Land Surface Evaluation for Engineering Practice

G e o l o g i c a l Society Special P u b l i c a t i o n s
Series Editor
J. S. GRIFFITHS

Special Publication reviewing procedures

The Society makes every effort to ensure that the scientific and production quality of its books matches that of its
journals. Since 1997, all book proposals have been refereed by specialist reviewers as well as by the Society's
Publications Committee. If the referees identify weaknesses in the proposal, these must be addressed before the
proposal is accepted.
Once the book is accepted, the Society has a team of series editors (listed above) who ensure that the volume
editors follow strict guidelines on refereeing and quality control. We insist that individual papers can only be
accepted after satisfactory review by two independent referees. The questions on the review forms are similar
to those for QuarterlyJournalof Engineering Geologyand Hydrogeology. The referees' forms and comments must be
available to the Society's series editors on request.
Although many of the books result from meetings, the editors are expected to commission papers that were not
presented at the meeting to ensure that the book provides a balanced coverage of the subject. Being accepted for
presentation at the meeting does not guarantee inclusion in the book.
Geological Society Special Publications are included in the ISI Science Citation Index, but they do not have an
impact factor, the latter being applicable only to journals.
More information about submitting a proposal and producing a Special Publication can be found on the Society's
web site: www.geolsoc.org.uk.

It is r e c o m m e n d e d t h a t reference to all o r p a r t o f this b o o k s h o u l d be m a d e in one o f the

f o l l o w i n g ways:

GRIFFITHS, J. S. (ed.) 2001. Land Surface Evaluation for Engineering Practice. G e o l o g i c a l

Society, L o n d o n , E n g i n e e r i n g G e o l o g y Special P u b l i c a t i o n s , 18.

FOOKES, P. G. & SHILSTON, D. T. 2001. Building the geological m o d e l : case s t u d y o f a r o c k

t u n n e l in S W E n g l a n d . In: GRIFFITHS, J. S. (ed.) Land Surface Evaluation for Engineering
Practice. G e o l o g i c a l Society, L o n d o n , E n g i n e e r i n g G e o l o g y Special P u b l i c a t i o n s , 18, 123-128.
Geological Society Engineering Geology Special Publication No. 18

Land Surface Evaluation for

Engineering Practice

EDITED BY

J. S. G R I F F I T H S
Department of Geological Sciences
University of Plymouth
Drake Circus
Plymouth
Devon PL4 8AA

2001
Published by
The Geological Society
London
 THE GEOLOGICAL SOCIETY
The Geological Society of London was founded in 1807 and is the oldest geological society in the world. It received its
Royal Charter in 1825 for the purpose of 'investigating the mineral structure of the Earth' and is now Britain's national
society for geology.
Both a learned society and a professional body, the Geological Society is recognized by the Department of Trade and
Industry (DTI) as the chartering authority for geoscience, able to award Chartered Geologist status upon appropriately
qualified Fellows. The Society has a membership of 9099, of whom about 1500 live outside the UK.
Fellowship of the Society is open to persons holding a recognized honours degree in geology or cognate subject, or not
less than six years' relevant experience in geology or a cognate subject. A Fellow with a minimum of five years' relevant
postgraduate experience in the practice of geology may apply for chartered status. Successful applicants are entitled to use
the designatory postnominal CGeol (Chartered Geologist). Fellows of the Society may use the letters FGS. Other grades of
membership are available to members not yet qualifying for Fellowship.
The Society has its own Publishing House based in Bath,UK. It produces the Society's international journals, books and
maps, and is the European distributor for publications of the American Association of Petroleum Geologists (AAPG), the
Society for Sedimentary Geology (SEPM) and the Geological Society of America (GSA). Members of the Society can buy
books at considerable discounts. The Publishing House has an online bookshop (http://bookshop.geolsoc.org.uk).
Further information on Society membership may be obtained from the Membership Services Manager, The Geological
Society, Burlington House, Piccadilly, London W1V 0JU (E-mail: enquiries@geolsoc.org.uk; tel: +44 (0) 207 434 9944).
The Society's Web Site can be found at http://www.geolsoc.org.uk/. The Society is a Registered Charity, number 210161.

Published by The Geological Society from: Distributors

The Geological Society Publishing House USA
Unit 7 Brassmill Enterprise Centre AAPG Bookstore
Brassmill Lane PO Box 979
Bath BA1 3JN, UK Tulsa
(Orders: Tel. +44 (0)1225 445046 OK 74101-0979
Fax +44 (0)1225 442836) USA
Online bookshop: http://bookshop.geolsoc.org.uk (Orders: Tel. +1 918 584-2555
Fax +1 918 560-2652
The publishers make no representation, express or E-mail bookstore@aapg.org
implied, with regard to the accuracy of the
information contained in this book and cannot accept Australia
any legal responsibility for any errors or omissions Australian Mineral Foundation Bookshop
that may be made. 63 Conyngham Street
Glenside
9 The Geological Society of London 2001. All rights South Australia 5065
reserved. No reproduction, copy or transmission of this Australia
publication may be made without written permission. (Orders: Tel. +61 88 379-0444
No paragraph of this publication may be reproduced, Fax +61 88 379-4634
copied or transmitted save with the provisions of the E-mail bookshop@amf eom.au
Copyright Licensing Agency, 90 Tottenham Court
Road, London W1P 9HE. Users registered with the India
Copyright Clearance Center, 27 Congress Street, Affiliated East-West Press PVT Ltd
Salem, MA 01970, USA: the item-fee code for this G- 1/ 16 Ansari Road, Daryaganj
publication is 0267-9914/00/$15.00. New Delhi 110 002
India
British Library Cataloguing in Publication Data (Orders: Tel. +91 11 327-9113
A catalogue record for this book is available from the Fax +91 11 326-0538
British Library. E-mail affiliat@nda, vsnl.net.in
ISBN 1-86239-084-3 Japan
ISSN 0267-9914 Kanda Book Trading Co.
Cityhouse Tama 204
Tsurumaki 1-3-10
Tauama-shi
Tokyo 206-0034
Typeset by Aarontype Ltd, Bristol BS5 0HE, UK. Japan
(Orders: Tel. +81 (0)423 57-7650
Printed by The Alden Press, Osney Mead, Oxford, UK. Fax +81 (0)423 57-7651
 Contents

Preface: the Second Working Party vii

Section 1: Introduction
J. S. GRIFFITHS 8z; R. J. G. EDWARDS: The development of land surface evaluation
for engineering practice

Section 2: Techniques in land surface evaluation

J. H. CHANDLER: Terrain measurement using automated digital photogrammetry 13
J. H. CHARMAN: Desk studies 19
J. H. CHARMAN, P. J. CAREY • P. F. FOOKES: Engineering classification for 23
environmental performance
C. N. EDMONDS: Predicting natural cavities in chalk 29
J. S. GRIFFITHS: Engineering geological mapping 39
G. J. HEARN & J. S. GRIFFITHS: Landslide hazard mapping and risk assessment 43
E. M. LEE: Geomorphological mapping 53
P. NATHANAIL & A. SYMONDS: Geographical information systems 57
P. J. PHIPPS: Terrain systems mapping 59

Section 3: Case studies in land surface evaluation

R. J. ALLISON: Surface and groundwater resources survey in Jordan 65
G. P. BIRCH: Mapping for high pressure gas pipelines in South Wales 73
G. P. BIRCH: Rapid evaluation of ground conditions for the AVE railway, Spain 83
J. H. CHARMAN: Preliminary landslide hazard assessment in remote areas 91
C. N. EDMONDS: Subsidence hazards in Berkshire in areas underlain by chalk karst 97
R. J. G. EDWARDS: Creation of functional ground models in an urban area 107
P. G. FOOKES, E. M. LEE & M. SWEENEY: Pipeline route selection and ground 115
characterization, Algeria
P. G. FOOKES & D. T. SHILSTON: Building the geological model: case study of a 123
rock tunnel in SW England
J. S. GRIFFITHS: Development of a ground model for the UK Channel Tunnel portal 129
G. J. HEARN: Low-cost road construction and rehabilitation in unstable mountain 135
areas
G. J. HEARN, R. BLON6 & G. HUMPHREYS: Terrain hazard around the Ok Tedi 143
copper mine, Papua New Guinea
G. J. HEARN, I. HODGSON & S. WODDY: GIS-based landslide hazard mapping in 151
the Scotland District, Barbados
D. K. C. JONES: Ground conditions and hazards: Suez City development, Egypt 159
D. K. C. JONES: Blowing sand and dust hazard, Tabuk, Saudi Arabia 171
E. M. LEE & D. BRUNSDEN: Sediment budget analysis for coastal management, 181
west Dorset
E. M. LEE & R. MOORE: Land use planning in unstable areas: Ventnor, Isle of 189
Wight
E. M. LEE & C. F. SAKALAS: Subsidence map development in an area of 193
abandoned salt mines
R. P. MARTIN: The design of remedial works to the Dharan-Dhankuta Road, 197
East Nepal
P. NATHANAIL:Terrain evaluation for military purposes: examples from the Balkans 205
D. N. PETLEY: Hazard assessment in eastern Taiwan 209
P. J. PHIPPS: Slope instability within a residential area in Cleveland, UK 215
P. J. PHIPPS: Ground models for the design and construction of a high-speed rail link 221
D. T. SHILSTON: Slope stability hazard assessment: Coalport Railway Bridge, 227
Shropshire
vi CONTENTS

D. T. SHILSTON, N. E. HARRISON & D. J. FRENCH: Evaluation in urban and 233

industrial environments: the Docklands Light Railway Lewisham Extension,
London

Section 4: Conclusions and Recommendations

D. BRUNSDEN & J. S. GRIFFITHS: Land surface evaluation: conclusions and 241
recommendations

Index 245
 Preface
The Second Working Party on Land Surface Evaluation for
Engineering Practice
The original Working Party on Land Surface Evaluation for Engineering Practice, under the
chairmanship of Mr R. J. G. Edwards, reported in 1982. Their report was presented in the Quarterly
Journal of Engineering Geology, Volume 15, pages 265-316, and was based primarily on work that had
been carried out during the previous decade. Since the original report was published there have been
considerable advances in the subject. Therefore, the Committee of the Engineering Group of the
Geological Society set up the Second Working Party in January 1997. The membership of the Second
Working Party comprised the following.
Dr J. S. Griffiths: Chairman: Head of the Department of Geological Sciences at the University of
Plymouth and member of Plymouth Environmental Research Group; Committee Member of the
Engineering Group of the Geological Society (1997-2003).
Mr R. J. G. Edwards: Chairman of the First Working Party; Consultant Engineering Geologist; Director
of Earth Science Partnership.
Professor D. Brnnsden: Emeritus Professor in Physical Geography at Kings College, London; Member of
the First Working Party; First President of the International Association of Geomorphologists; and the
Fifth Glossop Lecturer (2001).
Mr J. H. Charman: Consultant Engineering Geologist; Chairman of the Engineering Group 1996-98;
Member of the Council of the Geological Society (1998-2000).
Dr P. Nathanail: Senior Lecturer at Nottingham University; Chairman of the Environment Group of the
Geological Society (1997-99).
Mr W. Rankin: Divisional Director of Foundations and Geotechnics at Mott MacDonald; Chairman of
the Association of Geotechnical Specialists (1996-98); Chief Executive of AGS (1998-2000); Chairman
of Ground Forum (2000-2001).
Mr P. Phipps: Senior Engineer at Mott MacDonald; First Glossop Award Winner (1997); Committee
Member of the Engineering Group (1997-2000).
During the early meetings of the Working Party it became apparent that the report would have to take a
different form from the normal Engineering Group Working Party publications. It was decided that the
range and breadth of the subject would require input from a large number of specialists if it was to provide
a realistic view of the state of the art. The result was that practitioners in land surface evaluation known to
the members of the Working Party were invited to contribute short papers to an edited volume.
In the compilation of the final edited volume, substantial assistance was provided by Dr Gareth J.
Hearn of Scott Wilson Kirkpatrick, and Mr E. Mark Lee of the University of Newcastle, who, along
with Professor Denys Brunsden, acted as sub-editors.
All members of the Working Party and the sub-editors gave freely of their time in the preparation of
this report. The backing of employers in giving professionals the space and time on this, and similar
working parties, is too often taken for granted. As Chairman of the Working Party I wish to record the
debt of gratitude that I owe to the many individuals, companies and university departments who have
provided such support. In addition, I wish particularly to thank Mott MacDonald and Scott Wilson
Kirkpatrick for providing financial backing towards the cost of printing the final report, thus allowing us
to use colour in some of the figures.
Dr James S Griffiths
Chairman
University of Plymouth
October 2000
Geological Society, London, Engineering Geology Special Publications
About this title

Geological Society, London, Engineering Geology Special Publications 2001; v. 18; p. NP

doi:10.1144/GSL.ENG.2001.018.01.36

© 2001 Geological Society of London

Geological Society, London, Engineering Geology Special Publications
The development of land surface evaluation for engineering practice
J. S. Griffiths and R. J. G. Edwards

Geological Society, London, Engineering Geology Special Publications 2001; v. 18; p. 3-9
doi:10.1144/GSL.ENG.2001.018.01.01

© 2001 Geological Society of London

The development of land surface evaluation for engineering practice

J. S. Grittiths 1 & R. J. G. Edwards 2

1Department of Geological Sciences, University of Plymouth, Devon, U K
2 Earth Sciences Partnership, Leatherhead, Surrey, U K

Definition investigation for ground engineering'. The 1982 Report

identified the method as an integral part of the site
The First Working Party Report (Anon. 1982) defined investigation process typically undertaken at the re-
'Land surface evaluation' for engineering practice as: 'The connaissance and feasibility study stages of a project.
evaluation and interpretation of land surface features and The techniques were also seen as a specific method of
recorded surface data using one or a combination of the data collection and compilation for selected aspects
ground mapping, interpretation, classification and visual of investigation or monitoring at the design, construc-
remote sensing techniques outlined in this report'. The tion and post-construction stages. The Second Working
techniques outlined in the report were land classification, Party fully endorses this view but, on the basis of
remote sensing and geomorphological mapping. The identified practice, has identified that land surface evalu-
expression 'land surface evaluation' was adopted in pref- ation techniques also have a clear role to play in general
erence to 'terrain evaluation' or 'terrain classification' planning of engineering developments.
because the varied uses of the terms had created confu-
sion and led to misunderstandings. For this same reason
in this Second Working Party Report the expression
'land surface evaluation' has been maintained. Develop- Land surface evaluation techniques
ments since the 1982 Report was published require re-
vision to the definition of 'land surface evaluation'. The recommended by the 1982 Report
definition proposed by the Second Working Party is:' The
evaluation and interpretation of land surface and near The main land surface evaluation methods recom-
surface features using techniques that do not involve ground mended in the 1982 Report were geomorphological
exploration by excavation or geophysics'. This rather ground mapping and aerial photographic interpretation
broader definition allows land surface evaluation to be based on a framework of land classification. These
seen in its most common context as the process of data recommendations arose out of the experience gained on
compilation, interpretation and conceptual ground mod- projects undertaken during the previous four decades.
elling prior to undertaking engineering ground or site Of particular importance was the initial development of
investigation work. It therefore specifically includes the aerial photographic interpretative procedures (Belcher
integration of all existing ground information (desk 1948) and the subsequent geomorphologically based
study) whether it is surface or subsurface information. developments in land classification or terrain systems
The first objective of a land surface evaluation study is (Mitchell 1973). In engineering and resource survey
to acquire the most comprehensive conceptual ground projects the use of land classification based on terrain
model that can be generated in order to maximize the systems was pioneered by the Transportation Road
value and justify the cost of subsurface investigation Research Laboratory (TRRL) (Dowling 1968; Dowling
fieldwork and allied laboratory testing. The second & Bevan 1969). Other important contributions to the
objective is to minimize the engineering geological un- subject included: the development of land unit maps for
knowns that currently generate disproportionate unfore- data banks in South Africa (Brink & Partridge 1967;
seen contractual costs and not infrequently jeopardize Brink et al. 1968; National Institute for Road Research
the in-service facility or its design life. 1971); the PUCE terrain classification system in Aus-
tralia (Aitchison & Grant 1967, 1968); and the Oxford-
MEXE-Cambridge System of Classification (Beckett &
Webster 1969; MEXE 1969; Perrin & Mitchell 1969). The
Role of land surface evaluation MEXE (Military Experimental Establishment) system
was developed for military use and adopted by the
The 1982 Report stated that 'Land surface evaluation TRRL for the location of highways in the humid tropics
should be regarded and used as a normal method of (Lawrence 1978). It is interesting to note that, whilst not

From: GRIFFITHS,J. S. (ed.) Land Surface Evaluationfor Engineering Practice. Geological Society, London, Engineering Geology
Special Publications, 18, 3-9. 0267-9914/01/$15.00 9 The Geological Society of London 2001.
4 J. S. GRIFFITHS & R. J. G. EDWARDS

set out in specific detail, the value of these techniques was Developments in the UK since 1982
referred to in the original 1950 Civil Engineering Code of
Practice No. 1 updated for site investigations as CP2001 Throughout the 1980s and 1990s a significant number of
(British Standards Institution 1957). engineering case study applications of the recommended
Most of the initial land classification studies were techniques of land surface evaluation were reported in
undertaken at a reconnaissance level using relatively the literature (e.g. Jones et al. 1983; Fookes et al. 1985;
small-scale mapping (less than 1:25000). During the Griffiths & Marsh 1986; Charman & Griffiths 1993).
1970s larger scale mapping specifically for engineering However, it is in the Applied Earth Science Mapping
projects was introduced to facilitate conceptual ground programme carried out by various companies for the
modelling for planning site investigation work and it Department of the Environment (DOE) (e.g. Doorn-
also became established as an important tool in the kamp 1988; Forster et al. 1987; Wallace Evans Ltd 1994)
interpretation and presentation of site investigation that the most widespread use of land surface evaluation
output. A strong impetus to the wider use of mapping techniques can be found. These DoE studies demon-
techniques, particularly in engineering practice, was strated that the techniques could be directly applied in
provided by the Engineering Group Working Party planning development and for engineering and environ-
Report on maps and plans (Anon. 1972). Although not mental studies (Brook & Marker 1987; Dearman 1987;
specifically addressed, many of the recommendations for DoE 1991). The range of scale adopted for these studies
good practice that were included in the Working Party varied but commonly involved compilation of a suite of
Report involved the application of land surface evalua- 'earth science' maps at a scale of 1:25 000. One of the
tion techniques. During the 1970s a range of mapping few larger scale studies (maps at 1:2500 scale) was car-
techniques were developed for engineering practice using ried out for the assessment of landslide risk on the Isle of
both field mapping and aerial photographic interpreta- Wight at Ventnor (Lee et al. 1991). This programme of
tion. These developments included engineering geologi- applied earth science mapping has continued through
cal mapping (Dearman & Fookes 1974), geotechnical the 1990s, mainly carried out by the British Geological
mapping (Clark & Johnson 1975), engineering geomor- Survey, and there are approximately 50 such studies
phological mapping (Brunsden et al. 1975; Doornkamp available for areas of the UK (Smith & Ellison 1999).
et al. 1979), and refinements in the use of aerial photo- The DoE programme of earth science mapping and the
graphic interpretation for the analysis of engineering various detailed studies of ground instability (e.g. GSL
geological problems (Edwards 1968; Verstappen & Van 1987) formed the basis for the publication of PPG14 and
Zuidan 1968). PPG14 Annex 1 (DoE 1990, 1995). These notes repre-
The First Working Party Report, therefore, was sent the Government's planning guidance concerned
published at a time when there was cause for cautious with the granting of planning permission for construc-
optimism that the value and range of potential uses of tion on unstable ground.
land surface evaluation in engineering practice as an Whilst land surface evaluation techniques clearly
integral part of site investigations was being recog- made an important contribution to planning studies
nized. However, in 1981, BS5930 was published by the during the 1980s and 1990s, the published records of
British Standards Institution (1981) and this Code of their use in engineering practice has been disappointing.
Practice has formed the basis of site investigation for The development of engineering geomorphology in the
civil engineering in the UK for the 18 years up to the UK, seen in the 1982 Working Party Report as a central
publication of the revised code in 1999 (British Stand- component of land surface evaluation, was discussed in
ards Institution 1999). Although the importance of Griffiths & Hearn (1990). This study indicated that
desk studies and the need for rational site investiga- during the eight years since the publication of the First
tion planning was identified, BS5930:1981 clearly placed Working Party Report, the use of geomorphology in
the emphasis on ground investigation using explora- engineering practice had not been particularly wide-
tory holes, geophysics, and field and laboratory test- spread. The causes of this were complex. In part it was
ing. The first edition of BS5930 drew attention to the due to a lack of appreciation by engineers of how
use of aerial photography, and geomorphological and important and cost effective ground modelling is in the
engineering geological mapping, but it made no refer- development of investigatory works for engineering
ence to the importance of integrating this information projects. This was exacerbated by a continuing failure
using land surface evaluation techniques to generate amongst clients to recognize that money spent wisely in
conceptual or predictive ground models. This may partly the initial phases of project development pays massive
explain why the expansion in the use of land sur- dividends in overall site investigation, design, construc-
face evaluation techniques envisaged in the 1970s, and tion and maintenance costs. In part it was also due to a
recommended in the 1982 Report, appears to have lack of opportunity for such specialists to prove the
been rather limited, as indicated by the published case value of the technique because of failures in commu-
studies. nication between geomorphologists with a geographic
 THE DEVELOPMENT OF LAND SURFACE EVALUATION FOR ENGINEERING PRACTICE 5

background, and engineers. In addition, such specialists monitoring equipment to establish critical process para-
frequently did not have the training or understanding meters or seasonal variation in ground conditions that
of either geology or engineering to be able to generate have been identified as being of engineering significance.
the most useful information in a form that could be Since the First Working Party reported, therefore,
understood and adopted by engineers in the context of land surface evaluation techniques have continued to
engineering design and construction studies. develop, although in the UK this has mainly been in
Since 1990, there have continued to be occasional development planning. However, work carried out over
publications indicating that land surface evaluation the last two decades has demonstrated that land surface
techniques are being used for engineering studies and
that the available techniques have been developing Table 1. Names and affiliation o f contributors
rapidly (Hearn 1995a, b, 1997; Griffiths et al. 1996;
Waller & Phipps 1996; Edwards 1997; Fookes 1997). Professor R. Allison, Department of Geography, University of
Of particular importance are the changes in the remote Durham
sensing platforms that are able to provide data at scales G. P. Birch, Consultant Engineering Geologist, Sevenoaks,
that are suitable for engineering practice (e.g. Griffiths Kent
et al. 1994) with resolutions of +3 m now being com- Professor R. Blong, Natural Hazards, Research Centre,
Macquarie University, Australia
mercially available. Aerial photographic interpretation Professor Brunsden, Emeritus Professor of Physical Geography,
continues to provide the basis of land surface evaluation King's College, London
in many engineering studies (Verstappen 1983; Lawrence Dr P. Carey, School of Earth and Environmental Sciences,
et al. 1993). The modern techniques of analytical photo- University of Greenwich
grammetry (Chandler & Moore 1989; Chandler & J. H. Charman, Consultant Engineering Geologist, Milford,
Brunsden 1995) and digital photogrammetry (Brunsden Guildford, Surrey
& Chandler 1996) using both aerial and terrestrial Dr J. H. Chandler, Department of Civil and Building
photography have added an important new dimension Engineering, Loughborough University
to the technique. The full potential for using digitized R. J. G. Edwards, Earth Science Partnership, Leatherhead,
Surrey
airborne multispectral imaging systems for application to Dr C. N. Edmonds, Peter Brett Associates, Reading, Berkshire
civil engineering projects has yet to be realized despite the Professor P. G. Fookes, Consultant Engineering Geologist,
fact that it is universally recognized as a fundamental Winchester, Hampshire
tool in military engineering and has been for many years. D. J. French, W. S. Atkins Consultants Ltd, Epsom, Surrey
In conjunction with these developments the new Dr J. S. Grittiths, Department of Geological Sciences,
facilities available for data handling and analysis asso- University of Plymouth, Devon
ciated with the use of computers is comprehensive. The N. E. Harrison, W. S. Atkins Consultants Ltd, Epsom, Surrey
most immediately important development is the Geo- Dr G. J. Hearn, Scott Wilson Kirkpatrick & Co Ltd,
graphical Information System (GIS). The capability of Basingstoke, Hampshire
1. Hodgson, Scott Wilson Kirkpatrick & Co Ltd, Basingstoke,
GIS to handle, analyse and reproduce large volumes Hampshire
of spatial data, and then to permit revision and updating Dr G. Humphreys, School of Earth Sciences, Macquarie
of this data, is only just beginning to impinge on the University, Australia
world of civil engineering. It will provide a critical future E. M. Lee, Department of Marine Sciences and Coastal
development in promoting the scope and value of land Management, University of Newcastle, Newcastle upon Tyne
surface evaluation in engineering practice. Professor D. K. C. Jones, Department of Geography and
The unequivocal endorsement of conceptual ground Environment, London School of Economics, London
modelling, typically by the application of land surface Dr R. P. Martin, Geotechnical Engineering Office, Civil
evaluation techniques, presented appropriately in the Engineering Department, Government of the Hong Kong
Special Administrative Region
First Glossop Lecture (Fookes 1997), pointed to the fun- Dr R. Moore, Sir William Halcrow and Partners, Birmingham
damental principles set out in the 1982 Working Party Dr C. P. Nathanail, School of Chemical Environmental and
Report. Professor Fookes stated that field and labora- Mining Engineering, Nottingham University,
tory site investigation work normally should be a care- Nottinghamshire
fully planned, largely confirmatory study rather than an Dr D. N. Petley, Department of Geography, University of
exploratory exercise. The cost of such field and laboratory Durham
work and the type and range of techniques applied must P. J. Phipps, Mott MacDonald, Croydon, Surrey
be pertinent to the project development. Their deploy- W. Rankin, Mott MacDonald, Croydon, Surrey
ment typically should be concentrated where they are C. F. Sakalas, High-Point Rendel, Birmingham
D. T. Shilston, W. S. Atkins Consultants Ltd, Epsom, Surrey
needed to improve the existing intelligence database and M. Sweeney, BP-Amoco Exploration, London
to address 'problem' or 'unique' areas identified from the S. Woddy, Scott Wilson Kirkpatrick & Co Ltd, Basingstoke,
conceptual modelling. 'Type' areas typically only require Hampshire
confirmatory investigation work and the installation of
6 J . S . G R I F F I T H S & R. J. G. E D W A R D S

.a
 THE DEVELOPMENT OF LAND SURFACE EVALUATION FOR ENGINEERING PRACTICE 7

evaluation, particularly when used in conjunction with Section 2: A series of papers on techniques to be used in
comprehensive and effective desk studies, is one the most land surface evaluation.
cost-effective primary investigation techniques. One of Section 3: A collection of land surface evaluation case
its main strengths is in establishing conceptual ground studies from all types of engineering situations and
models that can be used for engineering purposes, as environments.
described by Fookes (1997). In this role, desk studies Section 4: A conclusion on the future role of land surface
developed around a land surface evaluation framework evaluation in engineering practice.
facilitate site investigation planning and enhance the
The names and affiliation of all contributory authors are
value of subsurface investigation work. The techniques,
presented in Table 1.
however, have been shown to have an ongoing role dur-
To facilitate use of the Report, Figure 1 presents a
ing construction to assist in the interpretation of ground
cross-tabulation of techniques, engineering situations
conditions. It is a matter of concern that the updated
and climatic or environmental zones against the authors
version of the Code o f Practice for Site Investigation
of papers presented in Section 3. References quoted by
(British Standards Institution 1999) failed to recognize
the authors can be found at the end of each paper.
or endorse this perspective by again placing the empha-
sis on actual ground investigations using exploratory
holes, geophysics, and field and laboratory testing, in the
same manner as BS5930: 1981. References
There are some important new developments in
techniques for data collection, compilation and analysis. AITCHISON, G. D. & GRANT, K. 1967. The P.U.C.E.
In land surface evaluation these will not replace tradi- programme of terrain description, evaluation and inter-
tional methods but will dramatically enhance the facil- pretation for engineering purposes. In: Proceedings of the
ities for information storage, manipulation, revision and Fourth Regional Conference in Africa on Soil Mechanics
presentation. This is an opportune time, therefore, to and Foundation Engineering (Cape Town), 1, 1-8.
AITCHISON, G. D. & GRANT, K. 1968. Proposals for the
undertake a review of the techniques of land surface application of the P.U.C.E. programme of terrain classi-
evaluation in the context of engineering practice and to fication and evaluation to some engineering problems. In:
compile working examples of good practice in order to Symposium on Terrain Evaluation for Engineering, Fourth
promote new standards of excellence for the next decade. Conference of the Australian Road Research Board, 4,
1648-1660.
ANON. 1972. The preparation of maps and plans in terms of
engineering geology. Report by the Geological Society
The Second Working Party Report Engineering Group Working Party. Quarterly Journal of
Engineering Geology, 5, 295-382.
The brief for the Second Working Party, set up in January ANON. 1982. Land surface evaluation for engineering practice.
1997, was to provide a revision of the First Working Party Quarterly Journal of Engineernig Geology, 15, 265-316.
Report incorporating the changes that had occurred since BECKET, P. H. T. & WEBSTER, R. 1969. A review of studies on
1982 and to provide a guide for best practice in the appli- terrain evaluation by the Oxford-MEXE-Cambridge group.
cation of land surface evaluation techniques. MEXE Report 1123, Christchurch, Hants.
Following extensive discussions it was decided that BELCHER, D. J. 1948. The engineering significance of land-
forms. Bulletin of the Highways Research Board, 13.
the wide range of techniques that had developed since BRINK, A. B. A. & PARTRIDGE, T. C. 1967. Kyalami land
the First Working Party precluded the development of a system: an example of physiographic classification for the
'manual of good practice'. The approach adopted for storage of terrain data. In: Proceedings of the Fourth
the report, therefore, was to invite state-of-the-art prac- Regional Conference for Africa on Soil Mechanics and
titioners in land surface evaluation to contribute papers Foundation Engineering (Cape Town), 1, 9-14
on techniques and engineering or planning case studies. BRINK, A. B. A., PARTRIDGE,T. C., WEBSTER,R. & WILLIAMS,
Thus the report is not intended to be an instruction book A. A. B. 1968. Land classification and data storage for the
that provides methodologies for applying all the vari- engineering use of natural materials. In: Proceedings of the
ous techniques because these are available in a range of Fourth Conference of the Australian Road Research Board,
1624-1647.
specialist publications. The aim is to outline the avail- BRITISH STANDARDS INSTITUTION. 1957. Site Investigations.
able methods of land surface evaluation, identify where British Standard Code of Practice CP2001. The Council
to find out details of how to use the techniques, and to for Codes of Practice, British Standards Institution,
provide examples of their application and output for a London.
range of engineering and environmental situationsl BRITISH STANDARDSINSTITUTION. 1981. Code of Practice for
The Second Working Party Report is divided into Site Investigations. BS5930: 1981. British Standards Insti-
four sections: tution, London.
BRITISH STANDARDSINSTITUTION. 1999. Code of Practice for
Section 1: An introduction on the development of land Site Investigations. BS5930: 1999. British Standards Insti-
surface evaluation for engineering practice. tution, London.
8 J. S. G R I F F I T H S & R. J. G. EDWARDS

BROOK, D. & MARKER, B. R. 1987. Thematic geological map- the field of engineering geology. MSc Thesis, Imperial
ping as an essential tool in land-use planning. In: CULSHAW, College, University of London.
M. G., BELL, F. G., CRIPPS,J. C. & O'HARA, M. (eds) Plan- EDWARDS R. J. G. 1997. A review of the hydrogeological
ning and Engineering Geology. Geological Society, London, studies of the Cardiff Bay Barrage. Quarterly Journal of
Engineering Geology Special Publications, 4, 211-214. Engineering Geology, 30, 49-62.
BRUNSDEN, D. & CHANDLER, J. H. 1996. The development of an FOOKES, P. G. 1997. Geology for engineers: the geological
episodic landform change model based upon the Black Ven model, prediction, and performance. Quarterly Journal of
mudslide 1946-1995. In: ANDERSON, M. J. & BROOKES, Engineering Geology, 30, 293-424.
S. M. (eds) Advances in Hillslope Processes. J. Wiley & FOOKES, P. G., SWEENEY, M, MANBY, C. N. D. & MARTIN, R. P.
Sons, Chichester, 2, 869-896. 1985. Geological and geotechnical engineering aspects of
BRUNSDEN, D., DOORNKAMP, J. C., FOOKES, P. G., JONES, low-cost roads in mountainous terrain. Engineering Geol-
D. K. C. & KELLY, J. M. H. 1975. Large-scale geomorph- ogy, 21, 1-152.
ological mapping techniques and highway engineering. FORSTER, A., HOBBS, P. R. N., WYATT, R. J. & ENTWlSLE,D. C.
Quarterly Journal of Engineering Geology, 8, 227-254. 1987. Environmental geology maps for Bath and the
CHANDLER, J. H. & BRUNSDEN, D. 1995. Steady state behav- surrounding area for engineers and planners. In: CUL-
iour of the Black Ven mudslide: the application of archival SHAW, M. G., BELL, F. G., CRIPPS, J. C. & O'HARA, M.
photogrammetry to studies of landform change. Earth Sur- (eds) Planning and Engineering Geology. Geological Soci-
face Processes and Landforms, 20, 255-275. ety, London, Engineering Geology Special Publications, 4,
CHANDLER, J. H. & MOORE, R. 1989. Analytical photogram- 221-236.
metry: a method for monitoring slope instability. Quarterly GRIFFITHS, J. S. & HEARN, G. J. 1990. Engineering geomorph-
Journal of Engineering Geology, 22, 97-110. ology: a UK perspective. Bulletin of the International
CHARMAN, J. H. & GR1FFITHS, J. S. 1993. Terrain evaluation Association of Engineering Geology, 42, 39-44.
methods for predicting relative hazard from mass move- GRIFFITHS, J. S. & MARSH, A. H. 1986. BS5930: the role of
ment, fluvial erosion and soil erosion in the developing geomorphological and geological techniques in a prelimin-
world. In: MERRIMAN, P. A. & BROWITT, C. W. A. (eds) ary site investigation. In: HAWKINS, A. B. (ed.) Site Investi-
Natural Disasters: Protecting Vulnerable Communities. gation Practice. Geological Society, London, Engineering
Thomas Telford, London, 167-183. Geology Special Publications, 2, 261-271.
CLARK, A. R. & JOHNSON, D. K. 1975. Geotechnical mapping GRIFFITHS, J. S., FOOKES, P. G., HARDINGHAM, A. D. ~r
as an integral part of site investigation two case studies. BARSBY, R. D. 1994. Geotechnical soils mapping for con-
Quarterly Journal of Engineering Geology, 8, 211-224. struction purposes in Central Saudi Arabia. In: FOOKES,
DEARMAN, W. R. 1987. Land evaluation and site assessment: P. G. & PARRY, R. H. G. (eds) Engineering Characteristics
mapping for planning purposes. In: CULSHAW, M. G., of Arid Soils. Balkema, Rotterdam, 69-85.
BELL, L. G., CRIPPS, J. C. d~; O'HARA, M. (eds) Planning GRIFFITHS, J. S., BRUNSDEN, D., LEE, E. M. & JONES, D. K. C.
and Engineering Geology. Geological Society, London, 1996. Geomorphological investigations for the Channel
Engineering Geology Special Publications, 4, 195-202. Tunnel portal and terminal. The GeographicalJournal, 161,
DEARMAN, W. R. & FOOKES, P. G. 1974. Engineering geo- 275-284.
logical mapping for civil engineering practice in the United GEOMORPHOLOGICALSERIES LTD. 1987. Review of research into
Kingdom. Quarterly Journal of Engineering Geology, 7, landsliding in Great Britain. Department of the Environ-
223-256. ment Volume Report, 14.
DEPARTMENT OF THE ENVIRONMENT. 1990. Development on HEARN, G. J. 1995a. Engineering geomorphological mapping
Unstable Ground. Planning Policy Guidance PPG14, and open-cast mining in unstable m o u n t a i n s - a case
HMSO, London. study. Transactions of the Institution of Mining and
DEPARTMENT OF THE ENVIRONMENT. 1991. Applied Earth Metallurgy, Section A, 104, A1-A17.
Science Mapping. HMSO, London, pp. 41. HEARN, G. J. 1995b. Landslide and erosion hazard mapping at
DEPARTMENT OF THE ENVIRONMENT. 1995. Development on Ok Tedi copper mine, Papua New Guinea. Quarterly
Unstable Ground." Landslides and Planning. Planning Policy Journal of Engineering Geology, 28, 47-60.
Guidance PPG14 Annex 1 HMSO, London. HEARN, G. J. 1997. Principles of low cost road engineering in
DOORNKAMP, J. C. (ed.) 1988. Applied Earth Science Back- mountainous regions, with special reference to the Nepal
ground." Torbay. Report to the Department of the Himalaya. Overseas Road Note 16, Transportation
Environment. Research Laboratory, Crowthorne.
DOORNKAMP, J. C., BRUNSDEN, D., JONES, D. K. C., COOKE, JONES, D. K. C., BRUNSDEN, D. & GOUDIE, A. S. 1983. A pre-
R. U. & BUSH, P. R. 1979. Rapid geomorphological assess- liminary geomorphological assessment of part of the
ments for engineers. Quarterly Journal of Engineering Geol- Karakoram highway. Quarterly Journal of Engineering
ogy, 12, 189-204. Geology, 16, 331-356.
DOWL1NG, J. W. F. 1968. The classification of terrain for road LAWRENCE, C. J. 1978. Terrain Evaluation in West Malaysia -
engineering purposes. Conference on Civil Engineering Part 2. Transportation Road Research Laboratory Report
Problems Overseas, Session V, Communications Paper 16, SR 378, TRRL, Crowthorne.
Institution of Civil Engineers, London. LAWRENCE, C. J., BYARD, R. J. & BEAVEN, P. J. 1993. Terrain
DOWLING, J. W. F. & BEVAN, P. J. 1969. Terrain evaluation for Evaluation Manual. Transportation Research Laboratory,
road engineers in developing countries. Journal of the State-of-the-Art Review 7, HMSO, London.
Institution of Highway Engineers, 16. LEE, E. M., DOORNKAMP,J .C., BRUNSDEN,D. & NOTON, N. H.
EDWARDS R. J. G. 1968. The use of aerial photographic 1991. Ground movement in Ventnor, Isle of Wight. Report to
interpretation for the analysis of engineering problems in the Department of the Environment.
 THE DEVELOPMENT OF LAND SURFACE EVALUATION FOR E N G I N E E R I N G PRACTICE 9

MEXE. 1969. Field trials of a terrain classification system. VERSTAPPEN, H. TH. 1983. Applied Geomorphology: Geomor-
MEXE Report 995, Christchurch, Hants. phological Surveys for Environmental Development. Else-
MITCHELL, C. W. 1973. Terrain Evaluation. Longmans, London. vier, Amsterdam.
NATIONAL INSTITUTEFOR ROAD RESEARCH. 1971. The produe- VERSTAVPEN, H. TH. & VAN ZUIDAN, R. A. 1968. Photo-
tion of soil engineering maps for roads and the storage of interpretation. ITC, Delft.
materials data. N I R R Technical Recommendations for WALLACE EVANS LTD. 1994. The presentation of earth science
Highways, 2. information for planning, development and conservation -
PERRIN, R. M. S. & MITCHELL, C. W. 1969. An appraisal of illustrated by a study of the Severn Levels. Summary Report
physiographic units for predicting site conditions in arid areas. to the Department of the Environment.
MEXE Report 1111, Vols I & II, Christchurch, Hants. WALLER, A. W. & PHIPPS, P. 1996. Terrain systems mapping
SMITH, A. & ELLISON, R, A. 1999. Applied geological maps for and geomorphological studies for the Channel Tunnel rail
planning and development: a review of examples for link. In: CRAIG, C. (ed.) Advances in Site Investigation
England and Wales, 1983 to 1996. Quarterly Journal of Practice. Thomas Telford, London, 25-38.
Engineering Geology, 32, Supplement, May 1999.
 THE DEVELOPMENT OF LAND SURFACE EVALUATION FOR E N G I N E E R I N G PRACTICE 9

MEXE. 1969. Field trials of a terrain classification system. VERSTAPPEN, H. TH. 1983. Applied Geomorphology: Geomor-
MEXE Report 995, Christchurch, Hants. phological Surveys for Environmental Development. Else-
MITCHELL, C. W. 1973. Terrain Evaluation. Longmans, London. vier, Amsterdam.
NATIONAL INSTITUTEFOR ROAD RESEARCH. 1971. The produe- VERSTAVPEN, H. TH. & VAN ZUIDAN, R. A. 1968. Photo-
tion of soil engineering maps for roads and the storage of interpretation. ITC, Delft.
materials data. N I R R Technical Recommendations for WALLACE EVANS LTD. 1994. The presentation of earth science
Highways, 2. information for planning, development and conservation -
PERRIN, R. M. S. & MITCHELL, C. W. 1969. An appraisal of illustrated by a study of the Severn Levels. Summary Report
physiographic units for predicting site conditions in arid areas. to the Department of the Environment.
MEXE Report 1111, Vols I & II, Christchurch, Hants. WALLER, A. W. & PHIPPS, P. 1996. Terrain systems mapping
SMITH, A. & ELLISON, R, A. 1999. Applied geological maps for and geomorphological studies for the Channel Tunnel rail
planning and development: a review of examples for link. In: CRAIG, C. (ed.) Advances in Site Investigation
England and Wales, 1983 to 1996. Quarterly Journal of Practice. Thomas Telford, London, 25-38.
Engineering Geology, 32, Supplement, May 1999.
Geological Society, London, Engineering Geology Special Publications
Terrain measurement using automated digital photogrammetry
J. H. Chandler

Geological Society, London, Engineering Geology Special Publications 2001; v. 18; p. 13-18
doi:10.1144/GSL.ENG.2001.018.01.02

© 2001 Geological Society of London

Terrain measurement using automated digital photogrammetry

J. H. Chandler
Department of Civil and Building Engineering, Loughborough University, Leicestershire, UK

Introduction represent the geometric characteristics of land and

terrain surfaces.
Photogrammetry has traditionally provided a means of
generating three-dimensional spatial data to represent
terrain surfaces, which complements traditional ground- Digital image acquisition
based surveying methods. Although techniques such as
airborne laser scanning (Lohr 1998) and synthetic aper- Digital photogrammetric methods rely upon the use of a
ture radar (Hogg et al. 1993; Vencatasawamy et al. 1998) digital image instead of the more traditional analogue
have developed, photogrammetry remains the primary contact diapositive. The digital image comprises a large
method of generating topographic maps (Wolf 1983; array of pixels, each representing a particular colour or
Capes 1998). One important advantage of photogram- grey-scale value to form the overall image. Space-borne
metry is the flexibility of scale that allows application imaging systems use push-broom sensors to create the
to imagery acquired from ground, air and space. Indeed, digital representation. For large and medium scales,
a new generation of high (i.e. 1 m) resolution satellite digital imagery can be obtained directly using a digital
sensors (Capes 1998) is likely to further increase the camera (Koh & Edwards 1996). Although digital camera
potential applications of photogrammetry. Despite many technology provides instant and appropriate imagery,
advantages, there have been several problems with the direct digital/vertical/aerial image acquisition is rarely
application of photogrammetry using traditional meth- used by the photogrammetric community (Maas & Ker-
ods. Most significantly, there was the requirement to use sten 1997). This is partly because digital cameras remain
an expensive and complex photogrammetric stereo- expensive, but, more crucially, such cameras only gener-
plotter. This ensured that the measurement process was ate images at comparatively low resolution (e.g. the
slow and generally required the skills of an experienced Kodak DCS460 camera costs s and provides an
operator, particularly if results of the highest accuracy image of only 3000 • 2000 pixels). The emulsions used in
were to be obtained. conventional aerial films represent an extremely efficient
Rapid developments in computing hardware and soft- means of storing image data at resolutions equivalent to
ware have allowed the science of photogrammetry to 50 000 • 50 000 pixels. Currently the most cost-effective
develop rapidly during the last ten years (Gruen 1994; means of obtaining digital imagery involves a hybrid
Atkinson 1996; Greve 1996). These developments have approach (Helava 1988) in which the conventional ana-
radically eased many of the problems and limitations logue photograph is converted into digital form using
associated with traditional analogue instrumentation. some form of scanning process. Scanning options and
Use of a purely numerical or analytical solution pro- costs vary widely, but to enable simplified processing
vides flexibility, which assists in two important ways. using full-format aerial images, a purpose-built geome-
Satellite imagery, oblique aerial photography and trically stable scanner should be used, (Warner et al.
ground-based imagery can be used, in addition to the 1996). These are expensive to buy, but bureau scanning
more traditional vertical aerial perspective. Similarly, services provide a cheaper alternative if production vol-
imagery acquired using a variety of non-photogram- ume is low.
metric cameras can be considered to be of value for It should be remembered that converting a photo-
spatial measurement. The most recent advance, known graph into digital form requires significant volumes of
as digital photogrammetry, now allows part of the file storage, which can cause serious system manage-
measurement process to be fully automated. This signifi- ment issues. For example, a normal black and white
cant development ensures that photogrammetry repre- aerial photograph (dimensions 230 • 230mm) scanned
sents an even more versatile and efficient method of at 25 micron resolution (1016dpi) with 256 grey levels
deriving dense digital elevation models (DEMs) to will generate an image consisting of 9200 x 9200 pixels,

From: GRIFFITHS,J. S. (ed.) Land Surface Evaluation for Engineering Practice. Geological Society, London, Engineering Geology
Special Publications, 18, 13-18. 0267-9914/01/$15.00 9 The Geological Society of London 2001.
 o

c~

o
o

o
ol~q

o
 TERRAIN MEASUREMENT USING AUTOMATED DIGITAL PHOTOGRAMMETRY 15

occupying 81 Mb of disk space. A minimum of two Application

images is needed to carry out three-dimensional mea-
surement and if colour images are scanned then these files Photogrammetric software packages, available commer-
need to increase in size by a factor of three. The manage- cially, have been developed for application using either
ment issues raised by working with digital imagery are stereo-satellite imagery (i.e. SPOT) or traditional vertical
described more fully by Colomer & Colomina (1994). aerial photography. One comparatively routine applica-
tion of automated digital photogrammetry is illustrated
by the work of Brunsden & Chandler (1996) who acquired
a new epoch of vertical aerial photography (photo scale
Image measurement and automatic 1:4000) of the Black Ven landslide, Dorset, UK, in
DEM generation March 1995. Automated methods were used to generate
a DEM consisting of 1 000 000 points within an area of
One of the important advantages of digital photogram- 1250 • 800m, a sampling density of one point every
metry is the replacement of the complex, specialized and metre (Fig. 1). Once a DEM had been created, it was pos-
expensive photogrammetric plotter with a more general, sible to create contours (Fig. 1), cross-sections and ortho-
and hence far cheaper, digital computer. Machines photos (Fig. 2). The orthophoto is particularly valuable
running the UNIX operating system have been widely for Earth scientists because it combines the interpretative
used in the recent past, but PC-based platforms are capabilities of the original aerial photograph with the
becoming ever more powerful and capable of manip- positional relevance of a map. The 1995 DEM of the
ulating the large images required. A competent UNIX Black Ven system was used to update a sequence of five
machine can be purchased for costs as little as s a lower resolution manually measured DEMs, which rep-
PC for less. Appropriate photogrammetric software is resented the morphology of the mudslides every ten years
required but competition for this market and the influ- since 1946 (Chandler & Cooper 1989). The improved
ence of software packages designed for remote sensing spatial resolution in combination with climatic and land-
applications have reduced costs significantly (e.g. Erdas slide incidence data allowed the revision of an evolu-
Imagine/OrthoMax and OrthoBase, PCI/EASI-PACE, tionary model (Brunsden & Chandler 1996).
R-WEL/Desktop Mapping System, VirtuoZo). Prices Although application using vertical aerial photo-
vary widely, currently within a range between s graphs will always remain most important, it is signifi-
and s 000. cant that automated digital photogrammetry can be
The second and perhaps most significant advantage of applied to both ground-based and terrestrial imagery. To
digital photogrammetry is the potential to automate ensure that the automated DEM extraction software
various aspects of the measurement process. The essen- remains successful, it is necessary to introduce an extra
tial capability relevant for land surface measurement is stage in the photogrammetric processing and although
the automated measurement of digital elevation models this is not documented in software user manuals, it
from an overlapping stereo-pair, which is now both is possible (Chandler 1999). One application of this
practicable and in an advanced state of development. approach is reported by Pyle et al. (1997) and involved
Such automation is based upon sophisticated image the creation of DEMs representing riverbanks using
correlation or image matching techniques that automa- ground-based oblique photography. The objective was to
tically identify and measure common image patches compare successive DEMs in order to map the spatio-
appearing on two overlapping digital images. Once temporal pattern of bank erosion. The precision of auto-
matched, these two image measurements are transformed matically generated DEMs was + 1 2 m m and, signifi-
into object coordinates using established photogram- cantly, the locations where individual clasts had been
metric methods, and the process repeated. With appro- removed from the gravel riverbank could be identified.
priate hardware and software, this cycle can recur at The same methods have been used to quantify the three-
speeds in excess of 100 points per second and so very dimensional form of exposed and subaerial riverbed
dense and consequently accurate DEMs can be gener- gravels, both in natural riverbed gravels (Butler et al.
ated. In most packages, the derived DEM is in the form of 1998) and in a flume (Stojic et al. 1998).
a regular grid of elevation estimates draped over the
desired area. This ability to measure very dense and
regular grid DEMs has instigated a return to grid-based Limitations
methods of manipulating and presenting height informa-
tion. This will perhaps reverse the trend towards using When considering using digital photogrammetric tech-
DEM processing methods based upon the Delauney niques, it is important to consider some of the require-
triangulation (Petrie & Kennie 1990). This algorithm ments and limitations associated with the method.
efficiently creates a surface from a limited sample of Although not theoretically essential, introducing photo-
terrain elevations in which the break-line assumes great control points into the object space eases photogram-
significance. metric processing significantly and should always be
 oo

~m

o
0
O~
o
,.~

0
r4
 TERRAIN MEASUREMENT USING AUTOMATED DIGITAL PHOTOGRAMMETRY 17

considered. Photo-control points are simply recognizable Developments in software implementing digital photo-
features that appear on the photography, and are situated grammetric methods are significant and readers are
at known geographical locations. Such points provide the encouraged to consider using such packages. The soft-
means to establish the transformation from measure- ware is also 'user-friendly', which allows the inex-
ments on the image, to positions in the desired ground perienced novice user to obtain results. Despite this, it is
coordinate system. The technology available to carry out perhaps important to conclude with several caution-
photo-control surveys has advanced in recent years (Lane ary comments. First, it is necessary to become famil-
et al. 1998). For surveys of small areas using terrestrial iar with the procedures briefly outlined in this paper.
photogrammetry, the modern total station is most effec- More guidance is provided by photogrammetry text-
tive, particularly if combined with an intersection method books (Wolf 1983; Atkinson 1996; Greve 1996) and
of surveying. Over larger areas and using vertical aerial other papers (i.e. Chandler 1999). It is easy to become
imagery, the Global Positioning System (GPS) (Schofield overly ambitious in terms of the size and number of
1994) is particularly appropriate, mainly because GPS DEMs that can be generated in a project. It is prudent to
does not rely upon maintaining 'line of sight' between start with a small area and extend only when experience
points. has been gained. Finally, it must be recognized that it
Although automation afforded by digital photogram- takes considerable care and some expertise to generate
metry provides a distinct advantage, such automation truly accurate surfaces on a routine basis. It is always
can also be interpreted as a weakness. The software will essential to assess the accuracy of generated surfaces
always produce some form of surface, but this may not using independent methods.
necessarily represent the surface that the user requires.
If, for example, the parameters used to control the deriva-
tion of the D E M are inappropriate, then erroneous sur-
faces will be generated (Smith et al. 1996). In vegetated References
areas, the surface generated automatically will represent
ATKINSON, K. B. 1996. Close Range Photogrammetry and
the tops of all visible vegetation and not the underlying
Machine Vision. Whittles, Caithness.
ground surface, which the user may perhaps require. BRtrNSDEN, D. & CHANDLER,J. H. 1996. The development of
Automated D E M acquisition is possibly best suited to an episodic landform change model based on the Black
situations in which there is either a lack of vegetation or Ven mudslide 1946-95. In: ANDERSON, M. J. & BROOKS,
where vegetation is homogenous in terms of foliage S. M. (eds) Advances in Hillslope Processes. John Wiley &
height above the true ground surface. Sharp disconti- Sons, Chichester, 2, 869-898.
nuities in the land surface can also downgrade final BUTLER, J., LANE, S. N. & CHANDLER, J. H. 1998. DEM
surface representation, particularly where resolution of Quality assessment for surface roughness characterisa-
the sampling points is low. It may then become neces- tion using close-range photogrammetry. Photogrammetric
Record, 16(92), 271-291.
sary to measure the three-dimensional characteristics of
CAPES, R. 1998. Developments in earth observation - how new
such discontinuities directly, using break-lines (Petrie & space-borne sensors will affect mapping. Surveying World,
Kennie 1990). 6(6), 24-27.
Many packages provide the opportunity to assess the CHANDLER, J. H. 1999. Effective application of automated
accuracy of points measured automatically using stereo- digital photogrammetry for geomorphological research.
superimposition (Smith et al. 1996). This practice is Earth Surface Processes and Landforms, 24, 51-63.
valuable but although such packages provide the oppor- CHANDLER,J. H. & COOPER, M. A. R. 1989. The extraction of
tunity to edit and correct erroneous height estimates, positional data from historical photographs and their
this process can become impracticable if the density of application in geomorphology, Photogrammetric Record,
generated data is high. 13(73), 69-78.
COLOMER, J. L. 8~;COLOMINA,I. 1994. Digital photogrammetry
at the Institut Cartogr~ific de Catalunya. Photogrammetric
Record, 14(84), 943-956.
GREVE, C. 1996. Digital Photogrammetry: an addendum to the
Conclusion manual of photogrammetry. American Society of Photo-
grammetry and Remote Sensing, Bethesda.
Digital elevation models can now be generated auto- GRUEN, A. 1994. Digital close-range photogrammetry- progress
matically using digital photogrammetry. Such automa- through automation. International Archives of Photogram-
tion has allowed the density of data used to represent metry and Remote Sensing, 5, 122-135.
terrain morphology to be increased by a factor of over HELAVA,U. V. 1988. On system concepts for digital automation.
International Archives of Photogrammetry and Remote
100 compared to manual photogrammetric measure- Sensing, 27/2, Kyoto, 171-190.
ment methods. This development is of distinct value to HOGG, J., MCCORMACK, J. E., ROBERTS, S. m. & GAHEGAN,
terrain measurement because it replaces the tedious and CHANDLER, J. H.M.N. 1993. Automated derivation of
expensive traditional manual procedures and generates a stream channel networks and selected catchment charac-
far higher resolution DEM. teristics from digital elevation models. In: P. M. MATTER
18 J . H . CHANDLER

(ed.) Geographical Information Handling- Research and PYLE, C. J., RICHARDS,K. S. & CHANDLER,J. H. 1997. Digital
Applications. John Wiley & Sons, Chichester, 207-235. photogrammetric monitoring of river bank erosion. Photo-
KOH, A. & EDWARDS, E. 1996. Integrating GPS data with Fly- grammetric Record, 15(89), 753-763.
on-demand digital imagery for coastal zone management. SCHOFIELD, W. 1994. Engineering Surveying. Butterworth-
Association of Geographic Information 1996 Conference Heinemann, Oxford.
Proceedings, 6.1.1-6.1.5. SMITH, M. J., SMITH, D. G. & WALDRAM, D. A. 1996.
LANE, S. N., CHANDLER, J. H. & RICHARDS, K. S. 1998. Experiences with analytical and digital stereoplotters.
Landform monitoring, modelling and analysis: landform Photogrammetric Record, 15(88), 519-526.
in geomorphological research. In: LANE, S. N., RICHARDS, STOJIC, M., CHANDLER, J. H., ASHMORE, P. & LUCE, J. 1998.
K. S. & CHANDLER, J. H. (eds) Landform Monitoring, The assessment of sediment transport rates by automated
Modelling and Analysis. John Wiley & Sons, Chichester, digital photogrammetry. Photogrammetric Engineering and
1-17. Remote Sensing, 645, 387-395.
LOHR, U. 1998. Digital elevation models by laser scanning, VENCATASAWAMY,C. P-, CLARK, C. D. & MARTIN, R. J. 1998.
Photogrammetric Record, 16(91), 105-109. Landform and lineament mapping using radar remote
MAAS, H.-G. & KERSTEN, T. 1997. Aero-triangulation and sensing. In: LANE, S. N., RICHARDS, K. S. & CHANDLER,
DEM/Orthophoto generation from high-resolution still- J. H. (eds) Landform Monitoring, Modelling and Analysis.
video imagery. On the potential of digital cameras onboard John Wiley & Sons, Chichester, 165-194.
an aircraft. Photogrammetric Engineering and Remote WARNER, W. S., GRAHAM, R. W. & READ, R. E. 1996. Small
Sensing, 63(9), 1079-1084. Format Aerial Photography. Whittles, Caithness.
PETRIE, G. & KENNIE, T. J. M. (eds). 1990. Terrain Modelling in WOLF, P. R. 1983. Elements of Photogrammetry. McGraw Hill,
Surveying and Civil Engineering. Whittles, Caithness. Singapore.
Geological Society, London, Engineering Geology Special Publications
Desk studies
J. H. Charman

Geological Society, London, Engineering Geology Special Publications 2001; v. 18; p. 19-21
doi:10.1144/GSL.ENG.2001.018.01.03

© 2001 Geological Society of London

D e s k studies

J. H. Charman
Consultant Engineering Geologist, Milford, Guildford, Surrey, UK

Introduction investigation programme. If carried out by a suitably

qualified and competent engineering geologist or geo-
In many, if not most, instances of civil engineering or technical engineer, it is arguably the most cost-effective
building projects that have suffered time or cost over- stage of the investigation.
runs or have required premature remedial works, the
reason can be attributed to geotechnical problems as a
result of inadequate planning or poor interpretation of
site investigation (Site Investigation Steering Group Sources of information
1993). The difficulties experienced in civil engineering
projects which can be related to the ground conditions Information that aids the implementation of the ground
occur either as a result of unawareness or as a result of investigation and the overall project includes planning
failure to grasp the implications of a certain set of ground and statutory restrictions, land ownership and access
conditions on the proposed engineering design (Fookes considerations, and access to utilities. These sources are
1997a). The site ground conditions pertaining at the given in BS5930:1999 and are not considered further
present are the result of a long history of geological pro- here. This paper deals with those sources relevant to the
cesses, from global tectonics, through climatic change evaluation of the terrain and, thus, determination of
to relatively recent landscape-forming processes. Major the ground conditions and sources of construction mate-
impact may also have been caused by man's historical rials. Sources of information include published maps
usage of the site. Much of this history of usage can be and memoirs, records from national and local govern-
derived by the specialist from existing scientific records ment archives, aerial photographs, scientific papers and
and published maps. The desk study is a fundamental records from other development activity in the area.
first step in any site investigation programme. Its pur- Of increasing significance is the availability of informa-
pose is to access published information and other avail- tion on the World Wide Web and references to useful
able records pertinent to the region, area and immediate web sites are included below.
environs of the project development site. This would in- For terrain evaluation per se, primary information
clude an investigation of geology, geomorphology, aerial includes topographic maps, aerial photographs, geologi-
photographs and other archival data. cal maps and soils maps. These all provide base data that
The recently published updated British Standard allow interpretation by the specialist. Secondary infor-
Code of Practice for Site Investigation, BS5930: 1999, is mation such as reports and archival material may already
unequivocal in stating that Stage 1 of a site investigation include some form of interpretation or assessment.
comprises a desk study and site reconnaissance and it In developing countries the available information is
should be undertaken at the start of every investigation largely primary. In the UK and other developed coun-
(British Standards Institution 1999). tries, many sources of secondary information also exist.
The sources of information in the UK are comprehen-
sively described in Perry & West (1996). A summary of
the sources of primary information is provided below.
Objectives
Topographic information
The main objective of the desk study is to carry out a
preliminary assessment of the ground conditions based Current mapping information in the UK is available
on existing information, to use this as a basis for a from the Ordnance Survey (www.ordsvy.gov.uk) at vari-
reconnaissance of the site and to plan the scope of the ous scales in both hard copy and digital form. Use-
ground investigation. This objective can be achieved at ful maps for terrain evaluation are the 1:50 000 scale
relatively low cost and at an early stage in the Landranger and 1:25 000 scale Pathfinder, Explorer and

From: GRIFFITHS,J. S. (ed.) Land Surface Evaluationfor Engineering Practice. Geological Society, London, Engineering Geology
Special Publications, 18, 19-21. 0267-9914/01/$15.00 9 The Geological Society of London 2001.
20 J . H . CHARMAN

Outdoor Leisure series. For larger scales, urban map- at a scale of 1:50000 cover most of the UK and are
ping at 1 : 1250, rural mapping at 1 : 2500 and mountain accompanied by sheet memoirs. More detailed mapping
and moorland mapping at 1:10 000 scale are available. at 1:10000 scale is also available and often contains
An examination of historic topographic maps is detailed field notes made by the mapping geologist.
essential to trace, for example, previous industrial The Geological Society has developed a new web site
usage (Charman & Cooper 1987). It is worth remember- (www.geolsoc.org.uk) which allows the library catalogue
ing that such maps are a factual record of the ground to be searched and also acts as a gateway to other Earth
surface at a single point in time. If, for example, a mine science sites for world wide information. The library is
shaft is infilled it ceases to feature on subsequent map one of the largest U K sources of geological information.
editions. Therefore, a sequential review of all historic
maps for the site should be made. The most comprehen-
sive collection is available at the British Library (www. Borehole information
bl.uk) and includes all Ordnance Survey map editions The BGS National Geological Records Centre has a
back to the 1800s. The British Library also contains comprehensive collection of site investigation and other
many other antiquarian maps as does the library of the borehole records, referenced to the relevant Ordnance
Royal Geographical Society (www.rgs.org). Survey 1:10 000 map series. A charge is made for this
In developing countries topographic maps are gen- information.
erally available from government survey departments.
Security considerations often govern their availability
and it is essential that a letter of authority is provided. Soil survey maps
Alternatively, for former colonies, maps are often avail- Pedological soil maps classify the soil in the top 1.5 m of
able in the former colonial power. the soil profile. The properties of this layer reflect the
properties of the deeper soil and rock profile. This is
Aerial photographs and satellite imagery particularly true in tropical areas of the world where
residual soils dominate.
Aerial photographs and satellite imagery are particularly In many parts of the world the coverage of pedo-
useful for the preparation of base maps and for terrain logical soil maps is more extensive than geological map
interpretation in undeveloped areas. They are often coverage. Pedological soil nomenclature is complex and
available for areas where other material is subject to several classification systems exist. The most widely used
security limitations and they can be used for those are those of the Food and Agricultural Organisation of
areas where access to the site is a problem. As with old the United Nations FAO/UNESCO (1988), the US Soil
map editions, where historic aerial photography is avail- Conservation Service and Duchaufour (1982; for areas
able the previous history of the site can be determined. developed under French influence). Fookes (1997b) pro-
In addition, some sources of satellite imagery provide an vides a summary of these systems as an aid to geotech-
archive of regular images for a particular site going back nical classification in tropical regions.
over many years.
Stereoscopic study techniques used by the specialist
remain the most effective tool for the initial assessment of
the terrain. They allow identification of many features Method of approach
including landforms, drainage patterns, geological boun-
daries, unstable ground and solution features. Advances The key to an efficient desk study is to focus clearly
continue to be made in the use of colour, multispectral on the type of information that is potentially relevant to
scanners and sensor systems. the project. For purposes of land surface evaluation it
The National Library of Air Photographs is probably is essential that a base map is prepared. Therefore, the
the biggest UK collection and is held by the Royal Com- first step is to access the primary sources of information,
mission on the Historic Monuments of England (www. i.e. topographic maps and aerial photographs. If these
rchme.gov.uk), now part of English Heritage. are unavailable then satellite imagery will provide a true-
For satellite imagery the National Remote Sensing to-scale base map at a price. A good base map, at a scale
Centre (www.nrsc.co.uk) and the World Wide Web vir- to suit the project site and the project needs, provides the
tual library of remote sensing (www.vtt.fi/aut/rs/virtual/) key to presenting and referencing other data sources.
are good places to begin a search for available images. Once this map is prepared other sources of information
can be meaningfully accessed and referenced, Geogra-
phical Information Systems (GIS) exist at various levels
Geological information
of sophistication. The simplest is a base map with anno-
Geological maps are available from the British Geolo- tated reference points and a series of derivative maps
gical Survey (BGS) (www.bgs.ac.uk). A full catalogue showing particular themes. Such themes may be factual,
and guide is available at the web site. Geological maps representing for example change in slope angle or depth
 DESK STUDIES 21

of shallow m i n i n g across a site. The theme m a y be an DUCHAUFOUR, P. 1982. Pedology, Pedogenesis and Classifica-
interpretation, representing changes in f o u n d a t i o n con- tion (English edition translated by T. R. Paton). George
ditions across the site for shallow foundations. D a t a Allen and Unwin, London.
FOOD AND AGRICULTURAL ORGANISATION OF THE UNITED
relevant to the reference points are best s u m m a r i z e d in
NATIONS. 1988. Soil Map of the Worm Revised Legend.
tabular form. T h e full range of thematic m a p p i n g exer- World Soil Resources Report No. 60, FAO, Rome.
cises carried out in the U n i t e d K i n g d o m based primar- FOOKES, P. G. 1997a. Geology for engineers: the geological
ily on available i n f o r m a t i o n is discussed in Smith & model, prediction and performance. Quarterly Journal of
Ellison (1999). Engineering Geology, 30, 294-424
FOOKES, P. G. (ed.) 1997b. Tropical Residual Soils. Geological
Society Professional Handbooks, The Geological Society,
References London.
PERRY, J. & WEST. G. 1996. Sources of information for site
BRITISH STANDARDSINSTITUTION. 1999. BS5930:1999 Code of investigations in Britain. TRL Report No. 192, Transport
Practice for Site Investigation. British Standards Institu- Research Laboratory, UK.
tion, London. SITE INVESTIGATIONSTEERING GROUP. 1993. Site investigation
CHARMAN, J. H. & COOPER, C. G. 1987. The Frindsbury area, in construction - Part 1: Without site investigation ground is
Rochester: a review of historical data and their implication a hazard. Thomas Telford, London.
on subsidence in an urban area. In: CULSHAW, M. G., SMITH, A. & ELLISON, R. A. 1999. Applied geological maps for
BELL, F. G., CRIPPS, J. C. & O'HARA, M. (eds) Planning planning and development: a review of examples from
and Engineering Geology. Geological Society, London, England and Wales 1983 to 1996. Quarterly Journal of
Engineering Geology Special Publications, 4, 115-124. Engineering Geology, Supplement, 32, Sl-$44.
Geological Society, London, Engineering Geology Special Publications
Engineering classification for environmental performance
J. H. Charman, P. J. Carey and P. G. Fookes

Geological Society, London, Engineering Geology Special Publications 2001; v. 18; p. 23-28
doi:10.1144/GSL.ENG.2001.018.01.04

© 2001 Geological Society of London

Engineering classification for environmental performance

J. H. Charman 1, P. J. Carey 2 & P. G. Fookes 3

1Consultant Engineering Geologist, Milford, Guildford, Surrey, UK
2 School of Earth and Environmental Sciences, University of Greenwich, Kent, UK
3 Consultant Engineering Geologist, Winchester, Hampshire, UK

Purpose of investigations This paper gives an example of a classification of the

environmental performance of natural and man-made
The condition of natural soils and rocks reflects the structures within a coastal environment. A classification
impact of a historical sequence of geological processes system is described which introduces a score rating for
(Fookes 1997), including plate tectonics, depositional each of a series of identified factors contributing to the
environment, structural and diagenetic change. Climate performance of the structure. The scheme should be
influences the effect of the atmosphere in producing readily adapted to other environments, if the variation
surface-related weathering and, in particular, the clima- in the geomorphological conditions and factors that
tic changes of the Quaternary have significantly modified influence them can be defined in a similar way.
the properties of near-surface soils and rocks. Such
modifications continue under present-day environmental
conditions.
Man-made or engineering structures are much young-
Basis of the performance model
er compared to the geological time-scale but they also The approach starts by identifying the factors that
undergo change as a result of the effect of the natural are important to the performance of each coastal struc-
environment in which they have been placed. The impact ture. These included geology, local coastal environment,
of these changes is governed by the materials that have
age of structure, material of construction, condition and
been used in the construction, the way in which they visual assessment of the frequency of repair. This formed
have been incorporated in the design, and the quality of the basis of a pro forma on which factors were either
the workmanship. The materials are not always suitable descriptive and given tick boxes or quantified through a
for the environment in which they have been placed. rating index. Each of the rating categories is summarized
An example is the use of pre-cast concrete with shallow below, but for points of detail the reader should refer
reinforcement cover in a coastal environment where it is
to Figure 1.
subject to seawater attack.
It follows, therefore, that engineering structures may
be placed in a wide variety of environments in which Definition of structure
chemical and physical attack will vary from harsh to
benign. For each environment the factors which may The subdivision of the structure into separate elements is
influence the rate of attack can be developed into a critical and must be developed to suit the particular
rating system which will allow the assessment of the output required. For example, the assessment of a
condition of the structure and its constituent materials. coastal section (Fig. 2a) may require subdivision into
This provides a framework for: slope or beach types and individual coastal protection
structures, e.g. groynes, sea-wall, etc. Assessment of an
9 development of a monitoring and maintenance pro- individual structure (Fig. 2b) will require subdivision
gramme; into its components, e.g. piers, footway, balustrade, etc.
9 possible changes to more appropriate design;
9 selection of more appropriate construction materials;
9 possible improvements in workmanship. Geology
The use of such rating systems in engineering geology Changes in geological material imply changes in per-
is well established; for example, the assessment of 'stand formance such as the rate of weathering, resistance to
up time' for underground excavations, the Rock Mass erosion and slope processes, and therefore must be rated
Classification (Bieniawski 1974)- has been in use for when assessing a coastal section where the geology
many years. changes. The rating has been based on the fundamental

From: GRIFFITHS,J. S. (ed.) Land Surface Evaluationfor Engineering Practice. Geological Society, London, Engineering Geology
Special Publications, 18, 23-28. 0267-9914/01/$15.00 9 The Geological Society of London 2001.
24 J . H . CHARMAN E T AL.

GENERAL LOCATION AND REFERENCE DETAILS.

1, STRUCTURE I Tick box

5. ENVIRONMENT Score
1A NATURAL COAST
1A1 Building coast 5A SUB -TIDAL 1
1A2 Eroding coast 5B INTERTIDAL 3-10
5C SUPRA-TIDAL 9-3
1B MAN MADE COAST 5D BACK BEACH 2
1B 1 GroUches 5E INLAND 1
1B2 Wave return revetment 1
1B3 Armoured breakwater
~k, _v H.W.M.
1B4 Sea wall ~ F 110
F~ 9 _v H.W.M.
1B5 Promenade/path v L.W.M. v L.W.M.
1B6 I Other

3
v H,W.M,
2. GEOLOGY (Add description) Score

~ -v~.L'W'M" 1
Widely
2A1 spaced
(>600ram)
Strong Mod.
(breaks spaced 6. COASTAL PROCESS Score
2A2
with (200-
hammer) 600mm) 6A LOW-LYING 1
Closely 6B DEGRADED SLOPE 2-5
2A3 spaced 6C ACTIVE DEGRADING SLOPE 5-9
(<200ram) 6D CLIFF 10
Rock
Widely
2A4 spaced
(>60Onto)
Mod. 7. FREQUENCY OF REPA1R Score
Weak
2A5 (breaks spaced 6A NONE 1
in hand) (200- 6B OCCASIONAL 2
600ram)
Closely 6C EVERY 10 YRS 3
2A6 spaced 6D EVERY 5 YRS 8
(<200ram) 6E EVERY YEAR 10
2A7 Fine
Soil
2A8 Coarse
8. RATE OF DETERIORA TION Score

7A CHANGE OVER >100 YEARS 1

3. MATERIAL OF CONSTRUCTION Tick box 7B CHANGE OVER 50 YEARS 2
7C CHANGE OVER 10 YEARS 3
3A STONE/MASONRY
7D CHANGE WITHIN 5 YEARS 8
3B CONCRETE
7E CHANGE WITHIN A YEAR 10
3C BITUMINOUS BOUND
3D TIMBER
3E METAL
3F BRICK 9, CONDITION Score
3G FILL
3H OTHER 8A VERY GOOD I
8B GOOD 3
8C ADEQUATE 5
8D POOR 7
4. AGE Score 8E VERY POOR 9
8F FAILED 10
4A PRE-HISTORIC 1
4B PRE 1900 4
4C 1900 -1940 6
10. PUBLIC SAFETY Score
4D 1940 -1960 7
4E 1960 -1980 8
4F 1980 -1990 9A NO RISK 1
4G 1990 - PRESENT 9B LOW RISK 3
9C MODERATE RISK 5
9D HIGH RISK 8
9E VERY HIGH RISK 10

Fig. 1. Ratings for classification pro formas.

 ENGINEERING CLASSIFICATION FOR ENVIRONMENTAL PERFORMANCE 25

(a)

GENERAL LOCATION AND REFERENCE DETAILS

1. STR UCTURE/CHA1NA GE 0-150m 150-650m 650-655m 655-1200m 1200-1425m

1A NATURAL COAST
1A1 Buildin~ coast
1A2 Erodin~ coast

1B MAN MADE COAST

1B 1 Groynes
1B2 Wave return revetment
1B3 Armoured breakwater
1B4 Sea wall
1B5 Promenade/path
1B6 Other

2.GEOLOGY

3, MATERIAL OF CONSTRUCTION

3A STONE/MASONRY
3B CONCRETE
3C BITUMOUS BOUND
3D TIMBER
3E METAL
3F BRICK
3G FILL
3H OTHER

4, AGE

5. ENVIRONMENT

6. COASTAL PROCESS

7. FREOUENCY OF REPAJR

8. RATE OFDETERIORATION

9. CONDITION

10. PUBLIC SAFETY

11. RECOMMENDED ACTION

REPAIR U R G E N T L Y
REPAIR RECOMMENDED
M O N I T O R CONTINUOUSLY
INSPECT A N N U A L L Y
INSPECT AFTER 5 YRS

(b)
Replace Section 1 (above) b y the following:

SEAWARD LANDWARD
1. STRUCTURE RAIL WALL STEPS PATHWAY
PIER PIER
1A I FOOTBRIDGE

Fig. 2. Examples of classification pro formas: (a) for coastal section; (b) for individual structure.
26 J. H. CHARMAN E T AL.

properties of strength and discontinuity spacing in rock of repair, for example, reflects the rate at which the
and particle size (coarse or fine) in soil (Fig. 1). structure has deteriorated and the condition is a measure
of the present serviceability.
Local environment
Public safety
The coastal environment has been subdivided into five
divisions, which reflect the vulnerability to wave attack This factor assesses the potential threat of deterioration
and exposure to salt weathering. The nature of the to public safety on the basis of severity of failure and
coastal profile influences the distribution of the subdivi- public accessibility.
sions (Fig. 1), and the pro forma may be enhanced in
some instances by the addition of a contoured plan. Recommended action
This system is currently being developed so that the
Material of construction scoring system can be calibrated against a number of
As with the in-situ soil or rock, the materials of construc- case histories. The total score is used to prioritize a
tion are fundamental to the durability of the structure. programme of remedial measures (Fig. 2)
Comparison of the different elements of the pro forma
will demonstrate those materials that perform best in
any particular environment, and may lead to recom- Location
mendations for the most suitable repair materials.
A classification scheme has been developed here using
two case studies from the south Devon coast. Both local-
Age, rate of deterioration, frequency of repair
ities form part of the itinerary for one of the University of
and condition
Greenwich MSc Geomaterials Field Courses and have
These factors are used in combination to provide a been developed from student exercises over several years.
measure of the durability of the structure or its fitness
for purpose. The age of the structure is important
Example of a natural coastal section
because it measures the time over which the structure
has been exposed. The rate of deterioration, for natural The coastal section area consists of fault-bounded
coastlines, is a measure of the rate of erosion, e.g. rate of blocks of Devonian limestones and shales, and Triassic
slope retreat. For man-made structures the frequency sandstones and conglomerates, forming cliffs up to

Fig. 3. Coastal section.

 E N G I N E E R I N G CLASSIFICATION FOR E N V I R O N M E N T A L P E R F O R M A N C E 27

GENERAL LOCATION AND REFERENCE DETAILS.

Footbridge over railway

RAIL ROCK
1. STRUCTURE LANDWARD PIER SEAWARD PIER PATHWAY
WALL ARMOUR

'Am ~~176176 I I I I mill

2. GEOLOGY
I I I I I I
3. MATERIAL OF CONSTRUCTION
3A STONE/MASONRY m m ~ m m m m u ~ ~ m
3B CONCRETE m m m m m m m n
3C BITUMOUS BOUND
3D TIMBER m m , ~ m _ ~ - -~, - -m~ ~ mm mm
3E METAL mm~mmm m|
3F BRICK m m m m m m~m
3G HLL
3H OTHER

4. AGE
14141419141,1414191,141 ~ 191
I 5. ENVIRONMENT
1515151,1517171~1717171 7 191
6. COASTAL PROCESS
IIIIlllllll I I I
7. FREQUENCY OF REPAIR

8. RATE OFDETERIORATION
IIIIIIlllll I I I
9. CONDITION

10, PUBLIC SAFETY

I TOTALS

11. R E C O M M E N D E D A C T I O N

REPAIR U R G E N T L Y
REPAIR RECOMMENDED
iim
M O N I T O R CONTINUOUSLY ml
INSPECT ANNUALLY
INSPECT AFTER 5 YRS ,/ I ~" ~" ~"

Hg. 4. Example o f a completed pro forma for a structure.

28 J. H. CHARMAN ET AL.

Fig. 5. Victorian footbridge, Dawlish.

100 m high (Fig. 3). It has been a popular amenity since mented. The general methodology has been used success-
Victorian times and there are several coastal defence fully, for example to survey marine concrete structures
structures of this age as well as a number of more recent in the Middle East (Fookes et al. 1981) and in the U K
slope remediation measures. The classification system for a Plymouth car park (Fookes et al. 1983a, b, 1984).
provides a framework for the assessment of risk, both to Each site requires adaptation of the general scheme to
the amenities and to the public, and the formulation of a suit local conditions and objectives.
remedial action plan.

Example of a man-made structure References

Brunel's railway between Exeter and Teignmouth, east BIENIAWSKI, I. T. 1974. Geomechanics classification of rock
of Dawlish, runs along a low embankment on the old masses and its application to tunnelling. Proceedings of the
foreshore and is protected by both sea wall and rock Third Congress of the International Society on Rock Mechan-
armour. The completed pro forma (Fig. 4) illustrates the ics, Denver, 1, 27-32.
assessment of one particular cross-section which includes FOOKES, P. G. 1997. Geology for engineers: the geological
a Victorian footbridge over the railway (Fig. 5). The pro model, prediction and performance. Quarterly Journal of
forma allows the evaluation of the relative durability of Engineering Geology, 3, 294-424.
FOOKES, P. G., POLLOCK,O. J. & KAY, E. A. 1981. Middle East
the various bridge and coastal rock armour materials for Concrete (2)- Rates of deterioration. Concrete, 15(9),
the effective design of future remedial measures. 12-19.
FOOKES, P. G., COMBERBACH,C. D. 8s CANN, J. 1983a. Field
investigation of concrete structures in South-west Eng-
land (1). Concrete, 17(3), 54-56.
Summary FOOKES, P. G., COMBERBACH,C. D. 8/; CANN, J. 1983b. Field
investigation of concrete structures in South-west England
The classification methods described here are a model of (2). Concrete, 17(4), 60-65.
what can be used by relatively inexperienced engineers FOOKES, P. G., CANN. J. & COMBERBACH, C. D. 1984. Field
and provide a rapid and cost-effective means for initial investigation of concrete structures in South-west Eng-
assessment so that a maintenance strategy can be imple- land (3). Concrete, 17(4), 12-16.
Geological Society, London, Engineering Geology Special Publications
Predicting natural cavities in chalk
C. N. Edmonds

Geological Society, London, Engineering Geology Special Publications 2001; v. 18; p. 29-38
doi:10.1144/GSL.ENG.2001.018.01.05

© 2001 Geological Society of London

Predicting natural cavities in chalk

C. N. Edmonds
Peter Brett Associates, Reading, Berkshire, U K

Introduction The research found that for natural cavities to be

formed, certain conditions needed to be present. Visual
Chalk is a soluble carbonate rock with extensive karst comparison of cavity occurrence versus 13 qualitative
development. Natural cavity occurrence initially appears geological, hydrogeological and geomorphological fac-
to be random. In an area where the degree of influence tors (Table 1) was undertaken (Edmonds et al. 1987) for
of all cavity formational factors is similar, but dissolu- the natural cavities database. The following points serve
tion is focused on one set of joints rather than another, to illustrate some of the main conclusions of the research.
then solution feature occurrence is perhaps random. (i) Ninety-seven per cent of natural cavities occur on
This might be termed the 'microscale' view, measured at the Upper Chalk, 2% on Middle Chalk and 1%
a scale of metres. However, if the pattern of natural on Lower Chalk.
cavity occurrence is considered at a 'macroscale' level, (ii) Lower than average numbers of natural cavities
say measured in hundreds of metres or kilometres, then are associated with the harder chalks of Lincoln-
spatial patterns emerge suggesting dissolution is not as shire and Yorkshire, and with the tectonically
random as it might first appear. hardened chalks (e.g. The Hogs Back, Surrey).
(iii) Ninety-seven per cent of natural cavities are asso-
ciated with the presence of a post-Cretaceous
cover deposit (such as Palaeogene and Quaternary
Spatial characteristics of natural cavity deposits). Only 3 % of features are isolated from a
occurrence on the Chalk cover deposit, but still generally occur within
200 m of a cover deposit margin.
In order to analyse the spatial characteristics of natural
cavity occurrence it was first necessary to collect as
many records of solution feature occurrence as pos-
sible from published and unpublished sources including Table 1. Geological hydrogeological and geomorphological
local authorities, Construction Industry Research and factors
Information Association (CIRIA), Transport Research Geological factors
Laboratory (TRL), Building Research Establishment 9 Lithostratigraphic horizons of the Chalk
(BRE), National House Building Council (NHBC), the 9 Biostratigraphic horizons of the Chalk
Environment Agency, water companies, and site inves- 9 Tectonic structure
tigation reports. Particular emphasis was placed on 9 Presence of post-Cretaceous cover deposits overlying
fieldwork to record new features revealed in a wide range the Chalk
of engineering works, road and motorway construction, 9 Lithology of post-Cretaceous cover deposits
and on visits to large numbers of working/disused chalk Hydrogeological factors
quarries and aggregate workings. A database of 2226 9 Hydrogeological characteristics of the Chalk
natural cavities was compiled, composed mainly of solu- 9 Hydrogeological characteristics of post-Cretaceous cover
tion pipes, sinkholes (dolines) and swallow holes. deposits
The spatially related database (each cavity loca- 9 Water table level in relation to the Chalk surface and cover
tion being recorded by National Grid Reference) was deposit interface levels
9 Effects of topographic relief upon surface water drainage and
carefully scrutinized to determine the conditions com- subsurface groundwater infiltration
monly associated with cavity occurrence (Edmonds
1987). The spatial analysis was undertaken with refer- Geomorphological factors
ence to the ideas previously put forward by Higginbot- 9 Locations of former surface water drainage paths
tom (1979) on factors that appeared to influence natural 9 Effects of marine planation
9 Effects of glaciation
cavity occurrence on the Chalk that were linked to subsi- 9 Effects of periglaciation
dence risk levels.

From: GRIFFITHS,J. S. (ed.) Land Surface Evaluation for Engineering Practice. Geological Society, London, Engineering Geology
Special Publications, 18, 29-38. 0267-9914/01/$15.00 9 The Geological Society of London 2001.
30 C . N . EDMONDS

(iv) Natural cavities are recorded below both cohesive deeply into the Chalk such that it is fully saturated or
and granular cover deposits up to 45 m in thick- water flows in the Chalk are upwards (artesian) below the
ness. glacial cover deposits. In these circumstances there are
(v) Ninety-two per cent of natural cavities are asso- little or no opportunities for the downward flow of water
ciated with surface water drainage and groundwater into the Chalk from the cover deposit, hence no natural
infiltration flows directed onto/into the Chalk from cavities are recorded in the unfavourable circumstances.
overlying Palaeogene and Quaternary deposits. In extraglacial areas it is considered that periglacial
(vi) The potential for natural cavity occurrence in- weathering has been generally beneficial to natural cavity
creases with the increasing tendency of topographic formation. In particular, it is thought that the downward
relief and surface water drainage/groundwater in- percolation of carbon-dioxide-rich, cold groundwater,
filtration conditions to concentrate water flows into released by the melting of ground ice formed within cover
the ground. The land surface may be subdivided deposits overlying the Chalk, is responsible for the
into water flow 'concentrators'. Fifty-three per cent formation and reactivation of many natural cavities.
of natural cavities are associated with concave It seems likely that this would have been the main source
surface channels, 28 % are associated with concave/ of water infiltration down into the Chalk that was
convex side slopes and 19~ are associated with responsible for natural cavity creation below cohesive
other land surface forms. cover deposits such as Reading Beds clays and Clay with
Flints (Plateau Drift).
The role of former drainage paths, especially the proto-
Thames and proto-Solent river systems, is thought to
have been influential in the formation of solution fea-
tures in the past. It is notable that higher numbers of Development of natural cavity occurrence
natural cavities are recorded within the Chilterns Hills (hazard) mapping techniques
and the Dorset and Purbeck Downs, where these
extensive river systems were formerly active. It appears The qualitative spatial analysis described above estab-
that they must have caused large volumes of water to lished substantial evidence for causal relationships
percolate down into the Chalk resulting in much between natural cavity occurrence and certain control-
dissolution activity. ling factors. Following this, the controlling factors were
Other influences, such as marine planation and glacial ranked into major, moderate and minor influential
erosion, can be both destroyers and creators of natural factors as shown in Table 2. These are the factors that
cavity-forming conditions. If marine or glacial erosion were consistently found to have the most pronounced
results in removal of a cover deposit and the chalk and specific influence on cavity formation.
surface zone containing natural cavities, then clearly The next step in the development of the hazard
natural cavities are destroyed and the numbers of natural mapping technique was to express the qualitative influ-
cavities greatly reduced. Conversely, if, following the ential factors in a quantitative way. It was decided to
erosion, a new cover deposit is laid down and the passage quantify major factors using a scale of 0 to 20, moderate
of concentrated water flows down into or onto the Chalk factors using a scale of 0 to 10, and minor factors using a
is resumed, then a new phase of active natural cavity scale of 0 to 5. The scale ranges were intended to reflect
formation commences. This is a feature of natural cavity the relative degree of influence of the different factors. The
occurrence noted in East Anglia associated with the higher the number, the higher the potential for natural
glacial cover deposits. However, in the Lincolnshire and cavity occurrence and associated instability hazard.
Yorkshire Chalk areas adjoining the present coastline, The specific numerical values for each of the influ-
the combination of marine and glacial erosion has etched ential factors are shown in Tables 3 to 8 for the varying

Table 2. Major, moderate and minor influential factors Table 3. Chalk lithostratigraphic factor (Gl)

Major influential factors Choose one Value

9 Chalk lithostratigraphy
9 Post-Cretaceous cover deposits Upper chalk 20
Middle chalk 2
Moderate influential factors Lower chalk 1
9 Water table level
9 Topographical relief and surface water drainage/subsurface Guidance note
groundwater infiltration 1. Choose the appropriate Chalk lithostrati-
9 Former surface water drainage path graphic unit that underlies the surface below
Minor influential factor the area of interest and carry forward the
9 Glaciation numerical value for insertion into the formula
in Table 9.
 PREDICTING NATURAL CAVITIES IN CHALK

Table 4. Post-Cretaceous cover deposit factor (G2) (note: G2 = Gc + Gr + Gf)

Gc Gr Gf

Choose one Value Choose one Value Choose one Va/ue

(a) Tertiary cover deposit 14 Tertiary cover deposit present: (a) Reading Beds feathering 3
present with or without (a) Reading Beds - Woolwich Beds 20 margin
superficial Quaternary (all regions) (b) Thanet Beds feathering 4
cover (b) Thanet Beds (WND, END 15 margin
(b) Quaternary cover deposit 6 regions) (c) Crag feathering margin 2
only present (c) Crag (EA region) 8 (d) Tertiary margins where 1
(c) No cover deposit present, 2 (d) Thicker Tertiary sequences 4 thick Tertiary sequences
but within 200 m of a involving Thanet Beds overlain occur (WND, DPD
Tertiary or Quaternary by Woolwich-Reading Beds, and regions)
cover deposit margin Blackheath Beds and disturbed (e) Quaternary deposit 2
(d) No cover deposit present 1 Blackheath beds (WND region) feathering margin
and >200m from a Tertiary (e) Thicker Tertiary sequences 2 (f) Exception to (e) for alluvial 0
or Quaternary cover deposit involving Reading Beds overlain deposits where seasonal
margin and overstepped by London Clay drainage is not directed
and Bagshot Beds (DPD region) across Chalk (Category 2)
of topographic relief and
Quaternary cover deposit present: surface drainage/
(a) Proto-Thames and Proto-Solent 20 subsurface infiltration
terrace gravels (CH and DPD factor (H2)
regions) (g) No cover deposit (topsoil 0
(b) Alluvial deposits (all regions) 13 only)
(only applicable to seasonal (h) No feathering margins 0
drainage directed across Chalk
(Category 2) of topographic relief
and surface drainage/subsurface
infiltration factor (H2)
(c) Alluvial deposits (all regions) 1
(any topographic relief and
surface drainage/subsurface
infiltration condition except
above)
(d) Glacial deposits (CH, EA, LY 12
regions)
(e) Low level fluvial terrace gravels 6
or valley gravel (all regions)
(f) High level fluvial terrace gravels 6
or plateau gravel (all regions)
(g) Clay-with-flints (all regions) 3
(h) Brickearth (all regions) 1
(i) Solifluction deposits (all regions) 1
(j) No cover (topsoil only) (all 0
regions)

Guidance notes
1. The value for factor G2 is calculated by summing together the values of the subfactors Gc, Gr and Gf.
2. For reference to the challdand regions see Fig. 1.
3. When determining Gr, the presence of a Tertiary deposit, even if overlain by Quaternary deposit, takes precedence when
choosing the appropriate Gr value.
4. For the area of interest, choose the one appropriate set of conditions from each of Gc, Gr and Gf in turn to derive the G2 value.
Carry forward the G2 value for insertion into the formula in Table 9.

circumstances as they apply to the Chalk. A n u m b e r subdividing the Chalk outcrop and subcrop in southern
of the tables m a k e reference to particular conditions and eastern England.
that apply to certain chalkland regions. To assist with The following account provides a s u m m a r y o f the
interpreting appropriate values for the factors, Figure 1 procedures for applying the hazard m a p p i n g techniques
shows the chalkland regions as used by the research for to a particular area of interest. In the first instance the
32 C. N. EDMONDS

Table 5. Water table level factor (1tl) i n f o r m a t i o n to derive a p p r o p r i a t e values for factors G1,
G2 a n d H1 (see Tables 3-5) m a y be obtained by referring
Choose one Value to published geological a n d hydrogeological maps for
the area o f interest. A p p l y i n g factor H2 to the area of
(a) No cover deposit present, water table level below 10 interest requires the area to be geomorphologically
Chalk surface level
m a p p e d to subdivide the land surface into terrain units
(b) No cover deposit present, water table close to or 1
at Chalk surface level that reflect the surface drainage/subsurface infiltration
(c) Cover deposit present, water table below Chalk/ 10 characteristics, the so-called w a t e r flow 'concentrators',
cover deposit interface referred to above. In order to visualize these character-
(d) Cover deposit present, seasonal water table level 5 istics, Figure 2 shows a series o f block models to d e m o n -
fluctuation causes water table level to rise above strate c o m m o n water flow scenarios f o u n d in the C h a l k
Chalk/cover deposit interface in wet season and Figure 3 demonstrates the principle of subdividing
(e) Cover deposit present, water table level normally 1 the landscape into m a p p a b l e water flow c o n c e n t r a t o r
at or above Chalk/cover deposit interface units. This w o r k , carried out by walkover survey sup-
(f) Same as (e), but if artificial groundwater lowering 3
ported by reference to t o p o g r a p h i c maps a n d aerial
to take place
(g) Artesian groundwater conditions present in cover 0 p h o t o g r a p h interpretation, forms a precursor to the
deposit overlying Chalk h a z a r d m a p finally produced. I n f o r m a t i o n on former
drainage paths (factor GM1) can be obtained from
Guidance notes published literature (e.g. Jones 1981) and the details for
1. The conditions described in (d) to (g) only apply where the GM2 can be derived from published geological maps.
cover deposit and the Chalk are in hydraulic continuity The numerical values q u o t e d in the tables for the
with one another. They do not apply to situations where a spatial controlling factors were only arrived at after
separate perched water level exists within the cover deposit m u c h detailed analysis a n d experimentation. Iterative
that is unrelated to the water table level within the Chalk
application, assessment and reassessment of the results,
below.
2. Carry forward the H1 value for insertion into the formula in when the factors were applied to the recorded natural
Table 9. cavities database, was executed. Wherever possible,

Table 6. Topographic relief and surface drainage~subsurface infiltration factor (H2)

Choose one category

Category 1 Category 2 Category 3

Seasonal/permanent surface drainage and Seasonal surface drainage and subsurface Seasonal/permanent surface drainage and
subsurface infiltration directed onto/ infiltration directed towards/onto cover subsurface infiltration directed across
into Chalk from cover deposit deposit from Chalk Chalk (may be covered by topsoil
and/or solifluction deposits or alluvium
in floors of seasonal/permanent
streams)

Choose terrain unit Value Choose terrain unit Value Choose terrain unit Value

Terrain Unit 1 6 Terrain Unit 1 1 Terrain Unit 1 l

Terrain Unit 2 10 Terrain Unit 2 2 Terrain Unit 2 3
Terrain Unit 3 4 Terrain Unit 3 1 Terrain Unit 3 1

Guidance notes
1. For the area of interest first consider the form of topographic relief and surface/subsurface water movement conditions that
would apply to the area. This should be carried out with reference to Figure 2 to derive the category in the above table that most
closely fits the situation. Support information will also need to be taken from the published geology, hydrogeology and
topographic survey maps. Stereoscopic viewing of aerial photographs can also be helpful.
2. The next step is to subdivide the area of interest into terrain units which determine the degree of surface/subsurface water
concentration. This exercise is carried out by performing standard geomorphological mapping of the area recording the concave
and convex surfaces. This is best done by field survey combined with the use of aerial photograph interpretation. Finally a map
of the terrain units is created for the area of interest, as illustrated by Figure 3. The recording of slope angles within terrain units
is not particularly important for this exercise, the landform type is more important to establish. However, for guidance Terrain
Unit 3 will have slope angles of zero to 1~ or 2 ~ while Terrain Unit 2 often has slope angles in the range of 2 ~ to 5~ sometimes up
to 10~
3. Carry forward the H2 values applicable and insert them, in turn, into the formula in Table 9 to calculate the SHRN values.
 PREDICTING NATURAL CAVITIES IN CHALK 33

Table 7. Former surface water drainage path factor (GM1) was over- or underestimated, hence it was necessary
to adjust certain values to overcome such problems.
Choose one Value Eventually, utilizing the results of the natural cavity
occurrence statistics from the research database, the fac-
(a) Proto-Solent corridor (DPD region) l0
(b) Proto-Thames corridor (CH region) but not 10 tors were suitably numerically weighted against the
applicable to following: chosen simple arithmetic scales. Finally, the derived
(i) where glacial deposits directly overlie Chalk 0 scale value for each factor is entered into the mathema-
or, tical formula shown in Table 9. In recognition that the
(ii) where surface drainage/subsurface infiltration 0 occurrence of a natural cavity gives rise to the potential
is directed off the Chalk towards/onto a for land instability, the formula is utilized to derive a
Tertiary cover subsidence hazard rating value, SHRN. Once the numer-
(c) Outside above proto-river corridors 0 ical value of SHRN has been calculated it is compared
with Table 10 to obtain the final subsidence hazard
Guidance notes
1. To check the geographic location and extent of the former classification.
proto-Thames and proto-Solent drainage path corridors it is By this methodology the range of possible subsidence
advised that reference is made to appropriate published hazard classifications that apply to the area of interest
texts, e.g.D.K.C. Jones, Southeast and Southern England, are calculated. The classifications are then attached to
University Paperbacks, Methuen, 1981. each of the spatial landscape units to which they apply in
2. For reference to the chalkland regions see Figure 1. order to produce the finished subsidence hazard map.
3. When H a ~--1 factor GMI is not applicable hence G M 1 ~ - 0 This is further explained with reference to a case study
(DPD, CH regions). where the techniques have been used to generate a subsi-
4. If the area of interest lies within a former drainage path
dence hazard map for an area of northwest Reading,
corridor then choose the appropriate situation to derive the
factor value. Transfer the value for insertion into the formula Berkshire (Edmonds 2001).
in Table 9.

Limitations
Table 8. Glaciation factor (GM2)
It was recognized at an early stage that the database
Choose one Value would show some bias towards urban areas where fea-
tures had been discovered as a result of man's activities.
(a) Glacial deposits directly overlie the Chalk 5 However, despite this it became apparent that in some
(CH, EA, L, Y regions) areas features were plentiful while in others they were
(b) Glacial deposits overlie Crag upon Chalk 3
(EA region) absent given similar opportunities for discovery by
(c) Glacial deposits overlie Reading Beds and/or 0 excavation. This was also reinforced by the field evidence
proto-Thames terrace gravels upon Chalk in non-urban areas. It was determined that the pattern
(CH region) of solution feature occurrence is not an artefact of
(d) Glacial deposits absent (CH, EA, L, Y regions) 0 man's activities and nor is it of anthropogenic origin
(e) Extra-glacial areas 0 (Edmonds 1984).

Guidance notes
1. To check the geographic location and extent of the formerly
glaciated areas of the Chalk it is advised that reference is Applications
made to appropriate published texts, e . g . D . K . C . Jones,
Southeast and Southern England, University Paperbacks, Since the completion of the academic research, the
Methuen, 1981. natural cavities database has been greatly enlarged. The
2. For reference to the chalkland regions see Figure 1. database forms the core of a national natural cavities
3. If the area of interest lies within a formerly glaciated area
then choose the appropriate situation to derive the factor database prepared for the Department of the Environ-
value. Transfer the value for insertion into the formula in ment (now the Department of the Environment, Trans-
Table 9. port and the Regions), the results of which were
published by Applied Geology Limited in 1994.
It has been found by hazard mapping that, in general,
the higher the hazard rating the greater the statis-
derived values were based closely upon the proportional tical chance of natural cavities occurring, thereby
influence shown by the recorded cavities database in increasing the potential for land instability as well (see
response to a particular spatial factor. However, some- Table 11). The technique has been applied to more than
times when derived values were reapplied to the data- 200 sites underlain by Chalk. Consequently the results
base the potential for natural cavity occurrence (hazard) of the work have been subjected to academic and
34 C . N . EDMONDS

~0 100
a Yonkshire(Y) I I I
b Uncolnshire(L) Scale
c EastAngtia(EA)
d ChilternHills (CH)
e Berkshire& MarlboroughDowns (BMD)
f SalisburyPlain (SP)
g HampshireDowns 0"t[3)
h WestNorth Downs (WND)
i EastNorth Downs (END)
j Thanet(1")
k SouthDowns (SD)
I L~lehampton(LO
rn Portsdown(P)
n Isle of Wight (lOW)
o CranbomeChase (CC)
p Dorset~d PurbeckDowns (DPD)
q Devon-DorsetOutliers (DDO)

Fig. 1. Plan of chalkland regions.

commercial scrutiny. The hazard mapping technique ing development, as an aid to determining remedial
has been found to provide a generally reliable way of measures. Currently, the technique is being adapted for
expressing cavity occurrence and subsidence potential. use as a mapping tool to recognize the potential for con-
The hazard mapping approach and its uses in investiga- tamination susceptibility of the Chalk aquifer due to
ting sites underlain by Chalk is outlined in Kirkwood & karstic landforms.
Edmonds (1989).
The technique has been found to be useful in
providing prior warning of potential subsidence pro-
blems and related effects in relation to new road Previous research
construction and the development of housing, commer-
cial and retail premises. On a number of occasions Research of karstic spatial patterns is not new: many
the method has been used to interpret and understand accounts have been published from at least the 1960s
patterns of ground subsidence behaviour affecting exist- onwards. For the Chalk, some of the earliest research
 PREDICTING NATURAL CAVITIES IN CHALK 35

Category 1,
Surface drainage and groundwaterinfiltrationoriginatesupon Tertiary/Quaternarycover deposits, thenflowsonto and Percolates
down into ohalk below.

Surface drainage and groundwaterinfiltrationoriginatesupon chalk, then flows down ontoand intoTertiary/Quaternarydeposits
at foot of slope.

Surface drainage and groundwaterinfiltrationoriginatesupon chalk, then is directed down slope and concentratedin chalk
valley floor with surface (bourne)flow8along and surfacepe'cc4ationIntothe fluvial/sotiftuctbnde.sits in fillingthe valleyfloor.
Category normallyapplicableto uplandchalk areas.

Fig. 2. Regional topographic relief and surface drainage/subsurface infiltration models.

in solution feature patterns was undertaken by Prince (1972), Williams (1972), Troester et al. (1984) and Thorp
(1961, 1962, 1964) in East Anglia and later by Sperling & Brook (1984). Other studies have analysed spatial
et al. (1977, 1979) and Prince (1979) for the many sink- patterns by visual comparison, relying strongly on the
holes found on the Dorset heathlands. Elsewhere, exam- use of remote sensing and/or geophysics (e.g. Black
ples of statistical studies include McConnell & Horn 1984; Littlefield et al. 1984; Stewart & Wood 1984).
36 C.N. EDMONDS

Fig. 3. Schematic illustration of simplified topographic relief and surface drainage/subsurface infiltration terrain
unit mapping.

Further studies have considered combinations of pos- 1984; Day 1984; Fischer & Greene 1984; Hubbard 1984;
sible controlling factors such as stratigraphy, lithology, Ogden 1984). Once spatial pattern controlling factors
structure, hydrogeology, topography and geomorphol- were identified they were combined qualitatively or
ogy using multilayered datasets for visual comparison semi-quantitatively to produce simple subsidence hazard
(e.g. Benson & la Fountain 1984; Dalgleish & Alexander predictive models.
 PREDICTING NATURAL CAVITIES IN CHALK 37

Table 9. Subsidence hazard mapping formula Table 11. Statistical relationship of cavity occurrence with
hazard category
Formula: SHRN = (G1 + G2 + H1 + GM1 + GM2)H2
Subsidence hazard Number of Percentage of
Factor identities for the formula are shown below:
category solution features solution features
SHRN = subsidence hazard rating for natural cavities
G1 = chalk lithostratigraphic factor No anticipated hazard 2 0.1
G2 = post-Cretaceous cover deposit factor Very low hazard 10 0.5
H1 =water table level factor Low hazard 62 3
H2 = topographic relief and surface drainage/subsurface Moderately low hazard 73 3.6
infiltration factor Moderate hazard 147 7.2
GM1 = former drainage path factor Moderately high hazard 244 11.9
GM2 = glaciation factor High hazard 466 22.8
Very high hazard 1042 50.9
Guidance note Totals 2046 100
1. Substitute factor values from Tables 3 to 8 into formula and
calculate SHRN value. Guidance notes
2. Compare SHRN value with Table 10. 1. The table shows the increasing numbers of solution features
(natural cavities) that are found within each of the ascending
hazard categories when the subsidence hazard mapping
technique is applied to the research database (2226 natural
Table 10. Subsidence hazard classification cavity records).
2. The total number of solution features is 2046, rather than
SHRN Subsidence hazard category 2226, because 180 features could not be located sufficiently
accurately to determine the hazard category that was
<55 No anticipated subsidence hazard applicable.
55-89 Very low subsidence hazard
90-136 Low subsidence hazard
137-200 Moderately low subsidence hazard
201-300 Moderate subsidence hazard
References
301-400 Moderately high subsidence hazard
APPLIED GEOLOGY LIMITED. 1994. Review of Instability due to
401-600 High subsidence hazard
Natural Underground Cavities in Great Britain. Vols 1.1
>600 Very high subsidence hazard
to 1.10, 2.1 to 2.3 and Summary Report for the Depart-
ment of the Environment.
Guidance notes
BENSON, R. C. & LA FOUNTAIN, L. J. 1984. Evaluation of
1. Use SHRN value to derive subsidence hazard category.
subsidence or collapse potential due to subsurface cavities.
2. Reapply the subsidence hazard category to the terrain unit
In: BECK, B. F. (ed.) Sinkholes: their Geology, Engineering
for which all derivative factors were originally obtained.
and Environmental Impact. Proceedings of the First Multi-
3. Repeat the procedure for all other terrain units within area
disciplinary Conference on Sinkholes, Florida, Balkema
of interest to create a subsidence hazard map of the area.
Press, 201-215.
BLACK, T. J. 1984. Tectonics and geology in karst development
of Northern Lower Michigan. In: BECK, B. F. (ed.) Sink-
holes: their Geology, Engineering and Environmental Impact.
A series of papers by D o u g h e r t y (1989), Fischer & Proceedings of the First Multidisciplinary Conference on
L e c h n e r (1989) and K o c h a n o v (1989) illustrate how, Sinkholes, Florida, Balkema Press, 87-91.
in the USA, n a t u r a l cavity occurrence m a p p i n g a n d CULSHAW, M. G., BELL, F. G., CRIPPS, J. C. & O'HARA, M.
h a z a r d zoning is being used for d e v e l o p m e n t land m a n - 1987. Planning and Engineering Geology. Proceedings of
the 22nd Annual Conference of the Engineering Group
a g e m e n t by statutory authorities in Pennsylvania and
of The Geological Society. Engineering Geology Special
N e w Jersey. Publications, 4, The Geological Society, London.
F u r t h e r examples o f modelling technique a p p r o a c h e s DALGLEISH, J & ALEXANDER,E. C. 1984. Sinkhole distribution
for deriving land instability h a z a r d ratings for areas in Winona County, Minnesota. In: BECK, B. F. (ed.) Sink-
p r o n e to mining subsidence, landsliding a n d limestone holes: their Geology, Engineering and Environmental Impact.
subsidence are presented by papers in C u l s h a w et al. Proceedings of the First Multidisciplinary Conference on
(1987). Similar ideas and a p p r o a c h e s to u n d e r s t a n d i n g Sinkholes, Florida, Balkema Press, 79-85.
subsidence b e h a v i o u r and risk on the C h a l k have been DAY, M. 1984. Predicting the location of surface collapse within
published m o r e recently by Rigby-Jones et al. (1993) a n d karst depressions: A Jamaican example. In: BECK, B. F.
(ed.) Sinkholes: their Geology, Engineering and Environ-
for a variety o f geohazards in M a u n d & Eddleston
mental Impact. Proceedings of the First Multidisciplinary
(1998). The development, application and usage of haz- Conference on Sinkholes, Florida, Balkema Press, 147-151.
ard m a p p i n g m e t h o d s by engineering geomorphologists DOUGHERTY, P . H . 1989. Land use regulations in the Lehigh
is becoming m o r e c o m m o n for evaluating a wide range Valley: Zoning and subdivision ordinances in an envir-
o f geohazards. onmentally sensitive karst region. In: BECK, B. F. (ed.)
38 C . N . EDMONDS

Engineering and Environmental Impacts of Sinkholes and B. F. (ed.) Sinkholes: their Geology, Engineering and Envi-
Karst. Proceedings of the Third Multidisciplinary Con- ronmental Impact. Proceedings of the First Multidisciplin-
ference on Sinkholes, St. Petersburg Beach, Florida, ary Conference on Sinkholes, Florida, Balkema Press,
Balkema Press, 341-348. 189-195.
EDMONDS, C. N. 1984. Reply to Discussion 'Towards the MAUND, J. G. & EDDLESTON,M. 1998. Geohazards in Engineer-
prediction of subsidence risk upon the Chalk outcrop' by, ing Geology. The Geological Society, London, Engineering
C. N. Edmonds. Quarterly Journal of Engineering Geology, Geology Special Publications, 15.
17, 167-168. McCoNNELL, H. & HORN, J. M. 1972. Probabilities of surface
EDMONDS, C. N. 1987. The engineering geomorphology of karst karst. In: CHORLEY, R. J. (ed.) Spatial Analysis in Geo-
development and the prediction of subsidence risk upon the morphology, Part H - P o i n t Systems. Methuen, London,
Chalk outcrop in England. PhD Thesis, University of 111-133.
London. OGDEN, m. E. 1984. Methods for describing and predicting the
EDMONDS, C. N. 2001. Subsidence hazards in Berkshire in areas occurrence of sinkholes. In: BECK, B. F. (ed.) Sinkholes:
underlain by Chalk karst. This volume. their Geology, Engineering and Environmental Impact. Pro-
EDMONDS, C. N . , GREEN, C. P. & HIGGINBOTTOM, I. E. 1987. ceedings of the First Multidisciplinary Conference on Sink-
Subsidence hazard prediction for limestone terrains, as holes, Florida, Balkema Press, 177-182.
applied to the English Cretaceous Chalk. In: CULSHAW, PRINCE, H. C. 1961. Some reflections on the origin of hollows in
M. G., BELL, F. G., CRIPPS, J. C. & O'HARA, M. (eds) Norfolk compared with those in the Paris region. Revue de
Planning and Engineering Geology. Proceedings of the Geomorphologie Dynamique, 12, 110-117.
22nd Annual Conference of the Engineering Group of PRINCE, H. C. 1962. Pits and ponds in Norfolk. Erdkunde, 16,
The Geological Society. Engineering Geology Special 10-31.
Publications, 4, The Geological Society, London, 283-293. PRINCE, H. C. 1964. The origin of pits and depressions in
FISCHER, J. A. & GREENE, R. W. 1984. New Jersey sinkholes: Norfolk. Geography, 49, 15-32.
Distribution, formation, effects- Geotechnical Engineer- PRINCE, H. C. 1979. Discussion- Marl pits or dolines of the
ing. In: BECK, B. F. (ed.) Sinkholes: their Geology, Dorset Chalklands. Transactions of the Institute of British
Engineering and Environmental Impact. Proceedings of the Geographers, New Series, 4, 116-117.
First Multidisciplinary Conference on Sinkholes, Florida, RIGBY-JONES, J., CLAYTON,C. R. I. & MATTHEWS,M. C. 1993.
Balkema Press, 159-165. Dissolution features in the chalk: from hazard to risk.
FISCHER, J. A. & LECHNER, H. 1989. A karst ordinance- Paper 7. In: INSTITUTIONOF CIVIL ENGINEERS (eds) Risk
Clinton Township, New Jersey. In: BECK, B. F. (ed.) Engin- and Reliability in Ground Engineering. Thomas Telford,
eering and Environmental Impacts of Sinkholes and Karst. London, 87-99.
Proceedings of the Third Multidisciplinary Conference on SPERLING, C. H. B., GOUDIE, A. S., STODDART,D. R. & POOLE,
Sinkholes, St. Petersburg Beach, Florida, Balkema Press, G. G. 1977. Dolines of the Dorset Chalklands and other
357-361. areas of Southern Britain. Transactions of the Institute of
HIGGINBOTTOM, I. E. 1979. Suggested distribution of chalk pipes British Geographers, New Series, 2, 205-223.
and swallow holes with a tentative indication of relative SPERLING, C. H. B., GOUDIE, A. S., STODDART,D. R. & POOLE,
subsidence risk levels. Unpublished presentation to the G. G. 1979. Discussion- Origin of the Dorset Dolines.
Geological Society of London. Transactions of the Institute of British Geographers, New
HUBBARD, D. A. 1984. Sinkhole distribution in the central and Series, 4, 121-124.
northern Valley and Ridge province, Virginia. In: BECK, STEWART, M. & WOOD, J. 1984. Geophysical characteristics of
B. F. (ed.) Sinkholes: their Geology, Engineering and Envir- fracture traces in the carbonate Floridan Aquifer. In:
onmental Impact. Proceedings of the First Multidisciplinary BECK, B. F. (ed.) Sinkholes: their Geology, Engineering and
Conference on Sinkholes, Florida, Balkema Press, 75-78. Environmental Impact. Proceedings of the First Multi-
JONES, D. K. C. 1981. Southeast and Southern England. Univer- disciplinary Conference on Sinkholes, Florida, Balkema
sity Paperbacks, Methuen. Press, 225-229.
KIRKWOOD, J. P. & EDMONDS, C. N. 1989. Ground subsidence THORP, M. J. W. & BROOK, G. A. 1984. Application of double
movements upon the Cretaceous chalk outcrop in Eng- Fourier series analysis to ground subsidence susceptibility
l a n d - Sinkhole problems and engineering solutions. In: mapping in covered karst terrain. In: BECK, B. F. (ed.)
BECK, B. F. (ed.) Engineering and Environmental Impacts of Sinkholes: their Geology, Engineering and Environmental
Sinkholes and Karst. Proceedings of the Third Multi- Impact. Proceedings of the First Multidisciplinary Con-
disciplinary Conference on Sinkholes, Florida, Balkema ference on Sinkholes, Florida, Balkema Press, 197-200.
Press, 247-255. TROESTER, J. W., WHITE, E. L. & WHITE, W. B. 1984. A com-
KOCHANOV, W. E. 1989. Karst mapping and applications to parison of sinkhole depth frequency distributions in temper-
regional land management practices in the Commonwealth ate and tropic karst regions. In: BECK, B. F. (ed.) Sinkholes:
of Pennsylvania. In: BECK, B. F. (ed.) Engineering and their Geology, Engineering and Environmental Impact. Pro-
Environmental Impacts of Sinkholes and Karst. Proceedings ceedings of the First Multidisciplinary Conference on Sink-
of the Third Multidisciplinary Conference on Sinkholes, holes, Florida, Balkema Press, 65-73.
Florida, Balkema Press, 363-368. WILLIAMS, P. W. 1972. The analysis of spatial characteristics
LITTLEFIELD, J. R., CULBRETH, M. A., UPCHURCH, S. B. & of karst terrains. In: CHORLEY,R. J. (ed.) Spatial Analysis
STEWART, M. T. 1984. Relationship of modern sinkhole in Geomorphology, Part H - P o i n t Systems. Methuen,
development to large scale photolinear features. In: BECK, London, 135-163.
Geological Society, London, Engineering Geology Special Publications
Engineering geological mapping
J. S. Griffiths

Geological Society, London, Engineering Geology Special Publications 2001; v. 18; p. 39-42
doi:10.1144/GSL.ENG.2001.018.01.06

© 2001 Geological Society of London

Engineering geological mapping

J. S. Grittiths
Department of Geological Sciences, University of Plymouth, Devon, U K

Introduction poses, medium- or, more rarely, small-scale maps can be

appropriate. In detailed engineering geological studies,
The development, scope and examples of best practice plans and sections at scales as large as 1:500 or even
in the preparation of geological maps and plans for 1:100 are likely to be required.
engineering practice are discussed in detail in Dearman
(1991), and this remains the definitive British text on the
subject. In the UK engineering geological maps were Components of an engineering
recognized as forming a significant component of site geological map
investigations with the publication of the Geological
Society Working Party Report on maps and plans (Anon. Fookes (1969) states that the aim of engineering geo-
1972). In many parts of the world the practice is also logical mapping should be to produce a map on which
long established (e.g. Peter 1966; Popov et al. 1950), and the mapped units are defined by engineering properties
in 1976 UNESCO produced a guide for the prepara- or behaviour, and the limits of the units are determined
tion of engineering geological maps (Anon. 1976). Many by changes in the physical and mechanical properties of
examples of international studies were reported at the the materials. Clearly on such maps the boundaries
1979 IAEG Newcastle Symposium on Engineering Geo- of the mapped units may not correlate or coincide with
logical Mapping. In contrast, in the United Kingdom, the underlying geological structure or the lithostrati-
BS5930, the Code of Practice for Site Investigations, graphic units depicted on normal geological maps.
which was first published in 1981 and has guided the The information that has to be recorded on the engi-
scope and content of site investigations in the UK for the neering geological map was laid down in the Geological
past two decades, makes scant reference to geological or Society Working Party Report (Anon. 1972) and, with
engineering geological mapping (Griffiths & Marsh certain additions, these still provide an excellent check-
1986). As noted in Griffiths & Edwards (2001), this situa- list. An updated version of this list is presented as Table 1.
tion has not been corrected in the revision of BS5930 In addition to the observed mapped data, information
published in 1999. Fookes (1997), in the first Glossop on the location of previous site investigations should be
lecture, states 'that engineering geological mapping, even noted including the sites of boreholes, trial pits and geo-
sketch mapping, is particularly under used in British physical surveys. Similarly, notes should be made on all
practice', and this is despite its longstanding and success- mines and quarries, including whether active or aban-
ful track record (Dearman & Fookes 1974). As with doned, dates of working, materials extracted and whether
many facets of land surface evaluation, this situation or not mine plans are available.
needs to be addressed. Whilst this list provides a broad coverage, the require-
ments for an individual engineering geological map can
be tailored to suit the specific issues to be investigated.
For example, in an area of volcanic risk there is likely to
The distinction between engineering be more emphasis on the nature, frequency of occur-
geological maps and plans rence and runout distances of lava flows, pyroclastic
flows, lahars, ash clouds, lateral blasts and toxic gases.
A critical aspect of all maps is the scale, and for Similar detailed forms of investigation will be appro-
engineering geological mapping the system contained in priate for different types of geohazard evaluation. It is
the UNESCO system is the most appropriate (Anon. also possible that more than one map will be required.
1976): large-scale maps are 1:10 000 or larger: medium- Dearman (1991) makes reference to the value of
scale are between 1:10000 and 1:100000; and small- engineering geological zoning. This identifies areas on
scale are 1:100 000 or smaller. Most engineering studies the map that have approximately homogenous engineer-
will require work to be carried out on large-scale maps ing geological conditions. Such zones would normally
or plans, although for regional studies and planning pur- be derived from the factual data compiled on the base

From: GRIFFITHS,J. S. (ed.) Land Surface Evaluationfor Engineering Practice. Geological Society, London, Engineering Geology
Special Publications, 18, 39-42. 0267-9914/01/$15.00 9 The Geological Society of London 2001.
40 J. S. GRIFFITHS

Table 1. Information to be recorded on an engineering geo- civil engineering purposes and will normally be based on
logical map site plans made specifically for the location. As defined in
Anon. (5972), the engineering geological plan is taken to
Geological data include both maps and other methods of displaying field
9 Mappable units (on the basis of descriptive engineering
geological terms) data such as cross-sections, gate diagrams and exposure
9 Geological boundaries (with accuracy indicated) logs. In these circumstances the mapping techniques used
9 Description of soils and rocks (using engineering codes of in smaller-scale mapping might have to be supplemented
practice, e.g. BS5930) by detailed surveying and logging procedures to ensure
9 Description of exposures (cross-referenced to field the data are recorded with the necessary level of accur-
notebooks) acy. Dearman (5991) divides plans into two types.
9 Description of state of weathering and alteration (notes
depth and degree of weathering) 1. The pre-construction or site investigation plan. This
9 Description of discontinuties (as much detail as possible on is prepared during the early stages of an investiga-
the nature, frequency, inclination and orientation of all tion to order to allow the ground investigations to
joints, bedding, cleavage, etc) be planned and engineering problems to be antici-
9 Subsurface conditions (provision of subsurface information if pated. These plans would normally be at 1:5000 for
possible, e.g rockhead isopachytes) general purposes although may increase to 5:5000
Hydrogeological data for specific studies, for example of dam sites or
9 Availability of information (reference to existing maps, well tunnel portals (see Griffiths 2001).
logs, abstraction data) 2. Construction or foundation plans. These are pro-
9 General hydrogeological conditions (notes on: groundwater duced during construction when exposures can be
flow lines; piezometric conditions; water quality; artesian logged as foundations are excavated. These provide
conditions; potability)
a record of the actual ground conditions and allow
9 Hydrogeological properties of rocks and soils (aquifers,
aquicludes and aquitards; permeabilities; perched water the geological ground model to be refined. The
tables) scales for plans may be as large as 1:500.
9 Springs and seepages (flows to be quantified wherever Given the potential for construction claims associated
possible)
with unforeseen ground conditions, accurate engineering
Geomorphological data (see section on geomorphological geological plans clearly can play a significant role in
mapping) supporting or rejecting such petitions.
9 General geomorphological features (ground morphology,
landforms, processes)
9 Ground movement features (landslides, subsidence,
solifluction lobes; cambering) Data collection
Geohazards Primary mapping for engineering geology follows the
9 Mass movement (extent and nature of landslides, type and same basic rules and uses the same techniques estab-
frequency of landsliding, possible estimates of runout hazard) lished for conventional geological mapping (see Barnes
9 Flooding (areas at risk, flood magnitude and frequency,
1997). All graduates in geology in the UK are taught
coastal or river flooding)
9 Coastal erosion (cliff form, rate of coastal retreat, coastal these skills, thus engineering geological mapping should
processes, types of coastal protection) not prove to be a major problem to them. The initial
9 Seismicity (seismic hazard assessment) decisions to be made when undertaking engineering geo-
9 Vulcanicity (volcanic hazard assessment) logical mapping are to identify the types of data that are
to be collected to meet the survey requirements, at what
scale will mapping be carried out, and what methods are
to be used for data collection. The required end-product
map and therefore should not form part of the origi- must guide these decisions, i.e. what is the map to be
nal map but can be produced as an overlay. Zoning used for?
maps can be particularly effective in geohazard studies In most engineering situations there will be three
where the magnitude of the hazard can often be repre- phases to the work: desk studies, including aerial photo-
sented by an interpretative map with an ordinal scaling graph interpretation, where all existing data are com-
for the degree of hazard. piled; primary mapping in the field; interpretation and
preparation of the final maps. In most instances there will
be a requirement for primary data collection through
Engineering geological plans field mapping, even if only small-scale maps are required
for planning purposes or the amount of background
When engineering geological data are represented at material is quite comprehensive.
scales larger than 1 : 50 000 then the material is shown on Once the factual data have been collected and
a plan. Engineering geological plans are made for specific compiled then derivative maps or plans can be prepared
 ENGINEERING GEOLOGICAL MAPPING 41

Table 2. Types of applied geological map relevant to engineering found to be of value in engineering and planning situa-
tions is presented in Smith & Ellison (1999) based on the
Data points results of the U K Applied Geological Mapping Pro-
9 Location of exploratory holes and wells gramme (Table 2).
9 Distribution of geotechnical data test results
9 Point rockhead information
Disturbed ground (human activity)
9 Distribution (general) Map legend
9 Distribution of mines and mine workings (all types, including
surface and subsurface) The map legend used on an engineering geological map
9 Distribution of made-ground should be developed to suit the purpose of the study.
Superficial geology The general legend provided in the First Working Party
9 Soil types, extent, lithology and thickness Report on Maps and Plans (Anon. 1972), and presently
9 Drift thickness/rockhead contours being updated, remains the most useful basis for devel-
9 Geotechnical properties of soils oping a task-specific legend. Dearman (1991) also con-
Bedrock geology tains a wide range of examples of map legends that have
9 Rock types, extent, lithology, lithostratigraphy been used both in the U K and overseas. As with all
9 Structure contours legends it is critical that the symbol used or zone iden-
9 Geological structure tified is fully explained. This might take the form of an
9 Geotecbnical properties of rocks expanded legend in the key or a short report, either
Engineering geology accompanying the map or printed on the bottom of the
9 Foundation conditions map. As a general recommendation the map should be
9 Hydrogeological conditions able to stand alone and be understood by all potential
9 Ground conditions in relation to groundwater users without having to refer to a separate report.
9 Nature and distribution of geohazards (subsidence,
instability, flooding, earthquakes, etc.)
9 Engineering geological zones (i.e. areas of homogenous
engineering geological conditions)
9 Aggregate and borrow material sources Conclusions
Geomorphology Engineering geological mapping should be part of all
9 Geomorphological landforms and process site investigations and, in conjunction with other tech-
9 Drainage
niques of land surface evaluation, would form the basis
9 Areas of slope instability
9 Flood frequency limits for the development of the geological ground model, as
proposed in Fookes (1997). The data from the mapping
Derived construction constraints maps should be combined with ground investigation data and
9 Slope steepness an updated ground model developed prior to construc-
9 Ground instability (e.g. subsidence, cambering, landslides,
soft ground, etc.) tion. This model would then be further refined during
9 Landslide hazard and risk maps construction as new soil and rock exposures become
9 Previous industrial usage (brownfield sites and contaminated available.
land)
Derived resources maps
9 Nature, extent and properties of mineral resources
(superficial and bedrock) References
9 Groundwater resources
9 Distribution of aggregates ANON. 1972. The preparation of maps and plans in terms of
9 Sites of Special Scientific Interest (SSSIs) engineering geology. Quarterly Journal of Engineering
Geology, 5, 293-381.
Summary maps ANON. 1976. Engineering Geological Maps: A Guide to their
9 Development potential Preparation. The UNESCO Press, Paris.
9 Summary of construction constraints BARNES, J. W. 1997. Basic Geological Mapping. 3rd Edition.
9 Statutory protected land John Wiley & Sons, Chichester.
DEARMAN, W. R. 1991. Engineering Geological Mapping.
Based on Smith & Ellison (1999). Butterworth Heineman, Oxford.
DEARMAN, W. R. & FOOKES, P. G. 1974. Engineering geolog-
ical mapping for civil engineering practice in the United
that meet the specific requirements. Derivative maps are Kingdom. Quarterly Journal of Engineering Geology, 7,
obtained by summarizing, simplifying or combining 223-256.
factual data and presenting them in a different form (see EDWARDS,R. J. G. 2001. Creation of functional ground models
Edwards 2001). An indication of the wide range of maps in an urban area. This volume.
42 J. S. GRIFFITHS

FOOKES, P. G. 1969. Geotechnical mapping of soils and sedi- site investigation. In: HAWKINS, A. B. (ed.) Site Investi-
mentary rock for engineering purposes with examples of gation Practice: Assessing BS5930. Geological Society,
practice from the Mangla Dam project. G~oteehnique, 19, London, Engineering Geology Special Publications, 2,
52-74. 261-267.
FOOKES, P. G. 1997. Geology for engineers: the geological PETER, A. 1966. Essai de carte grotechnique. Sols-Soils, Paris,
model, prediction, and performance. Quarterly Journal of 16, 13-28.
Engineering Geology, 30, 293-424. PoPov, I. V., KATS, R. S., KORIKOVSKAIA,A. K. & LAZAREVA,
GRIFFITHS, J. S. 2001. Development of a ground model for the V. P. 1950. Metodika sostavlenia inzhenernogeologicheski-
UK Channel Tunnel portal. This volume. khikart (The Techniques of Compiling Engineering Geo-
GRIFFITHS, J. S. & EDWARDS, R. J. G. 2001. The development logical Maps). Gosgeolizdat, Moscow.
of land surface evaluation for engineering practice. This SMITH, m. • ELLISON, R. A. 1999. Applied geological mapping
volume. for planning and development: a review of examples from
GRIFFITHS, J. S. & MARSH, A. 1986. BS5930: the role of geo- England and Wales 1983 to 1996. Quarterly Journal of
morphological and geological techniques in a preliminary Engineering Geology, 32, S1-$44
Geological Society, London, Engineering Geology Special Publications
Landslide hazard mapping and risk assessment
G. J. Hearn and J. S. Griffiths

Geological Society, London, Engineering Geology Special Publications 2001; v. 18; p. 43-52
doi:10.1144/GSL.ENG.2001.018.01.07

© 2001 Geological Society of London

Landslide hazard mapping and risk assessment

G. J. Hearn 1 & J. S. Grittiths 2

Scott Wilson Kirkpatrick & Co Ltd, Basingstoke, Hampshire, U K
2 Department of Geological Sciences, University of Plymouth, Devon, U K

Definitions remote mountainside may pose little risk, while a small

fall of boulders in an urban area can have serious risk
Before landslide hazard mapping and risk assessment consequences for property and public safety.
are reviewed, it is important to define terms and con- Most maps published to date aim to identify and por-
cepts closely in order to avoid the confusion and misuse tray regions, areas or individual slopes that are likely to
that has occurred in some previously reported case be more prone to failure than others. These susceptibility
histories. The most widely accepted and basic definitions maps can vary significantly in scale, from the regional to
in landslide studies are those provided by Varnes (1984). the slope-specific. Landslide inventories and susceptibil-
ity maps produced for national or regional landslide regi-
Natural Hazard: the probability of occurrence within a stration and planning purposes are usually published at
specified period of time and within a given area of a scales of 1:25 000 to 1 : 100 000, and occasionally smaller,
potentially damaging phenomenon. while maps produced for local planning or engineering
Vulnerability: the degree of loss to a given element or set schemes are larger scale, usually between 1:5000 and
of elements resulting from the occurrence of a natural 1 : 25 000. Maps may also range from a simple depiction of
phenomenon of a given magnitude. It is expressed on the known distribution of landslides, determined from
a scale from 0 (no damage) to 1 (total loss). aerial photographs and field mapping, to Geographical
Specific Risk: the expected degree of loss due to a Information Systems (GIS)-based landslide susceptibility
particular natural phenomenon. It may be expres- maps that analyse the conditioning and triggering factors
sed by the product of Hazard and Vulnerability. that promote slope failure. These factors are usually eval-
Elements at Risk: the population, properties, economic uated from a combination of remotely sensed data, pub-
activities, including public services, etc., at risk in a lished information and field mapping and investigation.
given area. True landslide hazard maps that incorporate probability
Total Risk: the expected number of lives lost, persons and the potential to cause damage, are less common.
injured, damage to property or disruption of eco- Further discussion and illustration of landslide suscep-
nomic activity due to a particular natural phenom- tibility and hazard mapping techniques is contained in
enon. It is therefore the product of Specific Risk and Hutchinson (1995) and Aleotti & Chowdhury (1999).
Elements at Risk. Landslide risk assessment attempts to combine the
These definitions have been expanded by the Inter- hazard and risk parameters into a framework allowing
national Union of Geological Sciences (lUGS) Work- planning and project management decisions to be made.
ing Party on Landslides through its committee on Risk Quantification of these parameters (quantitative risk
Assessment (IUGS 1997) but the main elements are assessment or QRA) allows different options to be
essentially the same. examined and compared, and the costs and potential
Hazard, therefore, defines the potential to cause dam- benefits of hazard mitigation schemes or hazard avoid-
age. With respect to landslides it is necessary to: identify ance strategies to be evaluated in terms of reduced risk,
the existence of a landslide or a potential slope failure; i.e. offset economic loss and improved public safety.
establish its size, depth, speed and travel distance; and A comprehensive review of this subject is contained in
estimate or calculate its frequency of movement or its Cruden & Fell (1997).
probability of occurrence. Risk defines the vulnerability
and value of the elements at risk; that is, establishing the
damage or loss potential posed by the hazard should it Methodology
occur. Risk is measured in terms of economic loss, hard-
Landslide hazard mapping
ship, loss of livelihood and threat to public safety. Thus,
a phenomenon that poses a high hazard, such as a large, Rather than describing a specific technique with a set
fast-moving landslide or debris flow, occurring on a procedure, landslide hazard mapping represents a range of

From: GRIFFITHS, J. S. (ed.) Land Surface Evaluation for Engineering Practice. Geological Society, London, Engineering Geology
Special Publications, 18, 43-52. 0267-9914/01/$15.00 9 The Geological Society of London 2001.
44 G. J. H E A R N & J. S. G R I F F I T H S

Fig. 1. Landslide susceptibility mapping for route alignment through an unstable river basin in east Nepal.
L A N D S L I D E H A Z A R D M A P P I N G A N D RISK ASSESSMENT 45

Fig. 1. (continued)
46 G . J . HEARN & J. S. GRIFFITHS

techniques that have been developed by different prac- river undercutting, waterlogged ground and adverse
titioners, each with different approaches, differing data- rock structure orientations are also often identifiable on
sets with which to work and different specific mapping aerial photographs.
objectives. The unifying factor in virtually all attempts The manner in which the mapping procedure is set
to portray landslide susceptibility or landslide hazard up has a significant effect on the landslide controlling
mapping is that one, or more, of the following assump- factors that are selected, analysed and represented. Land-
tions is made. slide susceptibility and hazard mapping is a terrain-based
techniques of assessment and, as such, a study area
1. The location of future slope failure or ground move-
should be subdivided according to slopes and terrain
ment will be determined by the distribution of exist-
units for hazard analysis. The logical steps in any land
ing or past landslides, i.e. known landslide locations
surface evaluation involve the recognition of patterns in
will continue to be a source of hazard.
the geomorphological and geological landscapes, and
2. Future landslides or ground movement will occur
in this respect landslide hazard mapping is no different.
under similar ground conditions to those pertaining
However, landslide hazard mapping based purely on
at the sites of existing or past landslides, i.e. the
statistical correlation between landslide occurrence and
conditioning or controlling factors that gave rise
slope angle and rock type, for instance, has frequently
to existing landslides can be ascertained and their
been done on a grid square basis for ease of computer
distribution reliably mapped over wider areas. When
analysis. One of the earliest examples was that of
examined collectively, these factors can provide a
Carrara et al. (1978) using land unit parcels of 200m x
reasonable indication of the relative tendency for
200 m. Notwithstanding the obvious value of this early
slopes to fail.
work, the extent to which landslide controlling factors
3. The distribution of existing and future landslides
remain consistent over a 200 m square interval remains
can be approximated by reference to conditioning
debatable.
factors alone, such as rock type or slope angle.
Figure 1 shows a simple landslide susceptibility map-
Mapping techniques that have relied on the first assump- ping exercise undertaken in 1984 for road alignment
tion are loosely classified as direct mapping methods purposes in Nepal (Hearn 1987; TRL 1997). Slope angle
in that they are based on a known distribution of exist- was measured on a grid square basis from aerial photo-
ing landslides or slope failures. Methods that rely on graphs using a parallax bar. The grid cells were 230 m x
inference (the third assumption) are classified as indirect 230m and proved too coarse to allow slope angle to
mapping. These methods may rely on the knowledge or be significantly correlated against landslide distribution.
expectation that a given rock type is more prone to A map that depicted the distribution of physiographical
failure than another, while steeper slopes or the pres- zones, identified and classified from aerial photographs
ence of an undercutting river will create a higher failure according to slope morphology, proved far more appro-
potential, other factors being equal. An approach based priate (Fig. 1). One of the more recent examples is that of
on the combination of the two methods (the second Greenbaum (1995) who used Landsat TM pixels (circa
assumption) is most desirable, because a known landslide 30 m • 30 m) as the basic mapping unit. When landslide
distribution will help to clarify and confirm the relative hazard mapping is contemplated at the regional and
importance of controlling factors in landslide initiation. national scale, as in the cases described by Greenbaum,
Furthermore, representation of the geographical varia- then the use of a grid square basis for data gathering and
bility in these controlling factors allows an interpretation data analysis (black box method) becomes far less time
of landslide susceptibility to be extrapolated into areas consuming than a landform-based approach (white box
where, for whatever reason, failure has not yet occurred, method). It is preferable, with grid-based data, to employ
or into areas where landslide event data are unavailable. GIS methods of landslide hazard data management,
Landslide susceptibility mapping, therefore, is nor- analysis and mapping output (Greenbaum 1995). Each
mally based on geographical correlation between land- data layer, such as topography or lithology, can be repre-
slide distribution and landslide controlling factors over sented by an individual map. Algorithms are then used
large areas using existing data sources. Published geo- to undertake various spatial analyses on the data sets.
logical and topographical maps are usually available, Further discussion on this subject can be found in
and together provide information on lithology, geologi- Soeters & van Westen (1996).
cal structure, slope steepness and aspect. Some may
even show mapped landslide deposits. Aerial photo-
graphs and, to a lesser extent, satellite imagery can be Hazard analysis and quantitative risk
useful in identifying landslides and slopes prone to assessment (QRA)
failure. Furthermore, aerial photograph interpretation
can assist with the identification of geographical pat- Identifying which slopes are most prone to movement or
terns of landsliding, and the recognition and mapping failure is the first step in the assessment of landslide
of controlling factors. For example, colluvial deposits, susceptibility hazard and risk. The next steps relate
 LANDSLIDE HAZARD MAPPING AND RISK ASSESSMENT 47

to the actual risk from an identified hazard and the may be a 100% probablility of a debris flow destroying a
need to address the questions: How, When and With boundary fence, but only a 10% probability of the same
what effect? In carrying out a quantified (landslide) risk landslide reaching a building, and only a 1% probability
assessment in this way, the following parameters need to of fatality if the building collapses or a pedestrian is
be ascertained: caught in the debris. Event probability is derived either
from historical data (most rarely) or through computer
9 the zone of influence of the hazard, i.e. its area of simulation (most frequently). As with any probabilistic
movement and its runout distance; analytical procedure based on limited factual data, the
9 the frequency of movement, e.g. the frequency with retention of plausible argument in the final conclusion is
which failed masses travel over a given distance or paramount.
range of distances, such as boulders impacting on Undertaking actions 6 and 7 requires moving outside
buildings and public amenities located at varying the normal area of expertise of the engineering geologist/
distances from a rockfall source; geomorphologist. Input from social scientists and
9 the risk consequences attached to different event out- economists is needed in order for the social and economic
comes occurring, measured in economic terms, e.g. by costs associated with different levels of risk (action 6) to
reference to road maintenance costs, the costs of tem- be calculated. It is also necessary for the acceptable level
porary road closures, the repair or reconstruction costs of risk to be defined (action 7), and this is likely to be a
to damaged buildings and other infrastructure, and the function of government legislation, legal precedents and
cost of personal injury, fatality or loss of livelihood. the general societal appreciation of natural and man-
made risks. However, it is critical that engineering geolo-
The following actions and input data are required before
gists and geomorphologists have an input in defining
a landslide risk assessment, culminating in a QRA, can
acceptable risks otherwise these may be set at impossibly
be fully achieved:
high levels (Fell & Hartford 1997).
1. preparation of a landslide location and susceptibility
map to identify potential landslide sources;
2. estimation of the volumes of potential landslide Limitations
masses likely to be derived from these sources;
3. estimation of the likely areal influence and runout The quality of the landslide hazard mapping output is
distance of these landslide masses; dependent on that of the data used and the assumptions
4. assessment of frequency or return periods of differ- made in its derivation. At the site or project area scale
ent landslide and runout scenarios; (1 : 1000 to 1 : 25 000), resources and practicalities will
5. list of potential consequences and vulnerability of usually permit detailed mapping and geotechnical model-
the elements at risk to these landslide and runout ling to be carried out. As the size of the area to be covered
scenarios; by the mapping increases to the regional scale (1 : 25 000
6. calculation of the economic loss and evaluation of to 1 : 50 000) the level of field-derived data input reduces,
the public safety implications associated with out- and the analysis becomes broader and more generalized,
comes likely to take place over the lifetime of an exist- and may be limited to a simple overlay of geology with
ing or proposed development, for example 25 years in topography. Most landslide hazard mapping falls into
the case of a low-cost road, or 100 years in the case of the regional scale category, and is essentially a remote
a housing development; exercise, conditioned by the data available from existing
7. if the risk levels calculated in action 6 are unaccept- sources. These data may often be incomplete, and may
able, calculation of the cost of mitigation works and lack information on a slope parameter that is considered
decision as to the most appropriate strategy for risk central to the initiation of landslides in that area. With-
management. In the extreme, this may mean can- out this information, landslide hazard mapping will be
cellation of the project. flawed from the beginning. Furthermore, certain con-
trolling factors are often judged to be more important
In terms of mapping, landslide hazard assessment might than others in causing slope failure, and weighting
succeed in delineating areas and frequencies of land- factors are assigned accordingly. The weighting system
slide runout, such as debris flows from a mountain slope illustrated in Table 1 was applied to rock slope hazard
onto a terrace below, or it may be able to portray the mapping in Papua New Guinea on the basis of field
frequency of damage to roads as recorded in road main- observation (TRL 1997). The system was able to model
tenance records. Thus the first four actions identified reasonably well the distribution of recorded failures and
above can be achieved if sufficient mapping data and it was therefore used for predictive purposes. The weight-
records of past events exist (e.g. Lee 2000). ing system illustrated in Figure 1 for Nepal was also
Event and decision tree analyses (Wu et al. 1996; Lee based on observed factor map correlations with landslide
et al. 2000) can be used to define and assign probabilities locations. The use of weighting factors without this
to the various consequences (action 5). For example there observational support is not recommended.
48 G. J. H E A R N & J. S. G R I F F I T H S

Table 1. Rock slope hazard rating for a site in Papua New Guinea

Hazard index

Factor 1: Slope angle irrespective of lithological dip

Limestone slopes (based on graphical database)
Angles <30 ~
Angles 31-44 ~
Angles 45-60 ~
Angles >60 ~
Weighting factor, 3

Mudstone slopes (based on graphical database)

Angles <25 ~
Angles 26-30 ~
Angles 31-35 ~
Angles >35 ~
Weighting factor 3

Factor 2: Slope angle with respect to angle of unfavourable lithological dip

Limestone and mudstone slopes
Angle is at least 10~ less than dip angle
Angle is at least 5 ~ less than dip angle
Angle isl0 ~ greater than dip angle
Angle is > 10~ greater than dip angle
Weighting factor: 3 where dip direction is directly out of slope; 2 where dip direction
is obliquely out of slope

Factor 3: Adverse bedrock sequences (permeable overlying less permeable lithologies and]or strong
overlying weak lithologies)
None
Repetitive sequence of thin beds
Single sequence with thin reservoir or strong rock stratum
Single sequence with thick reservoir or strong rock stratum
Weighting factor, 2

Factor 4: Faulted or sheared rock masses (only one of Factors 4 and 5 may be included in rating)
None
Yes, but with favourable structural orientation
Yes, with unfavourable structural orientation
Weighting factor, 2

Factor 5: Rock-mass dilation (only one of Factors 4 and 5 may be included in rating)
Massive with closed joints
Massive with open joints or moderately fractured with closed joints
Highly fractured with open joints
Crushed and dilated (as in H = 1 or H -- 2 for Factor 4)
Weighting factor, 2

Factor 6: Drainage
Well defined channelled runoff and dry slopes
No evidence of surface runoff, sinks or seepages
Sinkholes, internal drainage or ponding
Active springs and seepages
Weighting factor, 1 for limestone slopes and 2 for mudstone slopes

Factor 7: Slope unloading by toe erosion or failure from below

None
Localized and infrequent or due to ancient slope failure
Periodic or due to recent slope failure
Active
Weighting factor, 2

* Scale: 0 = least hazard; 3 = greatest hazard.

 LANDSLIDE HAZARD MAPPING AND RISK ASSESSMENT 49

A hazard map sets out to portray the nature and scale thorough analysis and reach an objective conclusion
of a hazard. However, the vast majority of hazard maps regarding the frequency of a landslide of a given size and
only portray the relative tendency or suseeptibifity of a given runout distance, and the value of the impact
slopes to fail. To indicate the level of hazard that might consequence. It is important to ensure that engineering
be generated they would have to incorporate frequency geological judgement prevails when numbers are being
or probability and some indication of areal influence sought to satisfy the needs of the computer analysis.
such as potential runout distances. These maps can, at Putting values on potential for loss of land, loss of earn-
best, portray relative hazard only; they do not provide an ings, disruption to communications, damage to prop-
indication of factor of safety, nor do they usually give an erty, and personal injury and fatality is a difficult task.
indication of how large, or how often? Therefore, from a A community's perception and tolerance, or even accept-
planner's or engineer's point of view most hazard maps ance, of disruption and risk to personal safety is hugely
will fall short of what is ultimately required for decision variable. This can lead to a disaster scenario being
making, unless they are backed up by some form of risk regarded as unacceptable in one country, while it is
assessment. If a hazard map is to progress from a docu- accepted as a fact of life in another.
ment that displays relative hazard to one that provides Hong Kong is a useful example of what can be
an indication of absolute hazard, and from then to a achieved. In Hong Kong, aerial photographs have been
document that portrays true risk, the quality and range taken of the territory since 1942, and there are fairly com-
of the necessary input data need to increase considerably. prehensive records of slope failures where these have
Landslide hazard maps also rarely take into account taken place in areas of urban development and related
the effects of triggering factors. Most landslides in infrastructure. QRA has therefore been possible in these
mountainous regions are small shallow failures involving circumstances (Malone 1998; Hardingham et al. 1998;
colluvium, soil or weathered rock. These landslides are Reeves et al. 1998). However, in areas where landslide
usually triggered by intense rainstorms which, unless events are less well documented and where less is known
generated by demonstrably marked orographic effects, about landslide controlling factors and triggering mech-
are just as likely to occur in one location as another over anisms, the prospects for landslide Q R A are far less.
a long enough period of time. However, this is little help
to a practitioner who has derived a landslide hazard map,
only to find that the following year an intense localized
rainstorm triggers numerous failures in an area where, Appfications
either through lack of data or lack of precedent, a low or
moderate hazard had been assigned. Earthquakes are Landslide hazard maps can be extremely valuable to
other obvious triggers that tend to initiate or reactivate regional planning and preliminary site selection, route
deeper failures. Whilst the location of active faults may alignment corridor identification, and in the provision
be known, the process of converting seismicity into geo- of a framework for slope management and slope
graphical landslide hazard is an uncertain one. maintenance (TRL 1997). A register of large landslides -
In hazard studies the timing of unknown failures will preferably GIS-based with an integrated attribute
also usually be related to a triggering mechanism and database describing their geology and geomorphology,
is virtually impossible to evaluate except in terms of probable cause, failure mechanisms, and extent and
return periods, and even then usually with a considerable periodicity of movement - represents an extremely valu-
degree of uncertainty. Known failures and susceptible able source of information for planning purposes. The
slopes, however, can be monitored to ensure there is pre- register of landslides recently compiled from aerial photo-
warning of accelerated movement, although the warning graphs in Hong Kong is a good example (Evans 1998;
periods are often very short. For longer-term studies an Griffiths et al. 1999). Taking the database a stage further
indication of frequency might be gained from historical into formal landslide susceptibility and hazard map-
records, aerial photographs or geobotanical evidence, ping also offers significant advantages if the factors con-
but this is very much dependent on the availability and trolling or triggering landslide location and behaviour
relevance of these long-term records in the first place. can be adequately measured and subdivided geographi-
Finally, the same slope may pose a hazard from rock cally into meaningful landform elements. However, any
falls and debris slides or debris flows, but each may be inaccuracies in this process could blight areas of low
conditioned by a different combination of controlling hazard, and careful field verification is required in mar-
factors, and initiated by different trigger mechanisms and ginal zones.
trigger levels. The hazard maps produced by Keinholz In many applications it is not necessarily a problem
(1978) are amongst the few that have attempted to com- that landslide hazard mapping and risk assessment fail,
bine type of hazard with qualitative assessment of risk conceptually, to evaluate absolute hazard, multiple haz-
on the same map. ards and true risk. A combination of landslide suscept-
The main limitation with landslide risk assessment lies ibility mapping and landslide runout analysis will often
in the usual paucity of data with which to undertake a suffice when route corridors for roads and pipelines are
50 G. J. HEARN & J. S. GRIFFITHS

being examined. This assumes that residual stability In a recent study in Hong Kong (Hadley et al. 1998),
problems can be overcome or managed, the structure has geomorphological mapping was used to identify poten-
to be built, and the objective is to find the most stable tial failure masses located on a granite mountainside
corridor in which to build it. High value elements of the above the proposed location for an expressway. A total
construction, such as bridges and construction camps, of 43 potential landslide sources were identified and
need to be located at sites where landslide, flood and the volumes calculated from the mapping data. The
scour hazards are at a minimum, but these sites are runout geometry of a large debris flow that had occurred
usually selected on the basis of a detailed examination of in 1990 was used to test a number of empirical run-
candidate sites identified from engineering criteria. In the out formulae derived from other mountain regions and
usual case where there is uncertainty over the magnitude, identify the one which was able to match, most closely,
impact potential, timing or frequency of landslides or the 1990 Hong Kong case. All of the observed poten-
slope movements, it will always be preferable to use tial failure masses were routed down the main drain-
whatever mapping and analytical methods can be applied age lines below. The computed debris flow runouts
to better define landslide hazards on engineering geo- reached the expressway embankments in only three of
logical grounds. If either the historical record, aerial the 12 valleys, and in all cases there was sufficient
photographs or field mapping indicate that large land- embankment height to retain the anticipated volumes
slide events have taken place in the past under simi- of debris. As a precautionary measure, checkdams were
lar geomorphological and climatic conditions to those scheduled to be constructed in the steepest drainage
that pertain at present, then it has to be assumed that lines upstream of the expressway alignment. This study
they will recur in the future. Their mitigation may be serves to illustrate the case where the frequency or return
cost-ineffective in the case of low to moderate value period of a landslide event cannot be reliably assessed,
investments but they will effectively become the 'design and the risk assessment is therefore required to identify
event' for such high value structures as expressways and a design event that represents the realistic worst-case
housing developments. outcome.

Fig. 2. Slope treatment and road reconstruction across the Jogimara Landslide, Prithvi Highway, Nepal.
 LANDSLIDE HAZARD MAPPING AND RISK ASSESSMENT 51

Slab failures onto large failed mass will continue to ovedoad Topography Geology
that slope causing it t-o fail gradually. 40000m 3
( ~ Creep failure of upper 3-5m of fracture~ phyllite , ~ Failure scar Undifferentiated
(note overturned strata). 3000 - 6000m outcrop
Convex change in
Q resh 0.5m vertical displacement at base of small cliff
in strong phyllite. 5000~8000m3 slope (slope ~ Limestone~
steepens) ....o...
~ mminent failure of spur. 10000 - 20000m 3 77" 9.:.....: Phyllite
Cliff
Imminent failure of spur. 10000m 3 Shaly phyllite
Slopes in this area X~ Slip scar or
contain limestone /5o tension crack $=k Schist
breccia boulders !
and talus Slope angle s,r Strata dip angle
! T'~ (degrees) and direction
44o Persistentjoint
angle and
& ~ Limestone direction
breccia boulders
3m cliff in 38~, and phyllite rafts .~/ Overturned strata
limestone 1 due to rock creep
breccia Sheared strata

Valley left
"hanging"
by failure

in shale

Quarry
floor

Faulted

6~
Expect cut slope
\ failure along
Culvert joint 47/040.
Dilation of steep
rock nose. Likely to
fail
50 0 50 100 150 Metres
I I I I I I I I I
Approx scale 1: 30000

Fig. 3. Engineering geological map of the Jogimara landslide showing anticipated failure volumes.

A study undertaken in Nepal illustrates the case where and delineated. The volumes of these rock masses were
close to the worst-case outcome had already taken place. estimated (Fig. 3) and time periods assigned during
The main highway into Kathmandu was repeatedly which their failure was considered most likely. These
blocked by rock falls from a particularly unstable rock time periods ranged between 12 months and five to ten
slope during the monsoon seasons of the late 1980s years. A number of stabilization and risk reduction
and early 1990s (Fig. 2). Access would be disrupted options were considered, ranging from minor, essentially
for several days at a time and queues of traffic would cosmetic, slope treatments and trap walls, through rock
build up for several kilometres on either side. Fatalities anchoring, bolting and shotcreting schemes to rock shel-
have been common. A combination of aerial photo- ter and covered tunnel options for road and traffic protec-
graph interpretation and engineering geological mapping tion. These options were costed and, given the relatively
assisted in identifying the main causes and mechanisms short time period in which the majority of remaining
of failure. From the mapping, several zones of fractured, unstable rock masses were considered likely to fall, the
adversely jointed and overhanging rock were identified recommendation was made to the client to adopt the
52 G. J. HEARN & J. S. GRIFFITHS

option of minor slope treatment with a rock trap wall. Landslides in Hong Kong. Atkins China Ltd Report GEO
Seven years later, the stability of the slope appears to 19/97 to Geotechnical Engineering Office, Hong Kong.
have improved, and initial indications would therefore HADLEY, D., HEARN, G. J. & TAYLOR,G. R. 1997. Debris flow
suggest that engineering geological mapping and judge- assessments for the Foothills Bypass, Hong Kong. In: LI,
ment proved able to assist in a slope management K. S., KRAY,J. N. & HO, K. K. S. (eds) Slope Engineering
in Hong Kong. Balkema, Rotterdam, 153-162.
decision where, through lack of data, quantitative risk HARDINGHAM,A. D., DITCHFIELD,C. S., HO, K. K. S. & SMALL-
assessment would have been difficult to apply. WOOD, A. R. H. 1998. Quantitative risk assessment-
a case history from Hong Kong. In: LL K. S., KRAV,J. N.
& HO, K. K. S. (eds) Slope Engineering in Hong Kong.
Conclusions Balkema, Rotterdam, 145-151.
HEARN, G. J. 1987. An evaluation of geomorphological contri-
butions to mountain highway design with particular reference
Landslide hazard mapping now has an established 'track
to the Lower Himalaya. PhD Thesis, University of London.
record' and, whilst most studies only evaluate relative HUTCHINSON, J. N. 1995. Landslide hazard assessment. Pro-
rather than absolute hazard, their value is not questioned ceedings of the Sixth International Symposium on Land-
for many applications, particularly at the reconnaissance slides, Christchurch, New Zealand, 3, 1805-1842.
or planning level. Genuine risk assessment, however, is lUGS 1997. Quantitative risk assessment for slopes and land-
still in its infancy as far as application to landslide studies slides- the state of the art. In: CRUDEN, D. & FELL, R.
is concerned. The concept has been developed in far more (eds) Landslide Risk Assessment. Proceedings of the Inter-
detail in connection with the manufacturing and nuclear national Workshop on Landslide Risk Assessment, Hon-
industries (Gerrard 2000) and was given strong impetus olulu, Hawaii. Balkema, Rotterdam, 3-12.
KIENHOLZ, H. 1978. Maps of geomorphology and natural
by the Royal Society Commission into risk analysis (The
hazard of Grindelwald, Switzerland, scale 1:10000. Arctic
Royal Society 1992). With respect to landslides and and Alpine Research, 10, 169-184.
indeed most natural hazards, with the exception of a few LEE, E. M. 2000. The use of archive records in landslide risk
notable examples such as Hong Kong, the data require- assessment: historical landslide events on the Scarborough
ments are presently far too onerous for full quantitative coast, UK. In: BROMHEAD,E., DIXON, N. & IBSEN, M.-L.
risk assessments to be normally undertaken. Clearly the (eds) Landslides in Research, Theory and Practice. Proceed-
need is for further development of data collection tech- ings of the Eighth International Symposium on Landslides,
niques to allow risk assessment of all natural hazards to Thomas Telford, London, 905-910.
be part of any terrain evaluation study. LEE, E. M., BRUNSDEN,D. & SELLWOOD,M. 2000. Quantitative
risk assessment of coastal landslide problems, Lyme Regis,
UK. In: BROMHEAt), E., DIXON, N. & IBSEN, M.-L. (eds)
Landslides in Research, Theory and Practice. Proceedings
References of the Eighth International Symposium on Landslides.
Thomas Telford, London, 899-904.
ALEOTTI, P. & CHOWDHURY,R. 1999. Landslide hazard assess- MALONE, A. W. 1998. Risk management and slope safety in
ment: summary review and new perspectives. Bulletin of Hong Kong. In: LI, K. S., KRAY, J. N. & HO, K. K. S.
Engineering Geology and the Environment, 58, 21-44. (eds) Slope Engineering in Hong Kong. Balkema, Rotter-
CARRARA,A., CATALANO,E., SORRISO-VALVO,M., REALI,C. & dam, 3 17.
Ossl, I. 1978. Digital terrain analysis for land evaluation. REEVES, A., CHAN, H. C. & LAM, K. C. 1998. Preliminary
Geologia Applicata e ldrogeologia, 13, 69-127. quantitative risk assessment of boulder falls in Hong
CRUDEN, D. & FELL, R. (eds) 1997. Landslide Risk Assessment. Kong. In: LI, K. S., KRAY,J. N. & HO, K. K. S. (eds) Slope
Proceedings of the International Workshop on Landslide Engineering in Hong Kong. Balkema, Rotterdam, 185-191.
Risk Assessment, Honolulu, Hawaii. Balkema, Rotterdam. SOEXERS, R. & VAN WESTEN, C. J. 1996. Slope instability
EVANS, N. C. 1998. The natural terrain landslide study. In: LI, recognition, analysis and zonation. In: TURNER, A. K. &
K. S., KRAY,J. N. & Ho, K. K. S. (eds) Slope Engineering SCHUSTER, R. L. (eds) Landslides: Investigation and Mitiga-
in Hong Kong. Balkema, Rotterdam, 137-144. tion. Transportation Research Board Special Report 247.
FELL, R. & HARTFORD, D. 1997. Landslide risk management. National Academy Press, Washington DC, 129-177.
In: CaUDEN, D. & FELL, R. (eds) Landslide Risk Assess- THE ROYALSOCIETY1992. Risk: Analysis, Perception and Man-
ment. Proceedings of the International Workshop on agement. Report of a Royal Society Study Group, The
Landslide Risk Assessment, Honolulu, Hawaii. Balkema, Royal Society, London.
Rotterdam, 51-109. TRL 1997. Principles of Low Cost Road Engineering in Moun-
GERRARD, S. 2000. Environmental risk management. In: tainous Regions. TRL Overseas Road Note 16. Transport
O'RIORDAN, T. (ed.) Environmental Science for Environ- Research Laboratory, Crowthorne.
mental Management. Prentice Hall, Harlow, 435-468. VARNES, D. J. 1984. Landslide Hazard Zonation: A Review of
GREENBAUM,D. 1995. Project summary report: Rapid methods Principles and Practice. UNESCO, Paris.
of landslide hazard mapping. Technical Report WC[95/30. Wu, T. H., WILSON,H. T. & EINSTEIN,H. 1996. Landslide haz-
British Geological Survey/DFID. ard and risk assessment. In: TURNER, A. K. & SCHUSTER,
GRIFFITHS, J. S., HUNGR, O., HUTCHINSON, J. N., HARDING- R. L. (eds) Landslides." Investigation and Mitigation. Trans-
HAM, A. D. & DITCHF~ELD, C. 1999. Scoping Study for a portation Research Board Special Report 247. National
Global Quantitative Risk Assessment of Natural Terrain Academy Press, Washington DC, 106-120.
Geological Society, London, Engineering Geology Special Publications
Geomorphological mapping
E. M. Lee

Geological Society, London, Engineering Geology Special Publications 2001; v. 18; p. 53-56
doi:10.1144/GSL.ENG.2001.018.01.08

© 2001 Geological Society of London

Geomorphological mapping

E. M. Lee
Department of Marine Sciences and Coastal Management, University of Newcastle,
Newcastle-upon-Tyne, UK

Introduction Egypt (Bush et al. 1980; Jones, 2001)); the delineation

of gypsum-related subsidence problems in the Ripon
There has been a long tradition of geomorphological area of the UK (Thompson et al. 1996); the assess-
mapping to support land use planning, especially in ment of landslide and erosion hazard at Ok Tedi
Poland (e.g. Klimaszewski 1956, 1961; Galon 1962) and copper mine, Papua New Guinea (Hearn 1995).
France (e.g. Tricart 1965; for a history of geomorpho- 9 Specific-purpose surveys to delineate and characterize
logical mapping in Europe, see Verstappen 1983). The particular landforms (e.g. the investigation of pre-
Geological Society Working Party Report on maps and existing landslide problems in and around the Channel
plans (Anon. 1972) identified examples of geomorpho- Tunnel terminal area, Folkestone (Griffiths et al. 1995;
logical mapping that could be of use to engineers. How- Griffiths 2001); the assessment of coastal erosion haz-
ever, the value of the technique was best highlighted by its ards at Blackgang, Isle of Wight (Moore et al. 1998)).
application to road projects in unstable terrain in Nepal
and South Wales during the early 1970s (e.g. Brunsden There is no single approach to geomorphological map-
et al. 1975a, b; Doornkamp et al. 1979; Jones et al. 1983).
ping. The method chosen will generally reflect the nature
The techniques have also been successfully applied to of the problem to be solved, the resources available and,
dryland problems (e.g. Brunsden et al. 1979; Bush et al. not least, the training and experience of the mappers.
1980, Doornkamp et al. 1980; Cooke et al. 1982, 1985; However, all maps should seek to subdivide the land-
Jones et al. 1986), soil erosion (e.g. Morgan 1995) and scape into units with similar surface form, materials and
river management (e.g. Doornkamp 1982; Richards et al. process characteristics. At the smaller scales these units
1987). Despite these applications, BS5930, C o d e o f Prac- will, inevitably, be terrain models or land systems (see
tice f o r Site Investigations (British Standards Institution
Phipps 2001; Fookes et al. 2001). Individual landforms
1981), contains little reference to geomorphological maps or terrain units (e.g. escarpments, dune ridges, river
(Griffiths & Marsh 1986) and the technique remains a channels, landslides) might be recorded on medium-scale
marginal skill practised by a few experienced engineering maps. Landform elements or geomorphological units
geomorphologists. Examples of good mapping practice (e.g. individual landslide blocks, within-channel bars,
can be found within the references cited in this paper. gullys) might be recorded on large-scale maps.
The method of recording geomorphological informa-
tion may change with map scale, but the basic approach
Methods to data collection should remain the same: mapping of
surface form, description of materials and recording
The style and format of a geomorphological map needs evidence of process. However, a key distinction must be
to reflect the nature of the environment and the prob- made between desk-based approaches (i.e. aerial photo-
lems that need to be addressed. The following are some graph interpretation and desk study) and field studies
of the more common types of map. (i.e. morphological mapping, recording near-surface
9 Regional surveys of terrain conditions, either to materials and evidence of surface processes).
provide a framework for land use planning (e.g. the
1:25 000 scale geomorphology map of the Torbay area Mapping surface form
(Doornkamp 1988)and the 1:50000 scale resources
survey maps of Bahrain (Doornkamp et al. 1980)) or as At the smaller scales, topographic map contours can
part of the baseline studies for environmental impact be used to subdivide the landscape into units, based on
assessment (e.g. Lee 1999). slope steepness (i.e. from the spacing of contour lines)
9 General assessments of resources (e.g. the Bahrain and slope form (i.e. from the shape of the contours).
Surface Materials Resources Survey (Doornkamp et al. However, at larger scales contours reveal little about
1980)) or geohazards (e.g. the investigation of ground landforms and their assemblages. The technique of mor-
problems and flood hazard in the Suez City area, phological m a p p i n g is the most convenient and efficient

From: GRIFFITHS,J. S. (ed.) Land Surface Evaluation for Engineering Practice. Geological Society, London, Engineering Geology
Special Publications, 18, 53-56. 0267-9914/01/$15.00 9 The Geological Society of London 2001.
54 E.M. LEE

I~ 10 5

,.! ,; V4--q '-;

Morphological mapping symbols

9 . Angularconvex v ~, Angularconcave v--v Srnoothh/c~wex

breakofMope breakofsiope changeof~ope

~_.~ Smoothlyconcave 11 . Dirtmionofslope i ,' - , . m Cliffs(bedrock

changeofslol~ (sngle in degreeS} 40"ormom)
....... * BmaksofMope -r-r-rr *Changemofslope ~',, -- Convex slope untt

* Convex and concsve too close together t o allow I , Concs~slopeunit

use of separate symbols

Fig. 1. Morphological mapping system (after Savigear 1965, from Cooke & Doornkamp 1990).

way of recording the surface morphology and allows Hand-held Global Positioning System equipment may
later interpretation of form and process (Waters 1958; be used, although it is of restricted value in densely vege-
Savigear 1965). Breaks or changes of slope are identified tated terrain. In some instances it may be appropriate to
from aerial photographs or in the field, and recorded record slope morphology 'by eye' or to use conventional
using standard symbols (Fig. 1). The degree of generali- survey equipment to accurately record the detail of
zation used in a morphological mapping survey will surface features (e.g. using a series of surveyed markers).
depend upon the scale of the base map used and the
purpose of the exercise.
Morphological mapping in the field often involves two
Recording near-surface materials
people and the use of a tape measure (25-50 m length) and
compass. One person walks along a fixed traverse The distribution of near-surface materials should be re-
(oriented up and down the slope) with the tape until a corded at all exposures (e.g. landslide backscars, stream
break or change in slope is reached. The second person or wadi channels, cliffs, quarries or borrow pits), using
then records on the field map (i) the position (i.e. distance standard methods of description (e.g. BS5930; Geotech-
and compass bearing from the last point), (ii) the change nical Control Office 1988; Geotechnical Engineering
in slope form (e.g. a convex change of slope) and (iii) the Office 1996). The descriptions can be recorded in a note-
slope angle using a clinometer or Abney level, and 'fills-in' book or on a pro forma developed or adapted for the
the map by joining this break or change in slope with the mapping exercise. The second team member (i.e. the non-
same feature recorded on previous traverses, adding to mapper) should be responsible for making the field notes
the map any features observed between the traverses. A or filling in the pro formas. It is essential to record the
useful rule is that all morphological lines must join up. location of the observations. Photographs should be
The morphological map is built up through repeated taken of the exposure and the photo number and subject
traverses up and then down the slope (or vice versa). The recorded in the notebook or on the pro forma. Particular
distance between traverses will be dependent on the com- attention should be given to describing the nature of
plexity of the landscape and the density of the vegetation: weathered materials and the texture of transported soils
in dense undergrowth traverses need to be closer together as these may give an indication of their origin and the
than out in the open. A scale-rule, pencil, rubber (eraser) nature and scale of the processes that occur within
and sharpener are essential parts of the field kit, as on- the landscape. For example, debris flow deposits consist
going modifications in the mapping are commonplace. of poorly sorted, large clasts embedded in a matrix of
It is pointless to continue mapping if you are unsure of fine material. Boulders may be concentrated at the top
your location: it will either lead to a poor map or large of the deposit (i.e. reverse grading) because of the
amounts of rubbing out. buoyant forces and dispersive pressures within a debris
 GEOMORPHOLOGICAL MAPPING 55

flow. By contrast, flash flood deposits tend to be cross- the area. In terrain susceptible to debris flow activity, the
stratified and show fining-upward grading. mapping exercise should consider the way different land-
scape elements (e.g. hill slopes, valley floors and stream
channels) interact to generate the potential for large flow
Recording evidence of surface processes events. In other circumstances, the recognition of clay-
Evidence of surface process should be recorded (either rich solifluction sheets indicates the potential for pre-
active, recent or historic). Landsliding can be identified by existing shear surfaces that might be reactivated by
the presence of characteristic landslide morphology or inappropriate construction works.
'indicators of instability' (e.g. tension cracks, back-tilted
trees, heave structures, etc.); floods or debris flows can be
recorded by the upper limits of flow (e.g. from damaged Engineering geomorphology
tree trunks or strandlines) on the margins of floodplains;
the net direction of sand transport can be identified from The geomorphological map should provide a spatial
the build-up of material adjacent to groynes, nebkha framework for appreciating the processes and mechanics
(bush mounds in deserts) or windbreaks. The recording of of landscape change over a range of timescales, particu-
surface drainage features, including areas of seepage, larly timescales relevant to engineers. The spatial frame-
should form part of the mapping exercise. This is particu- work provides 'context' for investigating problems. The
larly important in areas of known or suspected instability. old adage "site and situation' is clearly important in en-
suring that potential geohazards are identified at an
early stage in the investigation process (e.g. Fookes &
Geomorphological map production Vaughan 1986; Fookes 1997). Hillside sites need to be
seen in the context of the whole slope, river valleys as
A geomorphological map is produced by an ongoing part of the whole catchment, coastal sites as part of a
interpretation of the data collected in the field (or from sediment transport cell (see Lee & Brunsden 2001).
satellite imagery or aerial photographs). A key stage in Experience has shown (e.g. Cooke & D o o r n k a m p
map creation is the identification of suitable mapping 1990) that it is often necessary to produce some form of
units to reflect the scale of mapping and the objectives of summary statement about the significance of the geomor-
the study. Ideally, each unit should have consistent geo- phological conditions (landforms, near-surface materials
morphological characteristics, although internal varia- and surface processes) to the engineering project for
bility of materials or rate of process may be a feature of which it was undertaken. The summary information can
the unit (e.g. in areas mantled by glacial tills). Three broad be presented as"
categories of geomorphological unit can be recognized: 9 a hazard map, e.g. showing the nature and distribu-
1. units reflecting the control of the underlying geology tion of landslide features within the area, suscept-
(e.g. plateau surfaces, lithological benches, cliffs); ibility to landsliding, etc. (e.g. Hearn 1995);
2. units reflecting the activity of surface processes (e.g. 9 a resource map, e.g. showing the distribution of
landslide, fluvial, aeolian features, etc.); aggregate resources (e.g. Cooke et al. 1982);
3. units reflecting modification of the landscape by 9 an extended map legend describing each map unit in
man (e.g. areas of cut-and-fill, quarries, made- terms that are relevant to the end user e.g. the land-
ground, etc.). slide hazard or aggregate resource potential within
each unit (e.g. the Ground Behaviour maps of the
Geomorphological map unit boundaries should gener- Ventnor area (Lee & Moore 1991)).
ally follow morphological boundaries, although some
Geomorphological mapping, however, is only one of a
of the boundaries on the morphological map may be
number of complementary approaches which, when used
redundant. Depending on the map scale these units can
in combination, can lead to the development of an effec-
be portrayed in blocks of contrasting colour or shading
tive geological model. The technique needs to be suppor-
(as used on the 1:2500 scale geomorphological maps of
ted by, inter alia, subsurface investigation, monitoring,
the landslides at Ventnor, Isle of Wight (Lee & Moore
desk study, historical records analysis (e.g. sequences of
1991)) or as stylized symbols (see Demek 1972; Demek &
aerial photos) and a review of the environmental con-
Embleton 1978; Gardiner & Dackombe 1983; Cooke &
trols, in order to fully establish the potential for geohaz-
Doornkamp 1990) or a combination of both.
ards or the availability of resources.
Geomorphological maps should not be limited to a
description of what has occurred; they should also be
able to convey to the user what might occur, i.e. the References
potential for geohazards or sensitivity to change. For
example, the combined body of evidence (site-specific ANON. 1972. The preparation of maps and plans in terms of engi-
and landform assemblages) can allow certain judgements neering geology. Quarterly Journal of Engineering Geology,
to be made about the potential for landslide activity in 5, 293-381.
56 E . M . LEE

BRITISH STANDARDS INSTITUTION. 1981. BS5930. Code of GRIFFITHS, J. S. & MARSH, A. H. 1986. BS5930: the role of
Practice for Site Investigation. British Standards Institu- geomorphological and geological techniques in preliminary
tion, London (updated in 1999). site investigation. In: Hawkins, A. B. (ed.) Site Investigation
BRUNSDEN, D., DOORNKAMP, J. C. FOOKES, P. G., JONES, Practice: Assessing BS5930. Geological Society, London,
D. K. C. & KELLY, J. M. N. 1975a. Large scale geo- Engineering Geology Special Publications, 2, 261-267.
morphological mapping and highway engineering design. GRIFFITHS, J. S. BRUNSDEN,D., LEE, E. M. & JONES, D. K. C.
Quarterly Journal of Engineering Geology, 8, 227-253. 1995. Geomorphological investigation for the Channel
BRUNSDEN, D., DOORNKAMP, J. C. HINCH, L. W. & JONES, Tunnel and Portal. Geography Journal, 161, 257-284.
D. K. C. 1975b. Geomorphological mapping and high- HEARN, G. J. 1995. Landslide and erosion hazard mapping at
way design. Sixth Regional Conference for Africa on Soil Ok Tedi copper mine, Papua New Guinea. Quarterly
Mechanics and Foundation Engineering, 3-9. Journal of Engineering Geology, 28, 47-60.
BRUNSDEN, D., JONES, D. K. C. & DOORNKAMP,J. C. 1979. The JONES, D. K. C. 2001. Ground conditions and hazards: Suez
Bahrain Surface Materials Resources Survey and its City development, Egypt. This volume.
application to planning. Geography Journal, 145, 1-35. JONES, D. K. C., BRUNSDEN,D. & GOUDIE, A. S. 1983. A pre-
BUSH, P., COOKE, R. U., BRUNSDEN,D., DOORNKAMP,J. C. & liminary geomorphological assessment of part of the Kara-
JONES, D. K. C. 1980. Geology and geomorphology of korum highway. Quarterly Journal of Engineering Geology,
the Suez city region, Egypt. Journal of Arid Environments, 16, 331-355.
3, 265-281. JONES, D. K. C., COOKE, R. U. & WARREN, A. 1986. Geo-
COOKE, R. U. & DOORNKAMP,J. C. 1990. Geomorphologyin Envi- morphological investigation, for engineering purposes, of
ronmental Management. Oxford University Press, Oxford. blowing sand and dust hazard. Quarterly Journal of Engi-
COOKE, R. U., BRUNSDEN, D., DOORNKAMP, J. C. & JONES, neering Geology, 19, 251-270.
D. K. C. 1982. Urban Geomorphology in Drylands. Oxford KLIMASZEWSKI, M. 1956. The principles of geomorphologi-
University Press, Oxford. cal survey of Poland. Przeglad Geograficzny, 28 (Suppl.),
COOKE, R. U., BRUNSDEN, D., DOORNKAMP, J. C. & JONES, 32-40.
D. K. C. 1985. Geomorphological dimensions of land devel- K LIMASZEWSKI,M. 1961. The problems of the geomorphological
opment in deserts - with special reference to Saudi Arabia. and hydrographic map of the example of the Upper Silesian
Nottingham Monographs in Applied Geography No. 4. industrial district. Polish Academy of Sciences Institute of
DEMEK, J. (ed.) 1972. Manual of Detailed Geomorphological Geography.
Mapping. Academia, Prague. LEE, E. M. 1999. In Amenas Gas, Algeria." baseline terrain
DEMEK, J. & EMBLETON, C. (eds) 1978. Guide to Medium-scale evaluation report. Confidential Report to Environmental
Geomorphological Mapping. International Geographical Resources Management.
Union, Brno. LEE, E. M. & BRUNSOEN,D. 2001. Sediment budget analysis for
DOORNKAMP, J. C. 1982. The physical basis for planning in the coastal management, Dorset. This volume.
Third World. Third World Planning Review, 4, 11-3 I. LEE, E. M. & MOORE, R. 1991. Coastal Landslip Potential:
DOORNKAMP, J. C. (ed.) 1988. Planning and Development: Ventnor, Isle of Wight. Department of the Environment,
applied earth science background, Torbay. M1 Press, London.
Nottingham MOORE, R., CLARK, A. R. & LEE, E. M. 1998. Coastal cliff
DOORNKAMP, J. C., BRUNSDEN, D., JONES, D. K. C., COOKE, behaviour and management: Blackgang, Isle of Wight. In:
R. U. & BUSH, P. R. 1979. Rapid geomorphoiogical assess- MAUND, J. G. & EDDLESTON, M. (eds) Geohazards and
ments for engineering. Quarterly Journal of Engineering Engineering Geology. Geological Society, London, Special
Geology, 12, 189-204. Publications, 15, 49-59.
DOORNKAMP, J. C., BRUNSDEN,D., JONES, D. K. C. & COOKE, MORGAN, R. P. C. 1995. Soil Erosion and Conservation, 2rid
R. U. 1980. Geology, Geomorphology and Pedology of edn. Longman, Harlow.
Bahrain. GeoBooks, Norwich. PHIPPS, P. J. 2001. Terrain systems mapping. This volume.
FOOKES, P. G. 1997. Geology for engineers: the geological RICHARDS, K. S., BRUNSDEN,D., JONES, D. K. C. & MCCRAIG,
model, prediction and performance. Quarterly Journal of M. 1987. Applied fluvial geomorphology: fiver engineer-
Engineering Geology, 30, 290-424. ing project appraisal in its geomorphological context. In:
FOOKES, P. G. & VAUGHAN, P. R. 1986. A Handbook of RICHARDS, K. S. (ed.) River Channels: Environment and
Engineering Geomorphology. Blackie, Glasgow. Process. Blackwell, Oxford, 348-382.
FOOKES, P. G., LEE, E. M. & SWEENEY,M. 2001. Pipeline route SAVIGEAR, R. A. G. 1965. A technique of morphological map-
selection and ground characterization, Algeria. This volume. ping. Annals of the Association of American Geographers,
GALON, R. 1962. Instruction to the detailed geomorphological 53, 514-538.
map of the Polish Lowland. Polish Academy of Science, THOMPSON, A., HINE, P. D. GREIG, J. R. & PEACH, D. W. 1996.
Geography Institute of Geomorphology and Hydrography Assessment of subsidence arising .from gypsum solution.
of the Polish Lowland at Torun. Symonds Travers Morgan Report to the Department of
GARDINER, V. & DACKOMBE, R. 1983. Geomorphological Field the Environment.
Manual. George Allen and Unwin, London. TRICART, J. 1965. Principles et mdthodes de la gOomorphologie.
GEOTECHNICALCONTROLOFFICE. 1988. Guide to Rock and Soil Masson, Paris.
Descriptions. Geoguide 3, Hong Kong Government. VERSTAPPEN, H. TH. 1983. Applied Geomorphology: Geomorph-
GEOTECHNICAL ENGINEERING OFFICE. 1996. Guide to Site ological Surveys for Environmental Development. Elsevier,
Investigation. Geoguide 2, Hong Kong Government. Amsterdam.
GRIFFITHS, J. S. 2001. Development of a ground model for the WATERS, R. S. 1958. Morphological mapping. Geography, 43,
UK Channel Tunnel portal. This volume. 10-17.
Geological Society, London, Engineering Geology Special Publications
Geographical information systems
C. P. Nathanail and A. Symonds

Geological Society, London, Engineering Geology Special Publications 2001; v. 18; p. 57-58
doi:10.1144/GSL.ENG.2001.018.01.09

© 2001 Geological Society of London

Geographical information systems

C. P. Nathanail & A. Symonds

School of Chemical Environmental Engineering, Nottingham University, Nottingham, U K

Definition of Geographical Information GIS data models

Systems (GIS)
Present day GIS store information in both raster and
GIS software can be used to manipulate and display spa- vector formats. Raster data comprise 'pixels' or tiles that
tial information. Burrough & McDonnell (1998, p. 11) fill space, thus something is stored about everywhere.
provide a number of definitions of GIS, based upon the The raster data structure has its origins in remote sens-
concepts of the toolbox, database and organizational ing and digital image processing software. Vector data
systems. Examples of each include: are stored as points, lines or polygons, thus something
is stored only where a change occurs. The vector data
9 Toolbox definition: '... a system for capturing, storing, structure has its origin in computer aided design (CAD)
checking, manipulating, analysing and displaying data packages. Increasingly software allows users to store
which are spatially referenced to the earth ...' both types of data and to swap between one format
9 Database definition: ' . . . any manual or computer based and another for optimal data processing, display or
set of procedures used to store and manipulate geogra- reporting.
phically referenced data ...'
9 System definition: 'An organized collection of compu-
ter hardware, software, geographic data and personnel
designed to efficiently capture, store, update, manip- G IS uses and users
ulate, analyse, and display all forms of geographically
referenced information'. The core of a GIS is its database (Worboys 1995).
9 Organization based definition: '... a decision support Spatial data representing geographical phenomena are
system involving the integration of spatially referenced stored in this in terms of: their position with respect to
data in a problem solving environment ...' a known co-ordinate system; their attributes that are
unrelated to position; and their spatial interrelations
In the context of terrain evaluation the 'organiza- (Burrough & McDonnell 1998). Through computer
tion based' definition is considered to be the most hardware components (e.g. monitor, keyboard, mouse,
appropriate. This is because the intended output of printer, plotter, scanner, digitizer (Burrough & McDon-
terrain evaluation is to support a decision making nell 1988; p. 12-14; Worboys 1997, p. 1-3; Martin 1996,
process. More specifically, GIS can help the terrain p. 10-11)). GIS software is used to perform five generic
evaluation process answer questions such as those in functions on these data (Martin 1996; Burrough &
Table 1. McDonnell 1998):

(a) Data input and verification - the capture of spatial

Table 1. Spatial queries in terrain evaluation (Modified from
Nathanail 1994) data and conversion to a digital form
(b) Data storage and database management - the way
Where is the slope greater than x~ topology, attributes of geographical elements (e.g.
What material is at x metres depth? points, lines, polygons) are structured and orga-
How far from the fault will the excavation be at x metres depth? nized in the system
How many water tables are there and where are they? (c) Data output and presentation - the method of dis-
What is the orientation of bedding across the site? play and manner in which data are output to the user
Where is the strength of clay less than x kPa? (d) Data t r a n s f o r m a t i o n - deals with the aspect of
Can bored piles be used here?
Can the sandstones in these boreholes be correlated? data correction and data analysis
How long will it take the water to flow from here to there? (e) Interaction with the user - a range of menu driv-
Is more information needed; if so what information and where? en, macro & programming language and 'hotkey'
functions

From: GRIFFITHS,J. S. (ed.) Land Surface Evaluationfor Engineering Practice. Geological Society, London, Engineering Geology
Special Publications, 18, 57-58. 0267-9914/01/$15.00 9 The Geological Society of London 2001.
58 C.P. NATHANAIL & A. SYMONDS

Table 2. A summary of standard GIS functionafity As the price of computer systems and GIS fell in the
(after Symonds 1999) 1980s and 1990s, the range of users increased consider-
ably: emergency services; regulators; insurers; financial
Function Sub-sets of function institutions; estate agents; surveyors; environmental
Capture Digitizing/scanning groups; and, property developers (DOE 1994). As if to
Raster/vector conversion confirm its wide appeal, Burrough & Frank (1995) have
Co-ordinate/projection transformation described GIS as a 'generic toolbox ...', with the ability
Construct topology to solve many different problems for different user types.
Storage Relational/object-oriented databases The development in the resolution of global position-
Data integration ing systems and the drop in their price means there is no
excuse for not collecting accurate locations of where
Retrieval Data browsing observations were made or samples taken. Without such
Windowing location information, GIS can be of no use.
Query generation
At the time of writing desk top GIS have power and
Manipulation Map generalization capability that would have required workstations only a
Map abstraction couple of years ago. The key to any GIS implementation
Reclassification remains an adequately populated and maintained data-
Scale change
Linear and rubber sheeting base, well trained staff and operating procedures that
explicitly take the presence of GIS into account.
Analysis Statistical summary
Buffer generation
Polygon overlay/dissolve
Measurement - distance areas, volumes References
Least-cost routes
Network analysis BURROUGH, P. A. & FRANK, A. U. 1995, Concepts and para-
Spatial/environmental modelling digms in spatial information" are current geographical
Surface modelling and analysis information systems truly generic? International Journal of
Map algebra Geographical Information Science, 9, 10l- 115.
BURROUGH, P. A. & MCDONNELL, R. A. 1998. Principles of
Display Visualization GIS. Oxford University Press, Oxford.
Multiple map views DEPARTMENTOF THE ENVIRONMENT. 1994. Contaminated Land
Multiple feature displayed Research Report: Information systems and land contamina-
Tabular, chart, graphic output tion, CLR Report No. 5, DOE, London.
G3 SOLUTIONS1999. The Countryside Information System Year
From: G3 Solutions 1999. 2000 Project, available at: http://g3solutions.freeserve.co.uk
GRIMSHAW, D. J. 1994. Bringing Geographic Information Sys-
Worboys (1995) suggests that there are a number of tems into Business, Longmans.
HENDRIKS, P. H. J. 1998. Information strategies for GIS. Inter-
analytical requirements for a GIS: resources inventory;
national Journal of GIS, 12, 621-639.
network analysis; terrain analysis; layer based analysis; MARTIN, D. 1996. Geographic Information Systems: socioeco-
location analysis; spatial-temporal information. Martin nomic applications, 2nd edn. Routledge, London.
(1996, p. 59-60) adds reclassification and neighbour- NATHANAIL, C. P. 1994. Systematic modelling and analysis of
hood characterization. He also suggests that sequences digital data for slope and foundation engineering. PhD
of these functions are known as cartographic modelling. Thesis, London University.
Within most GIS are a large range of subsets of each of STRACHAN,A. J. • STUART,N. 1996. UK Developments in Envi-
these forms of analysis (Table 2). ronmental GIS. International Journal of GIS, 10, 17-20.
This functionality leads GIS to have many potential SYMONDS, A. 1999. Developing a contaminated land inspection
uses and users. For example, uses include: environmen- strategy for local authorities using a Geographical Informa-
tion System. MSc Thesis, University of Nottingham.
tal modelling (Strachan & Stewart 1996); forestry, land TOMLINSON, R. F. 1987. Current and potential uses of geogra-
registry, utilities, transport and engineering applications phical information systems: The North American experi-
(Tomlinson 1987; Martin 1996); business strategy devel- ence. International Journal of GIS, 1, 203-218.
opment (Hendriks 1998); business applications (90% of WORBOYS,M. F. 1997. GIS: A Computing Perspective. Taylor &
business users data spatially related, Grimshaw 1994). Francis, London.
Geological Society, London, Engineering Geology Special Publications
Terrain systems mapping
P. J. Phipps

Geological Society, London, Engineering Geology Special Publications 2001; v. 18; p. 59-61
doi:10.1144/GSL.ENG.2001.018.01.10

© 2001 Geological Society of London

Terrain systems mapping

P. J. Phipps
Mott MacDonald, Croydon, Surrey, U K

Description A terrain system covering a large area can be sub-

divided into a number of defining terrain facets such as a
The term 'terrain systems mapping' originated in the fiat-topped interfluve or major valley which assist in
1930s and early 1940s when the requirement was identi- describing the terrain system. It is the terrain facet that
fied to classify large areas of terrain for the purposes of provides the basic landform unit. There will undoubt-
locating potential agricultural and economic resources, edly be variations in the attributes of a terrain facet,
and identifying suitable sites for development in mainly although the surface materials, underlying geology,
undeveloped rural areas (Mitchell 1973). The view is water table, slope inclinations and slope stability will
adopted in this paper that 'terrain' and 'land' are syn- be relatively uniform within certain definable bound-
onymous, which follows both Christian & Stewart aries. Terrain facets have typically been mapped at
(1968) and Townshend (1981). 1:10000 to 1:60000 scale although Waller & Phipps
In many respects the methodology behind terrain (1996) have again shown their applicability at larger
systems mapping has not altered since one of its first sig- scales of 1 : 2500 to 1 : 5000.
nificant applications came to prominence in Australia, The terrain facets may themselves be further broken
with the publication of a series of reports summarizing down to terrain elements which are features that could
terrain systems mapping of the Australian Territories reasonably be identified and mapped directly in the field.
(Christian & Stewart 1952). Areas up to hundreds of A terrain facet that describes a major valley could com-
square kilometres were delineated in which characteristic prise terrain elements such as a stream channel, flood
assemblages of topography, soils and vegetation could plain, ox-bow lake etc. (Fig. 1).
be identified. This association is described in Cooke & The most significant advances in terrain systems map-
Doornkamp (1990) after Stewart & Perry (1953). These ping since the last working party report have not affected
authors established that the topography and soils are the general approach, but the techniques for the estab-
dependent on the nature of the underlying rocks lishment of hierarchical models have improved. The
(geology), the erosional and depositional processes that widespread availability of satellite imagery and high-
have produced the present topography (geomorphol- powered desk-top computing facilities have enabled
ogy) and the climate under which these processes have initial delineation of terrain systems boundaries to be
operated. Thus the land system is a scientific classi- carried out accurately and rapidly. Further detailed divi-
fication of country or landscape based on topography, sions at the terrain facet scale by aerial photograph
soils and vegetation correlated with geology, geomor- interpretation or photogrammetric techniques can then
phology and climate. be augmented by other Earth science information layers
The basic good practice for developing a suite of ter- which could include vegetation, ground investigation
rain systems maps has been succinctly covered in the 1982 information and hydrogeological readings. All the data
Engineering Group Working Party Report. The three can be managed and presented through computerized
main landform units of terrain system, terrain facet and geographical information systems. Geostatistical tech-
terrain element are defined and described. Terrain sys- niques are available which can be applied to the infor-
tems maps are hierarchical to an extent which depends mation to provide, for example, levels of confidence for
on the scale adopted, mapping requirements and brief potential soil thicknesses in a specific terrain facet.
(Fig. 1). The recognizable pattern forming the terrain However, there is still a fundamental requirement in
systems have normally been mapped at a scale of understanding the various geological, geomorphological
1: 250 000 to 1:1000 000. However, recent experience and climatological processes that combine to sculpt the
from studies for the Channel Tunnel Rail Link has terrain of any area. Site reconnaissance and/or ground
shown that the application of mapping at 1 : 50 000 scale truthing are required to provide the basis for delineat-
for civil engineering projects is most suitable when in- ing terrain systems and terrain facets, and in validating
tegrating detailed satellite imagery and re-evaluating the models and maps that are developed. Without the
geological mapping (Waller & Phipps 1996). fieldwork element the terrain systems approach is not

From: GRIFFITHS,J. S. (ed.) Land Surface Evaluationfor Engineering Practice. Geological Society, London, Engineering Geology
Special Publications, 18, 59-61. 0267-9914/01/$15.00 9 The Geological Society of London 2001.
60 P.J. PHIPPS

Terrain System Gentle hills with broad terraced river valleys

~"
Terrain Facets

~._ Hill

JL.
'~ 2. Mir ~r valley

[ ' ~ ' ~ ~ ~ , , ~ , ~ --,~ 4. Major valley

Field
Mapping
Features l

3b. Bluff 3a. Surface 2a. Floor 2b. Stream

f 4a. Abandoned
channel
4b. Stream
II ] l a. Slope
lb. Gully side
l d. Stream
4c. Floodplain lc. Gully floor
Fig. 1. Relationship between terrain systems and terrain facets.

rigorous and can lead to misinformation on a large scale visible on remotely sensed imagery whether satellite
that can be very costly to end users. mounted or from aerial photography. The resolution,
wavelength sensitivity, processing and enhancement will
all affect the specific nature of images that are to be used
for interpretation. Different images will highlight differ-
Limitations ent characteristics of the ground surface which will affect
visual distinctions. Remote sensing surveys commis-
The most significant limitation in carrying out a terrain sioned specifically for a mapping study will need to be
systems mapping programme is related to the quality specified with much care. Furthermore, the availability
and nature of the base information. System and facet of existing topographical and geological maps, ground
boundaries can only be delineated where distinct and investigation data and reports, and any other potential
 TERRAIN SYSTEMS MAPPING 61

supporting information will assist in developing more applications would include low-cost roads in the Hima-
representative mapping schemes and ground models at layas (Fookes 1997), a high-speed rail link in southeast
an early stage. These can then be augmented by targeted England (Waller & Phipps 1996), and road design in the
ground investigations, detailed field mapping or remote Libyan desert (Hunt 1979).
surveys to improve the information level in less well Within the above context of civil engineering appli-
understood terrains. cations, terrain system maps can be developed over a
The detail and appropriateness of the terrain system wide areal extent for assisting in identifying locations
maps produced are also dependent on the skill and for potential dam sites, quarries, airfields and large
experience of the person undertaking initial interpreta- structures.
tions, and information review and synthesis. An inter- Terrain systems mapping has also been adopted to
preter experienced in a specific terrain or environment assist in a wide variety of applications that are not
should be able to provide higher information levels and directly related to engineering issues. A summary of these
summarize the characteristics of facets in a more efficient has been provided by Mitchell (1973). Such usages
and cost-effective way than someone who is not familiar include soil science, agriculture and forestry, meteorol-
with that environment. ogy, microclimatology and hydrology, and resource
Finally, the scales adopted for terrain systems mapping analysis for landscape and recreational planning. Terrain
are such that even though large areas can be mapped systems maps have also been used for military purposes
rapidly and information summarized on a terrain facet in determining suitability of off-road mobility for differ-
level, detailed information at a site-specific scale is often ent vehicles, identifying sites of relatively easy excavat-
not available. To design a particular engineering struc- ability and for construction materials.
ture or earth structure, civil engineers and geotechnical
engineers require information on ground conditions and
geotechnical parameters specific to the location of the
structure. The variability inherent within a terrain facet
References
will normally be too great to allow all but the most CHRISTIAN,C. S. 8r STEWART,G. A. 1952. Summary of General
provisional of designs to be developed. Report on Survey of Katherine-Darwin Region 1946. Land
Research Series, 1, CSIRO, Australia.
CHRISTIAN,C. S. 8r STEWART,G. A. 1968. Methodology of Inte-
grated Surveys. Proceedings of Conference on Aerial Sur-
Applications veys and Integrated Studies, Toulouse, Unesco, 233-280.
COOKE, R. U. & DOORNKAMP, J. C. 1990. Geomorphology in
By far the most appropriate and effective use of terrain Environmental Management. Clarendon Press, Oxford.
systems mapping is for feasibility studies related to long ENGINEERING GROUP WORKING PARTY 1982. Land surface
linear civil engineering structures that include railways, evaluation for engineering practice. Quarterly Journal of
roads, water supply pipelines, canals and to a lesser Engineering Geology, 15, 265-316.
extent tunnels. The maps and supporting information FOOKES, P. G. 1997. Geology for engineers: the geological
model, prediction and performance. The First Glossop
produced can not only assist in identifying potential Lecture. Quarterly Journal of Engineering Geology, 30,
alignments, but also provide information on engineering 293-424.
geological and geotechnical issues such as topography, HUNT, T 1979. Geotechnical aspects of road design in Libya.
bedrock geology, soils, groundwater, geomorphological Ground Engineering, October, 15-19.
and geological hazards, aggregate resources, excavat- MITCHELL, C. W. 1973. Terrain Evaluation. Longrnan, London
ability, fill potential, foundation and slope stability. This STEWART, G. A. & PERRY, R. A. 1953. Survey of Townsville-
suite of information topics is dependent on engineering Bowen Region (1950). Land Research Series 2, CSIRO,
requirements and the specific environment. Identification Australia.
of facets with high saline potential for roads would be TOWNSnEND, J. R. G. (ed.) 1981. Terrain Analysis and Remote
Sensing. George Allen & Unwin.
important in desert systems and the depth of residual soil WALLER, A. M. & PHII'PS, P. J. 1996. Terrain systems mapping
development and weathering profiles in tropical environ- and geomorphological studies for the Channel Tunnel Rail
ments for excavatability of pipeline trenches. Published Link. In: CRAIG,C. (ed.) Advances in Site Investigation Prac-
examples of the terrain systems mapping approach for tice. Proceedings of the International Conference, London,
long linear structures are numerous, but some recent 30-31 March 1995. Thomas Telford, London, 25-38.
Geological Society, London, Engineering Geology Special Publications
Surface and groundwater resources survey in Jordan
R. J. Allison

Geological Society, London, Engineering Geology Special Publications 2001; v. 18; p. 65-71
doi:10.1144/GSL.ENG.2001.018.01.11

© 2001 Geological Society of London

Surface and groundwater resources survey in Jordan

R. J. Allison
Department of Geography, University of Durham, Durham, U K

Purpose of survey In the northeast Badia of Jordan there are two major
sources of water (A1-Homoud et al. 1995). Due to the
Water is one of the most valuable physical resources topographic effect of the Jebel Druz, precipitation totals
in the arid zone. Much effort is made by engineers to exceed 500 mm a -1 in the north, declining to <50 mm a -1
maximize availability and minimize wastage. In the in the south. During winter months, runoff can be
Hashemite Kingdom of Jordan water scarcity has been considerable but no reasonable data exist on parameters
exacerbated in recent years by rapidly rising demand. such as wadi discharge, infiltration rates and drainage
In the early 1990s total national water consumption in basin contributing areas under storms of a given mag-
Jordan approached 730 • 106m 3. It is estimated that nitude. Groundwater is found in three aquifers. Numer-
demand will rise to 1200 x 106m 3 by the year 2000. ous government-operated and private wells have recently

Fig. 1. The study area.

From: GRIFFITHS, J. S. (ed.) Land Surface Evaluation for Engineering Practice. Geological Society, London, Engineering Geology
Special Publications, 18, 65-71. 0267-9914/01/$15.00 9 The Geological Society of London 2001.
66 R. J. ALLISON

been drilled to exploit groundwater. There is little must be based on the development of a suitable geo-
information on the recharge:extraction balance, chan- logical and geomorphological ground model.
ging spatial patterns of water availability and temporal
changes in water quality, despite trends which hint at a
depleting resource.
The purpose of the research was to provide data for The site
enhancing efficient use of a scarce water resource. Sup-
porting objectives for the surface water study included The northeast Badia encompasses 11 200km 2 of land
determining water availability, establishing patterns of (Fig. 1), around 14% of the total land area of Jordan.
runoff generation, quantifying sediment mobilization The spatial limits of the study broadly coincide with the
and transport rates and locating potential water harvest- margin of late Tertiary and early Quaternary basalt lava
ing sites. Supporting objectives for the groundwater flows (Burdon 1959; Bender 1974, 1975), which spread
study included determining the rate of aquifer draw- from local eruptive centres (Fig. 2). The resulting basalt
down and recovery as a consequence of pumping, plateau is between 50 km and 170 km wide from east to
quantifying any permanent fall in water table height west and 180 km from north to south. The basalts are
and aquifer storage, identifying the causes of aquifer alkaline-olivine in character (Ibrahim 1992). Absolute
pollution and establishing patterns of groundwater age determination using K-Ar techniques has identified
quality deterioration. Interaction between the surface a number of basalt lava flows, with ages ranging from
and groundwater components of the project is signifi- 13.7 Ma to less than 0.5 Ma for the exposed rocks. Some
cant. Understanding the overall hydrological regime of the unexposed flows which constitute part of the
requires synthesis of both parts of the study and these groundwater system date at 23 Ma.

Fig. 2. Geological map and stratigraphic column of the northeast Badia (source: Bender 1974).
 SURFACE AND GROUNDWATER RESOURCES SURVEY IN JORDAN 67

Topography rises from a low of 400 m in the south to is poorly understood, hydrothermal and heavily sul-
1200 m around the town of Mafraq. There is an accom- phurous. The middle aquifer dips beneath the study area
panying increase in rainfall, which is seasonal and often at depths of 400 m to 700 m. Its transmissivity is low at
concentrated in high magnitude/low frequency runoff 35 m 2 d to 450m 2 d and water quality is variable. The
generating storms. The region is dominated by low, upper aquifer is the most important groundwater re-
gently undulating hills and elevation differences of 25 m source. It consists of recent sediments, Tertiary basalts
to 30 m between high and low points across the land- and limestones and is separated from the poorer quality
scape. Gradients are seldom steep and there are few middle and lower aquifers by a thick marl aquitard. The
sudden breaks of slope. Topographic highs provide local elevation of the upper aquifer is little more than 50 m in
watersheds, particularly where they occur in lines along the north to around 350m near Safawi. Flow is radial
dykes or major fault systems. Regional structure is towards the Azraq basin, with the saturated thickness
relatively simple. Three dominant fault systems run east decreasing from >300m in the north to <50m in the
to west, northwest to southeast and east-northeast to south. Pump tests show transmissivity to be highly
west-southwest. Structural lineaments have some con- variable but exceeding 3000 m 2 d in some places.
trol on groundwater flowpaths.
The age of the basalt flows determines the degree of
drainage network development, most readily observed
by wadi connectivity. The oldest lava flows have a gentle, Available information
rounded topography, well developed wadi systems and
a fine colluvial rill network. The most recently emplaced A desk study at the start of the project revealed a
basalts have a more rugged topography and poorly considerable amount of data, much as unpublished gov-
developed drainage network. As the basalts have weath- ernment reports. The northeast Badia has been topo-
ered since emplacement, boulder fields have evolved, graphically mapped at a scale of 1: 50000. Remote
with clasts covering the ground surface and overlying sensing imagery is available including swaths of Landsat
fine-grained, light orange sediments. The size of clasts TM and SPOT digital data and air photographs. The
and the degree of protection which they afford to the Geology Directorate of the Natural Resources Author-
ground surface varies. The Abed weathers to large, ity have mapped stratigraphy and structure at a scale of
rounded boulders, with exposed bare ground. The Bish- 1:250000. Subsurface geophysical surveys have been
riyya produces a fine reg of basalt chips, which leaves completed, revealing detail about the structure and the
very little of the underlying ground surface exposed. The groundwater reserve.
ground surface boulder cover affects surface runoff and Weather records confirm that much of the year
sediment mobility. is dominated by low precipitation and high potential
Different stratigraphic units act as either aquifers or evapotranspiration (Korzon 1974). Precipitation seldom
aquitards, depending on their physical properties (Fig. 3). exceeds 500mma -1 and potential evapotranspiration
The lower aquifer occurs at depths of 1.3 km to 3.4 km, can approach 2000mma -1, a consequence of mean

= ~,
WEST "~
~ r~, ~- ~" FAST
,o >,

< ~ :~ ~ ~ Pettic
1000 - ~ Recent alluvial sediments :~

.,, x x ....... "" " ~ x ~ " ~ o -

.go
: o

-soo I I' ' ' ! I !

50 100 150 200 250
Distanee (kin)

Fig. 3. Cross-section highlighting the main geological units relevant to groundwater flow patterns in the
northeast Badia.
68 R . J . ALLISON

annual maximum temperatures of 34~ to 37~ The radiation and the net radiation balance. The data were
Water Authority of Jordan monitors the three ground- used to establish the hydrological flux and balance
water systems but principally the Upper aquifer. Hun- between times of net water surplus and deficit. Field
tings Technical Survey has undertaken soil mapping infiltration tests were undertaken using a constant head,
and analysis as part of the National Soils and Land double-ring infiltrometer. Tests were conducted on a
Use Project. variety of surfaces to partition the landscape into zones
A number of the information sources comprise incom- of high, medium and low infiltration.
plete data sets. There is no guarantee that a complete Surveys in wadis and across their bounding slopes used
groundwater sampling programme will be undertaken an electronic distance measurer. Wadi gradients were
during site visits, for example, and time-series data determined for channel reaches at different points along
include breaks in the record. Variations in the hydro- drainage networks. Clasts were sampled and measured
logical conditions both at and beneath the ground along their a- b- and c-axis for palaeohydrological
surface have been so rapid in recent time that there is reconstruction. Lines of painted stones of differing size
doubt as to whether the resolution of available informa- were installed along wadi cross-sections. Sites were
tion is adequate for establishing patterns of change. revisited to see whether flow competence during wet
seasons reached magnitudes sufficient to move clasts,
thereby permitting discharge calculations. Sequences
of slope profiles were surveyed between topographic
Techniques used highs and lows to establish gradient, length and run-
off potential.
Surface hydrology runoff pathways were established by Groundwater studies included tests undertaken at the
digitizing 1 : 50 000 topographic maps. Contour patterns well-head and samples collected for laboratory analysis.
and drainage networks were recorded. The Geographi- All samples were collected from purged boreholes.
cal Information System package ARC-Info was used to A stable well-head electrical conductivity was used to
examine the hydrological regime. An automatic weather confirm adequate pre-sample pumping. An electronic
station and data logger were installed towards the water level probe was used to establish the height of the
middle of the study area to measure precipitation, wind water table and pump tests were undertaken to quantify
speed and direction, wet and dry bulb temperature, solar the rate of draw-down and recovery. Total dissolved

Table 1. Details of wells which are characteristic of Upper and Middle aquifer water extraction sites in the northeast Badia

Well Location Aquifer type Pump rate Conductivity Temperature Dissolved p H Total Total
no. Eastings Northings and lithology ( m 3 s -1 h -1) mmmho/cm (~ oxygen cations anions
(%) (meq1-1) (meq1-1)

1 North of region Basalt 58 300 29.7 88 8.42 2.79 2.81

Druze foothills
316570 184875
2 Lower Druze foothills Basalt and 45 610 34 87 8.14 5.14 5
322300 174500 rijam
limestone
3 Azraq wellfield north Basalt and 180 610 26 100 8.28 5.51 5.05
321508 152443 rijam
limestone
4 Azraq town Basalt 10 920 23.2 90 7.15 18.21 18.05
321425 143800
5 Central Azraq basin Rijam n/a 1400 24.5 59 7.47 12.05 11.73
320750 138900 limestone
Eastern margin of Basalt 60 1770 22.3 25 7.12 17.47 17.77
Azraq basin
338200 140550
South-west margin of Azraq 30 2140 23.5 89 7.05 23.89 23.05
Azraq basin formation
316420 130033
 SURFACE AND GROUNDWATER RESOURCES SURVEY IN JORDAN 69

oxygen was monitored during pump tests using a meter sediment. On the youngest basalts, surface water flow is
with an attached thermometer to correct for tempera- highly localized. Water runs into small, closed depres-
ture. Electrical conductivity and pH were recorded. sions or pans known as qa. The qa are usually no more
Samples of water were collected for well-head titration than 100 m 2 or so in size and seldom linked. On older
tests to determine alkalinity. Water (50 ml) was titrated basalts there are qa which are much larger in size, some
against 0.16N or 1.6N H2SO4 using phenolphthalein exceeding 20 km 2. The pans are fed from extensive areas
indicator if CO~ anions were present and screened and are often supplied by wadis as well as overland flow.
methyl orange to determine the concentration of HCO~ Large qa have high surface evaporation rates and are
anions. Bottles of water were returned to the laboratory usually saline, limiting the value of water once it has
for detailed geochemical analysis. Two 30ml, filtered entered the depression. A third type of pan, known
samples were collected at each site. One sample was locally as marab, evolves where sediment is deposited
acidified with three drops of 50% HC1. A 50 ml sample along the course of wadis if channelled systems open out
was collected for isotope analysis and at six sites a across wide areas. Marab gradients encourage the move-
sample was collected for 14C dating. ment of water from their up-stream to their down-
stream end. Most are fed by an extensive up-basin
network of wadis, resulting in significant water inunda-
tion during wet seasons.
Conceptual model Wadis drain radially into the Azraq basin. Wadi
gradients range from 5% to 10% on the footslopes of
The northeast Badia is an environment where water is the Druze and 3% to 5% in the south of the region.
limited and demands on the resource are increasing. During the study wadi flow was limited and many of
Developing technologies permit surface water harvesting the cross-section painted boulder lines did not move.
and groundwater extraction in increasingly large vol- Palaeohydrological analysis of wadi bed deposits con-
umes, which cannot be sustained. Alterations to one firmed that significant discharge levels do occur and there
component of the hydrological regime will affect other is potential to use surface water more efficiently. Infiltra-
parts of the system. tion rates are generally high in wadi beds but decrease
The surface water regime is characterized by long rapidly on qa, marab and towards interfluves. Where a
periods of the year with a negative hydrological balance. combination of high infiltration rates and appropriate
Precipitation-induced runoff results in wadi flow, with structural controls exist, there is the potential to sup-
flood hydrographs passing rapidly through the drainage plement groundwater recharge.
network. Groundwater extraction is increasing, particu- Groundwater studies confirm that the total available
larly from the upper aquifer, where the water table is resource and its quality are declining. Recharge to the
close to the ground surface. Access to groundwater is upper aquifer is mainly through direct infiltration in
determined by the thickness of permeable strata and the the northern part of the basin and through wadi beds.
regional dip. The present rate of water use cannot be Contemporary recharge estimates vary between 22 and
maintained. Both quality and volume of the upper 36 Mm 3 a -l . Water quality in the upper aquifer is gen-
aquifer are declining and action needs to be taken to halt erally good but salinity levels are increasing. They range
over-exploitation. from a few hundred to 4000ppm. At some well-
heads the total dissolved solids is approaching a point
where the water becomes marginal for human consump-
tion, principally because of NaHCO3 and NaC1 levels
What the survey established (Fig. 4). Extraction is largely uncontrolled from many of
the region's 600 boreholes, 75% of which are unlicensed.
Somewhere between 85% and 92% of precipitation is High evaporation rates where farmers are over-watering,
lost to evaporation, 5% to 11% is lost to infiltration and in combination with decreasing resource quality, is
2% to 4% generates runoff. Figures are highly seasonal. leading to ground surface salinization. Poor well-head
There are periods of the year during the spring, summer completion often leads to contamination. Lead concen-
and autumn when the hydrological flux is in deficit. trations in the south, for example, are the result of fuel
Much of the groundwater used for crop irrigation evap- spills and ingress down boreholes.
orates rather than infiltrating into the upper soil layers The general conclusion is that the hydrological regime
for plant uptake. During the wet season rainstorm events in the northeast Badia is a function of surface water
frequently generate overland flow. In places where the dynamics on the one hand and the groundwater regime
basalt boulder ground cover leaves exposed areas, sedi- on the other. Both are linked. Modifications to one
ment movement can be significant. component of either system have broad consequences.
Ground surface hydrology is affected by the degree Expedient and integrated use of surface and groundwater
of drainage network development on basalt lava flows of from the upper aquifer has the potential to support
different ages. Pans act as collection zones for water and sustainable development. The regional water authority
70 R. J. ALLISON

Fig. 4. Contour map of total dissolved solids of the upper aquifer complex.

plan is for the borehole network to the upper aquifer to of similar studies in dryland environments include work
be extended and abstraction raised by a further 3.9 x in Syria and countries of the Arabian Peninsula. Surface
lO6 m 3. Such a development, with no further integration water surveys have been completed in Bahrain as part of
of the surface and groundwater regimes, is likely to gen- its earth surface and materials resources survey (Bruns-
erate a gradually declining water resource in a hydro- den et al. 1979; Doornkamp et al. 1980). Wadi flow
logically sensitive area. regimes have been examined for their potential to
maximize available water in the Sultanate of Oman
(Doyel et aL 1984; Maizels & Anderson 1988). Work has
been completed on runoff dynamics and wadi discharge
Similar applications during high magnitude/low frequency events in the
Negev desert (Schick 1977; Yair 1983, 1992) and parts of
Aspects of the wadi surveys would not be possible in semi-arid United States of America (Patton & Baker
perennial channels and issues such as throughflow in the 1977; Abrahams et aL 1992).
unsaturated zone and the effects of vegetation would The groundwater programme is transferable to most
have to be considered in many environments. Examples aquifers. Sampling frameworks have to be designed
 SURFACE AND GROUNDWATER RESOURCES SURVEY IN JORDAN 71

relative to individual aquifer systems. The time between BURDON, D. 1982. Hydrogeological considerations in the
sampling a n d analysis is an i m p o r t a n t issue where stored Middle East. Quarterly Journal of Engineering Geology,
water is likely to deteriorate. Examples o f similar studies 15, 71-82.
CHARALAMBOUS, A. N. 1990. Hydrogeology of the Disi Sand-
in o t h e r d r y l a n d e n v i r o n m e n t s include w o r k in the
stone Aquifer. UNDP/DTCD Project JOR/97/O03, Hashe-
Sultanate of O m a n (Jones et al. 1988), Saudi A r a b i a mite Kingdom of Jordan Water Authority.
(Bakiewicz et al. 1982; L l o y d & Pim 1990) and Syria DOORNKAMP, J. C., BRUNSDEN,D., JONES, D. K. C. & COOKE,
( K h o u r i 1982). T h e r e are also studies in other parts o f R. U. 1980. Geology, Geomorphology and Pedology of
J o r d a n , particularly in the south of the c o u n t r y ( B u r d o n Bahrain. Geo Books, Norwich.
1982; C h a r a l a m b o u s 1990). DOYLE, W. W., AUBEL, J. W., DAVISON, W. D., GRAF, C. G.,
JONES, J. R. & KENNEDY, K. G. 1984. The Hydrology of
Acknowledgements. The work presented here was undertaken the Sultanate of Oman. Public Authority for Water
as part of the Jordan Badia Research and Development Resources Report No. 83.1, Muscat.
Programme, jointly sponsored by the Royal Geographical IBRAHIM, K. M. 1992. The geologicalframework for the Harrat
Society (with the Institute of British Geographers), London, Ash-Shaam basaltic super-group and its volcanotectonic
and the Higher Council for Science and Technology, Amman. evolution. Natural Resources Authority, Amman.
Support was generously provided by the University of Durham JONES, J. R., WEIER, H. & CONSIDINE,P. R. 1988. Geology and
Research Initiatives Fund. David Drury and Beatrice Gibbs hydrogeology of the pre-dune sand deposits of the Wahiba
completed parts of the groundwater survey and their input to Sands, Sultanate of Oman. Journal of Oman Studies, 3,
the Programme is gratefully acknowledged. 61-73.
KHOURI, J. 1982. Hydrogeology of the Syrian steppe and
adjoining areas. Quarterly Journal of Engineering Geology,
15, 135-154.
References KORZON, V. I. 1974. Atlas of Worm Water Balances. Hydromet,
Moscow.
ABRAHAMS, A., PARSONS, A. & HIRSCH, P. 1992. Field and LLOYD, J. & PtM, R. H. 1990. The hydrogeology and ground-
laboratory studies of resistance to inter-rill overland flow water resources development of the Cambro-Ordivician
on semi-arid hillslopes, southern Arizona. In: PARSONS,A. sandstone aquifer in Saudi Arabia and Jordan. Journal of
& ABRAHAMS, A. (eds) Overland Flow- Hydraulics and Hydrology, 121, 1-20.
Erosion Mechanics. UCL Press, London, 1-23. MAIZELS, J. & ANDERSON, E. W. 1988. Surface water in the
AL-HOMOUD, A. S., ALLISON, R. J., SUNNA, B. F. & WHITE,K. Sharqiyah: flash floods February/March 1986. Journal of
1995. Geology, geomorphology, hydrology, groundwater Oman Studies, 3, 217-230.
and physical resources of the desertified Badia environ- PATTON, P. C. & BAKER, V. R. 1977. Geomorphic response of
ment in Jordan. GeoJournal, 37.1, 51-67. central Texas stream channels to catastrophic rainfall and
BAKIEWICZ, W., MILNE, D. M. & NOORI, M. 1982. Hydro- runoff. In: DOEHRING, D. O. (ed.) Geomorphology in Arid
geology of the Umm Er Radhuma aquifer, Saudi Arabia. Regions. London, Allen & Unwin, 189-217.
Quarterly Journal of Engineering Geology, 15, 105-126. SCHICK, A. P. 1977. A tentative sediment budget for an
BENDER, F. 1974. Geology of Jordan. Gerbruder Borntraeger, extremely arid watershed in the southern Negev. In:
Berlin. DOEHRIN6, D. O. (ed.) Arid Geomorphology. John Wiley
BENDER, F. 1975. Geology of the Arabian Peninsula: Jordan. & Sons Ltd, New York, 139-163.
United States Geological Survey Professional Paper 5604. YMR, A. 1983. Hillslope hydrology, water harvesting and areal
BRUNSDEN, D., JONES, D. & DOORNKAMP, J. 1979. Bahrain distribution of some ancient agricultural systems, northern
Surface Materials Resources Survey and its application to Negev. Oecologia, 47, 83-88.
planning. The Geographical Journal, 145, 1-35. YAIR, A. 1992. The control of headwater area on channel runoff
BURDON, D. J. 1959. Handbook of the Geology of Jordan. in a small arid watershed. In: PARSONS,A. J. & ABRAHAMS,
Government of the Hashemite Kingdom of Jordan, A. D. (eds) Overland Flow-Hydraulics and Erosion
Amman. Mechanics. UCL Press, London, 53-68.
Geological Society, London, Engineering Geology Special Publications
Mapping for high pressure gas pipelines in South Wales
G. P. Birch

Geological Society, London, Engineering Geology Special Publications 2001; v. 18; p. 73-82
doi:10.1144/GSL.ENG.2001.018.01.12

© 2001 Geological Society of London

Mapping for high pressure gas pipelines in South Wales

G. P. Birch
Consultant Engineering Geologist, Sevenoaks, Kent, UK

Objectives The Project

In order to mitigate against environmental impact, to Gas supplies to the industrial valleys of South Wales
satisfy the design criteria and to ensure the long-term developed in piecemeal fashion outwards from the
integrity of the construction, the Transmission Depart- individual coking plants set up by collieries in the valley
ment of Wales Gas sought geotechnical advice on the floors. The introduction of natural gas from the North
routing and design of a major gas transmission system. Sea in the 1970s provided the opportunity to invest
This system was being built to reinforce the gas supplies in a new transmission system emanating from the
to the industrial valleys of South Wales. The opportunity new high-pressure grid on the north edge of the Coal
was taken to apply geomorphological mapping at an Field basin.
early stage in route planning so as to avoid abortive The 'backbone' of the project is a 30 km north-south,
design work on alignments which might subsequently high-pressure, welded steel transmission/storage pipeline
prove unsuitable or too costly to engineer (Fig. 1). Pipe- (oversized up to 1200 mm diameter to permit storage by
line engineers welcomed the approach, which com- pressure-packing) linking the North Sea feeder main at
menced by obtaining an understanding of the client's Dowlais, near Methyr Tydfil, with the Cardiff feeder
objectives and industry design guidance. Working as main at Nantgarw, near Caerphilly (Fig. 2). Congestion
part of the project team, engineering geomorphologists within the intensely developed valley floor dictated a
provided guidance from initial routing studies through mountain route along the valley divide (Fig. 3), with
detailed design to construction, which was completed smaller diameter cross-valley spurs to connect into the
without contractual conflict arising out of unforeseen existing low-pressure distribution system by way of
ground conditions. decompression stations.

I~ - - ' ~ - - - - 1

GEOMORPHOLOGICAL
I N P U T H E R E RED UCES
"l PLANNINGSTAGE ~
Is corridor free from ~ -(~
!
- - 4,
,, t
ABORTIVE
T H E RISK OF A B O R T I V E
DESIGN
D E S I G N WORK I

(approx. 0.1%
of project costs)
I OESIGNS QE IU
Is route free from ~- 0
:
- - .l
WORK

Fig. 1. Application of geomorphology to the development of a pipeline project.

From: GRIFF]THS,J. S. (ed.) Land Sutface Evaluation for Engineering Practice. Geological Society, London, Engineering Geology
Special Publications, 18, 73-82. 0267-9914/01/$15.00 9 The Geological Society of London 2001.
74 G.P. BIRCH

Dynevor ArmsLNG Dowlals Initial observations at the desk study stage revealed
that the routes for both N-S and E - W pipelines were
required to negotiate a wide range of topographic, geo-
i i~Tmd~ar morphological, geological and man-made constraints as
'--v.e illustrated in the 'generic' block diagram (Fig. 4).
Of particular significance to the project design was the

Al~rdar~ i large diameter relative to pipeline wall thickness and

the high pressure to be contained up to 70 bar. The sen-
sitivity of the 'thin wall' pipeline to tensile strain and,
more especially, compressive strain required particular
, Nelson
attention with regard to route selection and potential for
ground movement.
L
Pontypridd1 ~ ~ ~ Abercarn
Techniques
Caerphilly
At the planning stage, the emphasis was on the use of
existing information, such as geological maps and stereo
air photographs, to assess proposed alignments (Fig. 5).
This 'remote sensing' was particularly valuable where
the developer was not in a position to obtain access for
J ~2 L~aend field surveys. The main objectives were first, recognition
llrlean~ ExistingLP/MPdlslrlbutlonnetwork' of surface form (morphology) and its relationship to
9 \~ Exhdlno&newHPlocalgdd geology, and second, the identification of the origin
ExistingVHPnationalgrid of the features (morphogenesis) and their stage of
m NewVHPstorage pipeline development in relation to time (Brunsden et al. 1975).
The latter, in particular, provided the key to predicting
Fig. 2. Project layout. hazardous ground conditions or processes which had the

Fig. 3. The 1200mm diameter pipeline between Dowlais and Nelson was constructed in widely varying ground
conditions ranging between soft swampland, where a flotation jacket was required, to some of the UK's strongest
rock strata, where blasting was required.
 M A P P I N G F O R H I G H P R E S S U R E GAS PIPELINES IN S O U T H WALES 75

~ z
- - ~ . t
o~_z~-i
"~
,.,~uJ Z w ! "~
z ~
o ~ m ~ . o I "~

0>~

, ~

wm I "~~ "0

o~ <-
+il + +
o
r~

~ I

zomo !

"4
76 G . P . BIRCH

G E O T E C H N I C A L I N P U T - PIPELINE PLANNING STAGE

f STUDY OF PUBLISHED IJ~
~ C ~
= B

RECONNAISSANCE ROUTE INSPECTION AND PHOTOGRAPH~

. . . . . . '~Y ,~p, .._,..__~~

. • AIR
[ PHOTO EARLY O.S. b~
INITIAL APPROACH HISTORICAL
TO NCB, IGS, DOE AND
LOCAL INDUSTRIAL E
AUTHORITIES ARCHAEOLOGICAL E

CONSIDER [ASSESS
REROUTIES
- • ..... .

..t
GEOTECHNICAL DESK STUDY WITH |
_
NATURAL
. AND MAN-MADE
....
HAZARD
_ -PLANS L_ i

G E O T E C H N I C A L I N P U T - PIPELINE DESIGN STAGE

i DETAILED ENGINEERING

..... ~ ............. ,. ......... ....... ~

SITE INVESTIGATION ~ I NSID L. lar
FOR PIPELINE ROUTE AND M J CO ER ~ MINING I~
ABOVE GROUND [] I PIPEUNE SECURITY ~ SUBSIDENCE
,,,,;,~,,,,,o,,~ |1 ON STEEP SLOPES B PREDICTIONS i
"'"-'t"'-- ~ -"'-t --=-
,,~. ~
l WAYLEAVE ~ LABORATORY TESTING PIPELINE
REINSTATEMENT ~ OF SPOIL, LANDFILL

QUANTIFICATION OF ROCK EXCAVATION

!

!~
.2;:.1
RECOMMENDATIONS ON PIPELINE t
/

AND SPECIAL BACKFILL REQUIREMENTS

TRENCH STABILITY ASSESSMENT
J
DESIGN IN HAZARD AREAS, ON STEEP
SLOPES AND AT SPECIAL CROSSINGS
,...,,.,..,,,~.-~
!
I REPORTTO k

i" PROVISIONOF GEOTECHNICAL

! D A T A FOR TENDER DOCUMENTS ~.
. . . . ~_ - -

Fig. 5. Geotechnical input to pipeline planning and design.

MAPPING FOR HIGH PRESSURE GAS PIPELINES IN SOUTH WALES 77

Fig. 6. Selected natural and man-made hazards.

78 G . P . BIRCH

Fig. 7. Extract from a 1"10 000 scale hazard map.

 MAPPING FOR HIGH PRESSURE GAS PIPELINES IN SOUTH WALES 79

Fig. 8. Subsidence collapses or 'crown holes', are common in areas of abandoned shallow mining. Their
identification by geomorphological mapping provides an essential component of the hazard map.

potential to be reactivated during the construction period Once a preferred route had been established, the get-
or subsequently during the life of the project. morphological map provided the basis for the design of
The geomorphological and man-made features were subsurface investigations necessary for pipeline design.
recognized by a methodical study of the various ele- These comprised trial pits, boreholes and geophysical
ments of the landscape within, and in the vicinity of, the techniques. By distinguishing on the map between those
proposed route corridors. Air photos, viewed stereo- areas displaying essentially uniform or benign geomor-
scopically, provided a wealth of information on slope phological characteristics and those areas where difficult
morphology, drainage patterns, former land use and or hazardous ground conditions were indicated, it was
variations in soil type and wetness, enabling the devel- possible to optimize the layout of trial pits to maximize
opment of a model of the pattern of landforms and their value and save on their overall numbers.
development processes (Fig. 6). The 1:10 000 scale hazard mapping defined a number
The features observed were plotted onto a series of of areas where realignments were recommended to en-
1:10 000 scale Ordnance Survey route maps which were sure the long-term integrity of the pipelines, for example,
then developed into a set of hazard maps by highlight- to avoid areas prone to collapse of shallow nineteenth
ing those features which had the potential to impact century mine workings (Fig. 8).
on the design, construction or operation of the project At the design stage, detailed mapping at 1 : 2500 scale
(Fig. 7). was used to update the geomorphological information
Annotated in this way, the geomorphological map which, combined with the subsurface information, facili-
provided the pipeline engineer with a visually clear, low- tated an engineering geological evaluation of the pro-
cost and rapid evaluation of the route characteristics posed route (Fig. 5). By considering each element of the
and likely ground conditions. Whilst this information project and its engineering requirements, the air-photo,
could be provided by air-photo interpretation alone, it surface-photo and mapping information was used to
was desirable to carry out 'ground proofing' by field assess the likely impact of the pipeline construction on
reconnaissance, when areas of potential concern could the terrain (Hadley 1991). This provided an indication of
be assessed before consideration was given to subsurface any special design requirements, such as pre-contract
investigations, rerouting or special design. drainage in wet areas, or heavy-wall pipe thickness in
80 G.P. BIRCH

. ; - ..1'

~,.,~,~~ /
"_. o-
o

cq

o =
I
: ,

~ q

_- 9 c -

'- 0

t, ."

/'. S
o=
 MAPPING FOR HIGH PRESSURE GAS PIPELINES IN SOUTH WALES 81

Fig. 10. Perspective sketch illustrating the permanent alteration to groundwater movement resulting from
pipeline construction.

areas liable to ground movement or collapse. Not all Whilst posing no particular problems to construction
hazard areas were avoidable and detailed geomorpho- during summer months, it became evident once the
logical mapping enabled pipelines to be routed through groundwater table rose again that the pipeline trench had
extensive zones mapped by the BGS as undifferentiated permanently altered the local groundwater regime,
landslipping (Fig. 9). focusing spring water into the backfilled trench. As a
At the construction stage, rapid response was provided consequence, the trench backfill was blown out leaving
to resolve geomorphological or geotechnical problems the pipeline suspended and extensive underdraining work
relating to ancillary works or design alterations. Visits was necessary to reinstate the pipeline safely.
during construction provided the opportunity to vali-
date the predictions made at the planning and design
stages. Other applications
The techniques for using engineering geomorphology in
developing hazard assessments for linear development
Comparisons have been applied to a wide range of engineering pro-
jects, including, inter alia, the following:
The following two examples are included to illustrate
how geomorphological mapping might have averted 9 canal routing in dynamic environments (Peru) - deter-
costly post-construction remedials. The first, in Kent, is mining viability prior to site survey (Birch 1989);
where a pipeline was constructed along a line running 9 roadway repairs on unstable land ( U K ) - prioritiza-
obliquely up a hillside with a history of instability. As a tion of maintenance resources;
result, the pipe was subjected to high stresses from the 9 railway construction audit ( S p a i n ) - identification
ground attempting to carry it off downslope. The pipe and evaluation of problem zones and their causes
was exposed for stress relief and monitoring and subse- (see Birch 2001);
quently abandoned. 9 shallow tunnelling for urban metros (Greece) - deter-
The second illustration is in Wales where the pipeline mination of pre-development topography and hydrol-
trench penetrated an impermeable clay seal of natural ogy for the assessment of tunnelling hazard (Birch
boulder clay lining the floor of the valley (Fig. 10). et al. 1998);
82 G . P . BIRCH

9 railway maintenance strategy (UK) - prioritization of (5) a basis for engineering geological evaluations; and
repair and maintenance for coastal route. (6) an indication of possibly adverse environmental or
knock-on effects relating to pipeline construction.
In the application of engineering geomorphology, the
emphasis must be on focusing the input on the engin- In essence, geomorphological surveys are an effective
eering requirements. and low-cost tool which can streamline the planning and
design of pipeline projects and other linear projects, by
the early identification of potentially adverse ground
Conclusions conditions.

An understanding of landforms and of the natural pro-

cesses which have shaped, and are continuing to shape,
the landscape is now recognized as an important element References
in project planning and design. Geomorphological sur-
veys carried out at an early stage in project development BIRCH, G. P. 1989. Applications of geomorphology to small
can avoid abortive design work, and therefore wasted hydro schemes. Quarterly Journal of Engineering Geology
money, or worse, late abandonment. 22, 231-239.
In particular, the engineering geomorphological map BIRCH, G. P. 2001. Rapid evaluation of ground conditions for
provides: the 'Ave' railway, Spain. This volume.
BIRCH, G. P., LANCE, G. A. & HEWISON, L. R. 1998. Hazard
(1) an early indication of adverse or hazardous assessment for the Athens Metro, Greece. Proceedings of
processes; the International Conference on Urban Ground Engineering,
(2) a better understanding of topographic, geological Hong Kong, November 98. Institution of Civil Engineers,
and geotechnical conditions in advance of site London.
BRUNSDEN, D., DOORNKAMP, J. C., FOOKES, P. G., JONES,
access; D. K. C. & KELLY, J. M. H. 1975. Large scale geo-
(3) a conveniently scaled map for field reconnaissance morphological mapping and highway engineering design.
and for communication of relevant information on Quarterly Journal of Engineering Geology, 8, 227-53.
environmental and archaeological aspects; HADLEY, E. 1991. Engineering a greener pipeline: a practical
(4) a basis for the design of subsurface investigations; approach. Pipes and Pipeline International, Jan.-Feb.
Geological Society, London, Engineering Geology Special Publications
Terrain measurement using automated digital photogrammetry
J. H. Chandler

Geological Society, London, Engineering Geology Special Publications 2001; v. 18; p. 13-18
doi:10.1144/GSL.ENG.2001.018.01.02

© 2001 Geological Society of London

Terrain measurement using automated digital photogrammetry

J. H. Chandler
Department of Civil and Building Engineering, Loughborough University, Leicestershire, UK

Introduction represent the geometric characteristics of land and

terrain surfaces.
Photogrammetry has traditionally provided a means of
generating three-dimensional spatial data to represent
terrain surfaces, which complements traditional ground- Digital image acquisition
based surveying methods. Although techniques such as
airborne laser scanning (Lohr 1998) and synthetic aper- Digital photogrammetric methods rely upon the use of a
ture radar (Hogg et al. 1993; Vencatasawamy et al. 1998) digital image instead of the more traditional analogue
have developed, photogrammetry remains the primary contact diapositive. The digital image comprises a large
method of generating topographic maps (Wolf 1983; array of pixels, each representing a particular colour or
Capes 1998). One important advantage of photogram- grey-scale value to form the overall image. Space-borne
metry is the flexibility of scale that allows application imaging systems use push-broom sensors to create the
to imagery acquired from ground, air and space. Indeed, digital representation. For large and medium scales,
a new generation of high (i.e. 1 m) resolution satellite digital imagery can be obtained directly using a digital
sensors (Capes 1998) is likely to further increase the camera (Koh & Edwards 1996). Although digital camera
potential applications of photogrammetry. Despite many technology provides instant and appropriate imagery,
advantages, there have been several problems with the direct digital/vertical/aerial image acquisition is rarely
application of photogrammetry using traditional meth- used by the photogrammetric community (Maas & Ker-
ods. Most significantly, there was the requirement to use sten 1997). This is partly because digital cameras remain
an expensive and complex photogrammetric stereo- expensive, but, more crucially, such cameras only gener-
plotter. This ensured that the measurement process was ate images at comparatively low resolution (e.g. the
slow and generally required the skills of an experienced Kodak DCS460 camera costs s and provides an
operator, particularly if results of the highest accuracy image of only 3000 • 2000 pixels). The emulsions used in
were to be obtained. conventional aerial films represent an extremely efficient
Rapid developments in computing hardware and soft- means of storing image data at resolutions equivalent to
ware have allowed the science of photogrammetry to 50 000 • 50 000 pixels. Currently the most cost-effective
develop rapidly during the last ten years (Gruen 1994; means of obtaining digital imagery involves a hybrid
Atkinson 1996; Greve 1996). These developments have approach (Helava 1988) in which the conventional ana-
radically eased many of the problems and limitations logue photograph is converted into digital form using
associated with traditional analogue instrumentation. some form of scanning process. Scanning options and
Use of a purely numerical or analytical solution pro- costs vary widely, but to enable simplified processing
vides flexibility, which assists in two important ways. using full-format aerial images, a purpose-built geome-
Satellite imagery, oblique aerial photography and trically stable scanner should be used, (Warner et al.
ground-based imagery can be used, in addition to the 1996). These are expensive to buy, but bureau scanning
more traditional vertical aerial perspective. Similarly, services provide a cheaper alternative if production vol-
imagery acquired using a variety of non-photogram- ume is low.
metric cameras can be considered to be of value for It should be remembered that converting a photo-
spatial measurement. The most recent advance, known graph into digital form requires significant volumes of
as digital photogrammetry, now allows part of the file storage, which can cause serious system manage-
measurement process to be fully automated. This signifi- ment issues. For example, a normal black and white
cant development ensures that photogrammetry repre- aerial photograph (dimensions 230 • 230mm) scanned
sents an even more versatile and efficient method of at 25 micron resolution (1016dpi) with 256 grey levels
deriving dense digital elevation models (DEMs) to will generate an image consisting of 9200 x 9200 pixels,

From: GRIFFITHS,J. S. (ed.) Land Surface Evaluation for Engineering Practice. Geological Society, London, Engineering Geology
Special Publications, 18, 13-18. 0267-9914/01/$15.00 9 The Geological Society of London 2001.
 o

c~

o
o

o
ol~q

o
 TERRAIN MEASUREMENT USING AUTOMATED DIGITAL PHOTOGRAMMETRY 15

occupying 81 Mb of disk space. A minimum of two Application

images is needed to carry out three-dimensional mea-
surement and if colour images are scanned then these files Photogrammetric software packages, available commer-
need to increase in size by a factor of three. The manage- cially, have been developed for application using either
ment issues raised by working with digital imagery are stereo-satellite imagery (i.e. SPOT) or traditional vertical
described more fully by Colomer & Colomina (1994). aerial photography. One comparatively routine applica-
tion of automated digital photogrammetry is illustrated
by the work of Brunsden & Chandler (1996) who acquired
a new epoch of vertical aerial photography (photo scale
Image measurement and automatic 1:4000) of the Black Ven landslide, Dorset, UK, in
DEM generation March 1995. Automated methods were used to generate
a DEM consisting of 1 000 000 points within an area of
One of the important advantages of digital photogram- 1250 • 800m, a sampling density of one point every
metry is the replacement of the complex, specialized and metre (Fig. 1). Once a DEM had been created, it was pos-
expensive photogrammetric plotter with a more general, sible to create contours (Fig. 1), cross-sections and ortho-
and hence far cheaper, digital computer. Machines photos (Fig. 2). The orthophoto is particularly valuable
running the UNIX operating system have been widely for Earth scientists because it combines the interpretative
used in the recent past, but PC-based platforms are capabilities of the original aerial photograph with the
becoming ever more powerful and capable of manip- positional relevance of a map. The 1995 DEM of the
ulating the large images required. A competent UNIX Black Ven system was used to update a sequence of five
machine can be purchased for costs as little as s a lower resolution manually measured DEMs, which rep-
PC for less. Appropriate photogrammetric software is resented the morphology of the mudslides every ten years
required but competition for this market and the influ- since 1946 (Chandler & Cooper 1989). The improved
ence of software packages designed for remote sensing spatial resolution in combination with climatic and land-
applications have reduced costs significantly (e.g. Erdas slide incidence data allowed the revision of an evolu-
Imagine/OrthoMax and OrthoBase, PCI/EASI-PACE, tionary model (Brunsden & Chandler 1996).
R-WEL/Desktop Mapping System, VirtuoZo). Prices Although application using vertical aerial photo-
vary widely, currently within a range between s graphs will always remain most important, it is signifi-
and s 000. cant that automated digital photogrammetry can be
The second and perhaps most significant advantage of applied to both ground-based and terrestrial imagery. To
digital photogrammetry is the potential to automate ensure that the automated DEM extraction software
various aspects of the measurement process. The essen- remains successful, it is necessary to introduce an extra
tial capability relevant for land surface measurement is stage in the photogrammetric processing and although
the automated measurement of digital elevation models this is not documented in software user manuals, it
from an overlapping stereo-pair, which is now both is possible (Chandler 1999). One application of this
practicable and in an advanced state of development. approach is reported by Pyle et al. (1997) and involved
Such automation is based upon sophisticated image the creation of DEMs representing riverbanks using
correlation or image matching techniques that automa- ground-based oblique photography. The objective was to
tically identify and measure common image patches compare successive DEMs in order to map the spatio-
appearing on two overlapping digital images. Once temporal pattern of bank erosion. The precision of auto-
matched, these two image measurements are transformed matically generated DEMs was + 1 2 m m and, signifi-
into object coordinates using established photogram- cantly, the locations where individual clasts had been
metric methods, and the process repeated. With appro- removed from the gravel riverbank could be identified.
priate hardware and software, this cycle can recur at The same methods have been used to quantify the three-
speeds in excess of 100 points per second and so very dimensional form of exposed and subaerial riverbed
dense and consequently accurate DEMs can be gener- gravels, both in natural riverbed gravels (Butler et al.
ated. In most packages, the derived DEM is in the form of 1998) and in a flume (Stojic et al. 1998).
a regular grid of elevation estimates draped over the
desired area. This ability to measure very dense and
regular grid DEMs has instigated a return to grid-based Limitations
methods of manipulating and presenting height informa-
tion. This will perhaps reverse the trend towards using When considering using digital photogrammetric tech-
DEM processing methods based upon the Delauney niques, it is important to consider some of the require-
triangulation (Petrie & Kennie 1990). This algorithm ments and limitations associated with the method.
efficiently creates a surface from a limited sample of Although not theoretically essential, introducing photo-
terrain elevations in which the break-line assumes great control points into the object space eases photogram-
significance. metric processing significantly and should always be
 oo

~m

o
0
O~
o
,.~

0
r4
 TERRAIN MEASUREMENT USING AUTOMATED DIGITAL PHOTOGRAMMETRY 17

considered. Photo-control points are simply recognizable Developments in software implementing digital photo-
features that appear on the photography, and are situated grammetric methods are significant and readers are
at known geographical locations. Such points provide the encouraged to consider using such packages. The soft-
means to establish the transformation from measure- ware is also 'user-friendly', which allows the inex-
ments on the image, to positions in the desired ground perienced novice user to obtain results. Despite this, it is
coordinate system. The technology available to carry out perhaps important to conclude with several caution-
photo-control surveys has advanced in recent years (Lane ary comments. First, it is necessary to become famil-
et al. 1998). For surveys of small areas using terrestrial iar with the procedures briefly outlined in this paper.
photogrammetry, the modern total station is most effec- More guidance is provided by photogrammetry text-
tive, particularly if combined with an intersection method books (Wolf 1983; Atkinson 1996; Greve 1996) and
of surveying. Over larger areas and using vertical aerial other papers (i.e. Chandler 1999). It is easy to become
imagery, the Global Positioning System (GPS) (Schofield overly ambitious in terms of the size and number of
1994) is particularly appropriate, mainly because GPS DEMs that can be generated in a project. It is prudent to
does not rely upon maintaining 'line of sight' between start with a small area and extend only when experience
points. has been gained. Finally, it must be recognized that it
Although automation afforded by digital photogram- takes considerable care and some expertise to generate
metry provides a distinct advantage, such automation truly accurate surfaces on a routine basis. It is always
can also be interpreted as a weakness. The software will essential to assess the accuracy of generated surfaces
always produce some form of surface, but this may not using independent methods.
necessarily represent the surface that the user requires.
If, for example, the parameters used to control the deriva-
tion of the D E M are inappropriate, then erroneous sur-
faces will be generated (Smith et al. 1996). In vegetated References
areas, the surface generated automatically will represent
ATKINSON, K. B. 1996. Close Range Photogrammetry and
the tops of all visible vegetation and not the underlying
Machine Vision. Whittles, Caithness.
ground surface, which the user may perhaps require. BRtrNSDEN, D. & CHANDLER,J. H. 1996. The development of
Automated D E M acquisition is possibly best suited to an episodic landform change model based on the Black
situations in which there is either a lack of vegetation or Ven mudslide 1946-95. In: ANDERSON, M. J. & BROOKS,
where vegetation is homogenous in terms of foliage S. M. (eds) Advances in Hillslope Processes. John Wiley &
height above the true ground surface. Sharp disconti- Sons, Chichester, 2, 869-898.
nuities in the land surface can also downgrade final BUTLER, J., LANE, S. N. & CHANDLER, J. H. 1998. DEM
surface representation, particularly where resolution of Quality assessment for surface roughness characterisa-
the sampling points is low. It may then become neces- tion using close-range photogrammetry. Photogrammetric
Record, 16(92), 271-291.
sary to measure the three-dimensional characteristics of
CAPES, R. 1998. Developments in earth observation - how new
such discontinuities directly, using break-lines (Petrie & space-borne sensors will affect mapping. Surveying World,
Kennie 1990). 6(6), 24-27.
Many packages provide the opportunity to assess the CHANDLER, J. H. 1999. Effective application of automated
accuracy of points measured automatically using stereo- digital photogrammetry for geomorphological research.
superimposition (Smith et al. 1996). This practice is Earth Surface Processes and Landforms, 24, 51-63.
valuable but although such packages provide the oppor- CHANDLER,J. H. & COOPER, M. A. R. 1989. The extraction of
tunity to edit and correct erroneous height estimates, positional data from historical photographs and their
this process can become impracticable if the density of application in geomorphology, Photogrammetric Record,
generated data is high. 13(73), 69-78.
COLOMER, J. L. 8~;COLOMINA,I. 1994. Digital photogrammetry
at the Institut Cartogr~ific de Catalunya. Photogrammetric
Record, 14(84), 943-956.
GREVE, C. 1996. Digital Photogrammetry: an addendum to the
Conclusion manual of photogrammetry. American Society of Photo-
grammetry and Remote Sensing, Bethesda.
Digital elevation models can now be generated auto- GRUEN, A. 1994. Digital close-range photogrammetry- progress
matically using digital photogrammetry. Such automa- through automation. International Archives of Photogram-
tion has allowed the density of data used to represent metry and Remote Sensing, 5, 122-135.
terrain morphology to be increased by a factor of over HELAVA,U. V. 1988. On system concepts for digital automation.
International Archives of Photogrammetry and Remote
100 compared to manual photogrammetric measure- Sensing, 27/2, Kyoto, 171-190.
ment methods. This development is of distinct value to HOGG, J., MCCORMACK, J. E., ROBERTS, S. m. & GAHEGAN,
terrain measurement because it replaces the tedious and CHANDLER, J. H.M.N. 1993. Automated derivation of
expensive traditional manual procedures and generates a stream channel networks and selected catchment charac-
far higher resolution DEM. teristics from digital elevation models. In: P. M. MATTER
18 J . H . CHANDLER

(ed.) Geographical Information Handling- Research and PYLE, C. J., RICHARDS,K. S. & CHANDLER,J. H. 1997. Digital
Applications. John Wiley & Sons, Chichester, 207-235. photogrammetric monitoring of river bank erosion. Photo-
KOH, A. & EDWARDS, E. 1996. Integrating GPS data with Fly- grammetric Record, 15(89), 753-763.
on-demand digital imagery for coastal zone management. SCHOFIELD, W. 1994. Engineering Surveying. Butterworth-
Association of Geographic Information 1996 Conference Heinemann, Oxford.
Proceedings, 6.1.1-6.1.5. SMITH, M. J., SMITH, D. G. & WALDRAM, D. A. 1996.
LANE, S. N., CHANDLER, J. H. & RICHARDS, K. S. 1998. Experiences with analytical and digital stereoplotters.
Landform monitoring, modelling and analysis: landform Photogrammetric Record, 15(88), 519-526.
in geomorphological research. In: LANE, S. N., RICHARDS, STOJIC, M., CHANDLER, J. H., ASHMORE, P. & LUCE, J. 1998.
K. S. & CHANDLER, J. H. (eds) Landform Monitoring, The assessment of sediment transport rates by automated
Modelling and Analysis. John Wiley & Sons, Chichester, digital photogrammetry. Photogrammetric Engineering and
1-17. Remote Sensing, 645, 387-395.
LOHR, U. 1998. Digital elevation models by laser scanning, VENCATASAWAMY,C. P-, CLARK, C. D. & MARTIN, R. J. 1998.
Photogrammetric Record, 16(91), 105-109. Landform and lineament mapping using radar remote
MAAS, H.-G. & KERSTEN, T. 1997. Aero-triangulation and sensing. In: LANE, S. N., RICHARDS, K. S. & CHANDLER,
DEM/Orthophoto generation from high-resolution still- J. H. (eds) Landform Monitoring, Modelling and Analysis.
video imagery. On the potential of digital cameras onboard John Wiley & Sons, Chichester, 165-194.
an aircraft. Photogrammetric Engineering and Remote WARNER, W. S., GRAHAM, R. W. & READ, R. E. 1996. Small
Sensing, 63(9), 1079-1084. Format Aerial Photography. Whittles, Caithness.
PETRIE, G. & KENNIE, T. J. M. (eds). 1990. Terrain Modelling in WOLF, P. R. 1983. Elements of Photogrammetry. McGraw Hill,
Surveying and Civil Engineering. Whittles, Caithness. Singapore.
Geological Society, London, Engineering Geology Special Publications
Rapid evaluation of ground conditions for the AVE railway, Spain
G. P. Birch

Geological Society, London, Engineering Geology Special Publications 2001; v. 18; p. 83-89
doi:10.1144/GSL.ENG.2001.018.01.13

© 2001 Geological Society of London

Rapid evaluation of ground conditions for the AVE railway, Spain

G. P. Birch
Consultant Engineering Geologist, Sevenoaks, Kent, UK

Objectives Following the landslip, and a number of other very

worrying incidents, the Inspector General expressed
The development of high-speed railways across Europe, concern for the quality of construction and integrity of
spurred on by the commencement of the Channel Tunnel the earthworks, including in particular:
in 1986, saw the development of Spain's first Euro-
pean gauge high-speed line, the Alta Velocidad Espana 9 the absence of a project-wide geological study and site
(AVE), between Madrid and Sevilla, southern Spain, investigation;
aided by European Economic Community funding 9 settlements at embankment/structure transitions;
(Fig. 1). The construction timetable required completion 9 erosion and minor failures on embankments;
by April 1992, in time for inauguration by King Juan 9 instability in cuttings;
Carlos at the opening of Expo'92 in Sevilla. 9 adequacy of cross-track drainage and effect on adja-
Whilst the greater part of the 470km new align- cent railway;
ment from Madrid was advancing well, the final 9 adequacy of design of bridges and retaining walls;
125 km between Cordoba and Sevilla had to be squeezed 9 adequacy of quality control.
into a 289 year design and construct period. This fol- The client's perception, therefore, was that the entire
lowed a late decision to switch from an upgrade of the 125 km between Cordoba and Sevilla should be subject
existing route to a new alignment along the valley of to a complete geotechnical appraisal comprising a cam-
the River Guadalquivir, one of Spain's five major rivers paign of regularly spaced boreholes along the railway
(Fig. 2). formation.
A major landslip occurred during construction on the
Cordoba-Sevilla section destroying both the new works
and adjacent rural railway (Fig. 3). This prompted the
Railway Inspectorate for RENFE, the Spanish National Regional setting
Rail Network, to commission an independent technical
audit to address, in particular, the geotechnical aspects The route traverses the gently undulating flanks of
of the construction works. the Rio Guadalquivir immediately south of the Sierra
Morena which form the southern edge of the Span-
ish Mesata.
The strata encountered comprise largely loose or
partly consolidated gravels, sands, silts and soft clay of
.:i iii!ii:i:::: Quaternary age which have accumulated by the gradual
infilling of the Guadalquivir Depression (Vanney 1971).
Older basement rocks of Cambrian Age are encountered
in the region of Almodovar del Rio where the railway
tunnels (300m) beneath a castle built atop a granite
intrusion. A further outcrop of basement rocks is
exposed to the west in a deep cutting.
Between the alluvial gravels and the basement rocks
are contrasting blue-grey marly deposits laid down in a
quiet lacustrine regime which prevailed during Turto-
neien Times. These marls, known as 'margas azules', are
characterized by a high proportion of the swelling
clay minerals of the smectite group and this character-
istic has a very significant bearing on its geotechni-
Fig. 1. Location map. cal behaviour (Gonzalez & Galan 1986).

From: GRIFFITHS,J. S. (ed.) Land Surface Evaluationfor Engineering Practice. Geological Society, London, Engineering Geology
Special Publications, 18, 83-89. 0267-9914/01/$15.00 9 The Geological Society of London 2001.
 n~

oa
o
++ o
0 o

~,,-i, +i +
o a ,.-i
,.,,.,

B
l
| ~0
<
I=1

! N
B

0~
bC: o
o

o
O --

o o

Z)
0
.-i
LH

+,
t n~

--z

~B
|
| "~++,,I
o

! "-...J ,.Q
O
2
o

o
m
, ,.,.~
tll
r~
O ~ w
{g

+
W _
+
i +i! e4
O ..~
 RAPID EVALUATION OF GROUND CONDITIONS FOR THE AVE RAILWAY, SPAIN 85

Fig. 3. Landslide on the Cordoba-Sevilla section of the railway that occurred during construction.

Techniques 9 stereo aerial photographs at 1:18000 scale from

national coverage;
Following a rapid one-date route reconnaissance we 9 stereo aerial and oblique photographs taken for the
were able to persuade the client that a more cost-effec- project;
tive and beneficial solution would be a phased approach 9 meetings with representatives of the contractors.
in order to maximize the use of available information
before commitment to expensive and inappropriate sub- A major component of the desk study was the
surface investigations. The phased approach also pro- characterization of geotechnicai and geomorphological
vided the opportunity for feedback to the client at key conditions through air photo interpretation and field
stages of the programme. reconnaissance to provide 'ground truthing'. This phase
provided an early indication of the likely extent of
additional ground investigations necessary to confirm or
re-establish the geological conditions.
Phase One: desk studies and field The field reconnaissance commenced with a two-day
reconnaissance inspection (level 1 reconnaissance) using track-side
access roads, which are a feature of the project, wherever
The objective at this stage was to establish the possible.
geological/geomorphological setting to the project and Having confirmed the limited opportunity for geo-
the scope and quality of the geotechnical data gen- technical problems over those stretches of track at grade
erated during the design stage of each section of the and on level ground, attention was focused for the
works. remainder of the field reconnaissance period on a more
Sources of information included: detailed assessment on foot (level 2 reconnaissance).
This survey covered the section of route between
9 published topographic and orthophoto maps at Almodovar del Rio in the east, and the Rio Guadalqui-
1:25 000 scale (14 photomosaics); vir crossing in the west.
9 published geological maps at 1:50,000 scale and Field record sheets designed to capture rapidly a wide
associated memoirs (six sheets); range of factors pertinent to the geotechnical conditions,
9 published technical papers from The Geological were completed for the entire route as well as for each
Society, Imperial College and University of Sevilla major structure, including those falling within sections
(13 references); of track at grade (Fig. 4). The locations of each of the
9 unpublished technical reports and site investi- 148 field record sheets were indicated on the 1:50000
gation data from client and contractors (105 docu- scale route maps. Photographs were taken to illustrate
ments); specific features identified in the fieldwork.
86 G.P. BIRCH

e =oe?o~,~" , ~ ,,., ma.."tm-~-rm'i .m s I

m•4L• SEV~LLA A ~ I I
IIEP.
" _ _ _ _
A
REHFE S E V I L L A A~m
FIELD RECORD S H E E T
"I 2 : : ; - ~ "
I 1 ~ ' ~ 1 ~ ' ~ 1
. . . . . I
HO,MI D E R~FERENCMI DEL C.,4MPOI - r..Url~Jux~:.~ ~1, HO~4 D E REF'ERENC/A DEL C~ I ~ I ~ ~ CONSt~I~R~ .~

+ . M ~ a r s mm~ s ~
"++'-'-*" "-I i= s I ,~,-,+ +-,-o s.,,.z ~ <~o1"-~
,+- e.~. "s~',s~az (s-sO
U T I m l O~ I m O i l X ,

| ,,, ,.,,~,. )
!-~ ,,,++
(~121t~,',r~l)h, O~[T~S i . - + + 1 + 1 S. t+om ~ ~q+lt.11 + , e + M ~&l.1~mi~*'rej
r ,~t,~.-rc ,N. 4rr
s[vznlxy ol, ~,om.[ll, o z ~ i 'x z

..... +- t__~: i v,J~It4EA~

tmm.~0 ,
9 _i
u m,~umt

= . ,

~ LrOLeeu ~ O eem'[mmICAL meomu, m w imzlr, i.o,P.~'rlom I x r l o l [ i

+x.-
Ix ,
-:-, < A'

I.~. _.p 9 r
~ I M ~" <'
_- _'_
+ - - |

tl n to

,M~,"'. mm ~
P~.+ + ...o"" ~ ' "+/,.

A e CORDOBA .. I,,E~ ~ , - " * - " ~ 1 ,.,-"~- = ~ ,

RENFE SEVILLA
F I E L D RECORD S H E E T
--"] 2 - 3 , ~ - 3 ~ -
~ V l l ~ / m ~
,~
.E+E sEv.,,. A~m I [-~-'-~-c-~,__ , _, I 1
FIEI-D RECORD SHEET l ~ / ~ a ~ r 2 J ~ l m ~ |
~/OJA D E REFEREA~.JA DEL CAMPO B K ~ l ~ & b f J ~ l CONSULTORE$ -cA HOdA D E R E F E ~ A DEL C ~ . - ~ . ~ m ~ l e ' 1 1 1 1 l l ~ J CON~r~POI~'S SA.J

"'"+"m-I ~ I'=-" I :-~- ~"~-~" I.==x- IN.4N, N

/

~ ,
( ),+.-,+.++, [, 1..+,++..-,_.. z/ , -
<~-~m cscm+~ e,,., ~m+o r om e u . m l

wy ,m," M,

.wwrs'r~,m~v,~ ~ ~e-ooc~,v'.ms ,re

[xm~,
e . ~ e , vres

eN~sAem~.
</ j,,~iv, ~ c,,js.rR,,~,,,~

+O(T+
,.~-,..~..~.~ ~1.,~,,~..1'~..~. l ~ - IICT~II.
.............. I I," I mE~ ~.A 'm~ a
~'CtO+V

~[TCH, OUCXII,Tm~,

4"

o,m,.:C! :

lll~tl~ r~. elf ~ 93 enAm ram, _ ! A

Fig. 4. Field record sheets.

 RAPID EVALUATION OF GROUND CONDITIONS FOR THE AVE RAILWAY, SPAIN 87

Fig. 5. Generic model of the landslide failures in the marl.

Of the more serious problems noted by the field Following discussions with the client, a limited
reconnaissance, many involved the notoriously weak campaign of subsurface investigations was carried out
marls of the Guadalquivir Depression. The presence comprising 11 carefully targeted rotary cored bore-
and indeed the significance of these marls had evidently holes at four of the eight locations of residual problems.
not been recognized in the design of cuttings, and in In addition, general guidelines were given on the resolu-
some cases possibly foundations. The anticipation at tion of project-wide problems of a more generic nature
outline design of cut slope of l h to lv (45 ~ would relating to erosion and washout affecting both cut-
appear to have been over-ambitious in the light of con- tings and embankments and, more especially, bridge
struction experience. For the more serious problems, abutments.
specific recommendations were made as to what further
investigations were necessary in order to develop
remedial designs.
The findings of Phase One were presented as an
interim report accompanied by the field record sheets
Results
and 1:50 000 scale route maps highlighting those areas The absence of a desk study for the 125km route had
requiring special attention. left a gap in the design process such that inappropriate
parameters were being adopted. This project-wide over-
view of the geological and geomorphological develop-
ment of the region provided a basis for understanding of
Phase Two: field investigations the engineering difficulties experienced by individual
contracts and the underlying cause of the major land-
The primary function of Phase Two was to elucidate and slip processes. For example, geomorphological mapping
re-evaluate ground conditions at eight locations where from 1:18 000 scale stereo aerial photographs and pub-
specific geotechnical problems were indicated. The two lished 1 : 25 000 photomosaic maps revealed the location
most severe problem locations shared the same generic where the alignment traversed former stream courses
cause, namely triggering of incipient failures within the and abandoned meanders of the Rio Guadalquivir
troublesome marls (Fig. 5). (Fig. 6). It was, therefore, no coincidence to find severe
v~

~o

O
0

oo
 RAPID EVALUATION OF GROUND CONDITIONS FOR THE AVE RAILWAY, SPAIN 89

embankment settlements adjacent to piled bridge struc- References

tures at these locations.
The overall conclusion was that the client welcomed
GONZALEZ, I. & GALAN, E. 1986. Origin and environmental
a fresh, rapid and objective evaluation free from politi- conditions of the Tertiary marine deposits of the Sevilla
cal, contractual or technical preconceptions. The end- area, Guadalquivir basin, Spain. Tenth Conference on Clay
product was a survey of the land surface that would be Mineralogy and Petrology, Ostrara, 209-217.
directly accessed for the design of any future remedial VANNEY, D. L. 1971. Le Bas Guadalquivir. Ru Vase de
works. Velazqiez.
Geological Society, London, Engineering Geology Special Publications
Preliminary landslide hazard assessment in remote areas
J. H. Charman

Geological Society, London, Engineering Geology Special Publications 2001; v. 18; p. 91-95
doi:10.1144/GSL.ENG.2001.018.01.14

© 2001 Geological Society of London

Preliminary landslide hazard assessment in remote areas

J. H. Charman
Consultant Engineering Geologist, Milford, Guildford, Surrey, U K

Rationale of the model can be found in Fookes et al. (1985). Most

hill canals in the study area run through Zones 3, 4 and 5.
In remote areas decisions on the need for remedial and
maintenance works on linear projects such as roads
and canals may involve the evaluation of risk based on The technical problems
the identification and distribution of definable hazards.
A terrain model provides the basis for a classification Hill canals require an intake on a tributary channel
scheme to enable this to be done cost effectively. In addi- above the main valley. They must run to a predeter-
tion it facilitates the identification of problem areas that mined gradient to the irrigation area and thus inevitably
may require more detailed subsequent design investiga- cross the steep slopes of Zone 4 before emerging onto
tion. In the absence of more detailed investigation it pro- gentler Zone 3 slopes where the agricultural land is
vides the basis for a conceptual construction approach. located. The major problem for these canals is to iden-
tify the existing landslides and to use methods of design
and construction that do not exacerbate these, minimize
The site new instability elsewhere and are locally sustainable.

This example relates to hill irrigation schemes in the

Himalaya of Nepal and Bhutan and involved the up- Techniques used
grading of existing village canals irrigating between 50
and 200 hectares of agricultural land for rice production. Aerial photographs at a nominal scale of 1 : 25 000 were
Each canal ran across steep slopes and was experiencing available in stereo pairs for interpretation, but the ex-
difficulties with slope instability. In the feasibility stage treme relief sometimes limited stereoscopic viewing.
a limited number of projects had to be selected from a These were used to carry out an initial land classifica-
large initial listing. tion of each project area into mountain zones or units
(Fookes et al. 1985). When possible, the units were
further subdivided into land elements that were visually
Area model recognizable as a distinct surface type (Fig. 1). These
units also represented a distinct approach to appropriate
The Himalaya form part of an active fold mountain belt design and construction to sustain the life of the canal.
characterized by moderate to high rates of uplift, regular Other factors important to the occurrence of landslides
seismic events and extreme relief. They are also subject and erosion were identified. These included: climate, par-
to a tropical monsoon climate, at least at lower eleva- ticularly rainfall and temperature, both of which varied
tions, which causes significant weathering and erosion. significantly with altitude; geology, particularly soil and
Mountain-building activity is not uniform but com- rock type and rock structure; land use; and the location
prises episodes of uplift interposed between quieter of other development projects in the area. Existing infor-
periods. In active times rivers rapidly downcut and form mation on these factors was gathered from local sources.
steep incised valleys. In quieter periods weathering and Selected aerial photos were enlarged to provide 1: 5000
slope evolution and more mature rivers result in a less scale base maps for walk-over survey. This allowed
severe landscape. The model in Figure 1 uses this con- confirmation of the land elements and the addition of
cept to classify the landscape. other elements that were not distinguished in the initial
The active lower slopes of Zone 4 represent the cur- photo interpretation. During the survey the basic geol-
rent phase of slope steepening caused by stream incision, ogy and landforms were mapped, the land use was
while the overlying more gentle slopes of Zone 3 mark differentiated, and surface and groundwater conditions
an earlier break in tectonic activity. Zone 5 is the area were noted. Areas of existing instability were demar-
of deposition where alluvial terraces of varying age line cated and the probable mechanism of failure noted
the lower valley slopes and floor. A full description (Varnes 1978).

From: GRIFFITHS, J. S. (ed.) Land Surface Evaluation for Engineering Practice. Geological Society, London, Engineering Geology
Special Publications, 18, 91-95. 0267-9914/01/$15.00 9 The Geological Society of London 2001.
92 J.H. CHARMAN

No Deecriptlon D ec.criplJon
1 High altitude gacial and perigtacial areas subject"to glacial erosion,
mechanical weathering, rock and snow instability and solifluction
movements with tttn roeW soil, boulder fields, gfac/em, bare rock :
slopes, talus development and debris fails
Free rock face and associated steep debris slopes subject to
chemical and mechanical weathering, mass movement, talus creep,
freeze-thaw, and debas fan accumulation.
Degraded middle slopes and ancient valley floors forming shallow 3A Ancient erosional terraces covered with a weathered residual soil mantle generally
erosional su,fecss subject to chemical weathering, soil creep, upto 3mthick. Slope anglle generally
sheettlow, rill and gully development and stream incision., < 35~ and elable. Often farmer terraced. Plighl,t, susceptible to water erosion
313 Degraded colluvium coropdalng landslide debris of gravel, cobbies and boulders in a
matrix of silt and clay. Slope angle
< 35". Relatively stable. Often farmer terraced. Vadeble permeability
4 Steep acBve Jower slopes with chemical and mechanical 4A Bare rock slopes. Sleep slope angles > 60~. Stability dependent on orientation ef
weathenng, large-scale mass movement, gullying, undercutting at di=morCdnuitfa~ such as joints and bedding plating.
base and accumulation of debris fans and flows of marginal stability
413 Rock slopes with nlantta of residual soil esuslly < 2rn thick. Steep slope angles
> 45*. Prone to extensive shallow debris slides. Deeper instability as f~" 4A.
4(3 Active colluvium. Thick landslide debris often at base of sJope and subject to active
dyer eresion. Slope angle > 35~ Highly unstable, paltJcularty dunngwet season.
413 Degraded colluvium Thick landslide debris. Slope angle < 35~ Marginally stable
and suecept!ble to gl'Ad,=~!downsio~..,creep during wet ~,=,~,~m
Valley floors associated with fast flowing, sediment laden rivers, and 5A of old alluvial tenaces above present fiver level. Generally fiat to shallow, <
populated by sequences of river terraces. T~)~ Coarse granular and perrneabta soils May be covered by a less permeable
residual sollmantle.
5B Front scarp face of old alltNial terraces. Sleep slope angle > 650, but subject to
sudden collapse when cemer~ation breaks down under weathering o~ when subject
to toe erosion.

Fig. 1. A mountain classification (after Fookes et al. 1985).

Local ground models Element 4B represents the undisturbed slope compris-

ing a bedrock thinly covered by a weathering mantle
Completion of the desk study and field survey provided of residual soil. This has a high potential for debris
the necessary data to enable a landslide hazard assess- slides and scores 4. Once failure has occurred the
ment to be compiled. This was based on a pro forma relatively fresh rock of the back scarp (Element 4A)
which introduced a scoring rating for each of the iden- provides a firm foundation for the canal and scores 2.
tified causative factors (Fig. 2). A full description is The slipped debris either continues to be seasonally
provided in U N D P / I L O (1993). active (Element 4C) scoring 4 or reaches a stable angle of
repose (Element 4D) scoring 3. Zone 5 is divided into the
Terrain fiat terrace surfaces (Element 5A) scoring 1 and the steep
The basic terrain model was enhanced in this study to river-cut faces prone to sudden slumps (Element 5B)
enable individual elements to be identified. This was im- scoring 4.
portant to provide a basis for allocating a relative hazard
score. Each element was identified from field observa-
Geology
tions and defined the conceptual engineering approach.
Zone 3 is comparatively stable and scores 1. The steep The scoring system is applied to rock type on the basis
slopes of Zone 4 can be divided into several elements. of its relative resistance to weathering. For example,
 PRELIMINARY LANDSLIDE HAZARD ASSESSMENT IN REMOTE AREAS 93

PROJECT: l~t4~1~4 Completedby: ~

SheetNo: / ~ ~ Date; I I - Z - ~)Z-

. . . . . . . . . . . . . . . . .
CHAINAGE
FACTOR
SCORE O-ISO I~1"~$o "~/~'{-~$o ~t'~ZO0 IZBI'I~o ,IZEl'tJoe
,, ii ii

TERRAIN iLand Element 3 llilm

mm
......

!CLASS'N !Land Element 4A

, ...... , ....
z L
,Land Element 4B , ,
mm
Land Element 4(; . . . . . . . .
4
Land Element 4D
. . . . . . . . . . . . . . 3 3
Land Element 5A 1
Land Element 5B 4
i | , ,
1 ,,

GEOLOGY 1 Quartzite Marble 1

Rock Type Gneiss, Sandstone ,, ,
2 2
Limestone
. . . . . . . . . . . 3
,Phyllite .... 4
Mica Schist 4
tt, ,,,

GEOLOGY 2 Coarse Granular (gravel) I

Soil Type Fine Granular (sand,slit) 3 3 3 , r ,

Cohesive Iclay) 2
GEOLOGY 3 Dip out of,slope 4 q.
Structure
......
Dip into slope 2 Z
CLIMATE Sub-alpine ~ ) I
C,,ooltemperate I2(X]O-30(X~n) 2 2. "
?. I
z '
7. L z,
Warm temperate (12tx>20(X)m) 3
Sub-tropical (O-1200rn) .......... 4
LAND USE Dense forest .......
I
ScrubtQmss, ,, 2 l z 1.
Dry cultivation (khet) 2
Wet cultivation {Paddy) 4 q.
,,,,f,,,
Fallow ,,,,, ,,,
3
GROUND Dry,,, 1 'i ............ ! .... t
WATER S~ge 2 2.
Moderate flow
. . . . . . . . . . . . .
3
Heaw flow . . . . . . 4

HAZARD RATING // // (07) t3 /Z

Fig. 2. Terrain hazard assessment pro forma.

94 J. H. CHARMAN

Section

-%
Section 1. Rock s t r u c t u r e dips into
mountain
no risk of sliding
slopes > 75 ~ : half tunneling ",..\

slopes < 75 ~ : open bench and

masonry wall

Rock faces with ~!

2. Rock s t r u c t u r e dips o u t o f high risk of
mountain failure
Section - ; < 15 ~ :low risk of sliding
slopes >75 ~ : half tunneling
slopes <75 ~ : open bench and
masonry wall 1. Full t u n n e l with holes for excavation
-; 15-25 ~ : moderate risk of sliding: debris and light and air at suitable
full tunneling intervals
-; > 25 ~ : high risk of sliding:
full tunneling with
permanent Section
shuttering on the
up-dip side

Section
| I I
[ i1~11 i i iJ
I . . . . "~'" i,,1 ' ii
3. Rock s t r u c t u r e d i p s more
Rock face
steeply than rock face
> 75~
- no risk of sliding, and low
- some risk of toppling risk of failure
open bench and masonry wall with
covered canal in areas of toppling
2.Half t u n n e l with op'donal suppordng walls
and outer masonry wall

Section
4. Rock s t r u c t u r e dips steeply
into m o u n t a i n
- no risk of sliding,
- high risk of toppling :
••ection ~...<
~ . ~
~V--
Ensure no over
hanging slabs

Final cut
< 75~

~ \ .. ~
Mas~
slopes > 75 ~ : full tunneling
with permanent
shuttering on
up-dip side ~ Spread rubble
slopes < 75 = : open bench and Rock faces < ~ spoil as evenl~
masonry wall with 75 = but some ~ as possible on
covered canal in risk of failure ~ l o w e r slope
areas of toppling
3. Open b e n c h with outer masonry wall

Fig. 3. Alternative engineering approaches to design and construction in land element 4A.

a resistant quartzite scores 1 while a mica schist, highly Climate

susceptible to weathering, scores 4. For soil type a
coarse granular soil scores 1 while a weaker cohesive soil Climate is heavily influenced by elevation and changes
scores 2. A structure in which the main discontinuity dramatically over short distances. A tropical monsoon
dips out of the slope and provides a failure plane scores climate combining high temperature and high seasonal
4 and a dip into the slope scores 2. rainfall is the most destructive and induces the highest
 PRELIMINARY LANDSLIDE HAZARD ASSESSMENT IN REMOTE AREAS 95

rate of weathering: this scores 4. Conversely, a cool The hazard assessment described here could be car-
temperate climate at elevations of between 2000 m and fled out by local engineering staff under the guidance/
3000 m scores 2. supervision of a qualified engineering geomorphologist/
geologist. The method has not been mathematically
L a n d use tested and is open to improvement but it has been
employed and field tested with some success to classify
In an area where farmers strive to terrace the slopes and road and irrigation canal sections in the Himalaya
irrigate them profusely to generate rice paddy, land use (UNDP/ILO 1993).
has a significant effect on stability. Disturbance of this The subdivision of the general terrain model into
vegetation cover removes the binding effect of roots, the elements provides the basis for an appropriate engineer-
protection against rainsplash and rill erosion and allows ing approach to remediation and design. For example,
greater infiltration of water. Excess irrigation water is land element 4A is indicative of relatively fresh rock in
often poorly managed. Undisturbed dense forest scores 1 the back scarp. It is therefore a potential source of rock
while wet paddy cultivation scores 4. for stone and aggregate, but may be difficult to excavate.
The approach to canal construction in this terrain
Groundwater element is illustrated by the conceptual construction
outline given in Figure 3.
The level of groundwater in the slope and the excess
The level of engineering achievement attained on the
pore pressure that this implies directly influences poten-
basis of an approach such as this is dependent on super-
tial slope instability. Qualitative assessment allows dif-
vision by an experienced and practical construction pro-
ferentiation between saturated conditions which score 4
fessional, the ability to adapt to conditions as they are
through to dry conditions which score 1.
revealed, and the use of local materials and skills which
allow sustainable maintenance. Further reading on gen-
eral design and construction measures is contained in
Use of the ground models Fookes et al. (1985). Some examples of successful imple-
mentation are given in U N D P / I L O (1993).
In remote areas where access is limited, detailed ground
investigation may be impossible. Decisions on project
feasibility and the formulation of the detailed design
may have to be made on the basis of desk study and References
walk- or drive-over survey alone. Construction methods
are devised for implementation by a local workforce and FOOKES,P. G., SWEENEY,M., MANBY, C. N. D. & MARTIN, R. P.
local materials are utilized. 1985. Geological and geotechnical engineering aspects of
In these circumstances terrain classification is a par- low-cost roads in mountainous terrain. Engineering Geol-
ticularly important part of the decision-making process. ogy, 21, 1-152.
UNDP/ILO. 1993. A manual for environmental protection
It provides a semi-quantitative and rapid method of
measures for hill irrigation schemes in Nepal. Nepal SPWP
appraisal, and identifies the potential problem areas for Manual No. 1 International Labour Organisation, Geneva.
detailed geomorphological and engineering geological VARNES,D. J. 1978. Slope movement and types and processes.
mapping. This allows conceptual alternative designs to In: SCHUSTER, R. L. & KRIZEK, R. J. (eds) Landslides:
be developed so that they can be incorporated when Analysis and Control. Transport Research Board, National
construction reveals the actual site conditions. Research Council, USA, Special Report 176, Ch. 2.
Geological Society, London, Engineering Geology Special Publications
Subsidence hazard in Berkshire in areas underlain by chalk karst
C. N. Edmonds

Geological Society, London, Engineering Geology Special Publications 2001; v. 18; p. 97-106
doi:10.1144/GSL.ENG.2001.018.01.15

© 2001 Geological Society of London

Subsidence hazard in Berkshire in areas underlain by chalk karst

C. N. Edmonds
Peter Brett Associates, Reading, Berkshire, UK

Purpose of survey Locally the Reading Formation comprises mostly clays

in the upper portion of the stratum overlying mostly
During the last ten years a number of ground subsidence sands in the lower portion. The higher ground is capped
events have occurred in the northwest part of Reading. by Hill Gravel, a sandy gravel with some silt and clay
Many of the subsidence events resulted in structural horizons. Valley Loam is present along the floor of
damage to existing properties (see Plates 1 and 2). On the Hemdean Bottom extending southwards towards the
basis of the properties inspected to date it appears that Thames where it joins with the Alluvium that occupies
the local housing has been constructed mostly upon the Thames floodplain. The Valley Loam is a calcareous
conventional strip footings bearing onto naturally occur- silty sandy clay with occasional gravel, while the
ring soils. The increasing number of recorded subsidence Alluvium consists mostly of silt with occasional organic
events is of concern to planners, developers and insurers. horizons (Blake 1903).
Consequently the aim of the survey was to identify the A simplified geological plan is shown in Figure 2. The
nature and extent of subsidence hazard in the local area. Hill Gravel is now thought to be a Post-Anglian fluvial
terrace deposit laid down by the proto-Thames when
the river bed lay at a higher topographical level (Jones
1981). The Valley Loam appears to be a brickearth/
The site head deposit.
The northwest part of Reading, generally referred to as
Caversham, is shown in Figure 1. It largely comprises a
south to southeasterly dipping land surface, overlook-
ing the River Thames. North of the Thames the land
Geomorphology
surface is dissected by a NNW-SSE trending valley Tectonic downwarping of the London Basin in the early
feature known as Hemdean Bottom. This divides the Tertiary resulted in a number of transgressions which
westerly Caversham Heights area from the easterly laid down a sequence of Palaeogene strata across the
Caversham Park and Emmer Green areas. The Thames Thames Valley area containing Reading. Continuing
lies at just below 40 m AOD and northwards the land tectonic activity through into the Neogene, and erosion,
rises to above 80 m AOD. The floor of Hemdean Bottom resulted in the stripping back of the Palaeogene cover
generally lies between 40 m and 50 m AOD. (Jones 1981). Following removal of the London Clay
Formation cover across the area and a fall in water table
level below the sub-Palaeogene surface, the potential
Geology has existed for downward infiltration of groundwater
through the Reading Formation to initiate dissolution
The published geological map at 1:10 560 scale (British of the Chalk below. On the basis of erosional evidence
Geological Survey County Series Berkshire Sheet 29 SE) (e.g. unroofing of the Weald, Jones 1981) it seems likely
for this area shows the entire district to be underlain by that these conditions have been present from late
Cretaceous Upper Chalk, overlain by a Tertiary Read- Tertiary to early Pleistocene times onwards.
ing Beds outlier to the northeast side of Hemdean During the Pleistocene the Anglian ice front advanced
Bottom. The former London Clay Formation cover has as far south as the Vale of St Albans. Reading lay
been eroded away within the study area. The chalk is a beyond the ice front and was subjected to periglacial
very weak to moderately strong, white, porous, soluble, weathering (Jones 1981). Another important Pleistocene
carbonate rock with flint bands. The chalk surface is event was the diversion of the proto-Thames river by
highly weathered where exposed and frost shattered. the Anglian ice. Both before and after impedance of the
The Reading Beds are now referred to as the Reading proto-Thames river, fluvial and fluvio-glacial deposits
Formation, part of the Palaeogene Lambeth Group. have been laid down along its ancestral and modern

From: GRIFFITHS,J. S. (ed.) Land Surface Evaluation for Engineering Practice. Geological Society, London, Engineering Geology
Special Publications, 18, 97-106. 0267-9914/01/$15.00 9 The Geological Society of London 2001.
98 C . N . EDMONDS

Plate 1. Internal view of structural cracking damage to a property at Buxton Avenue, Reading following
subsidence over a solution feature.

Plate 2. External view of compaction grouting stabilization works being carried out at the same property at
Buxton Avenue, Reading.
 SUBSIDENCE HAZARD IN BERKSHIRE IN AREAS UNDERLAIN BY CHALK KARST 99

SU69 SU72 .............

!
ADMINISTRATIVE ", HIIHffI~wn
BOUNDARY TO . . . .

\
't
c.,
\ f "" -,' E M M E R

9\
---,.2-- / GREEN

/ 1
Fan11. . . .
CAVERSHAM
HEIGHTS Gavin+sham
HHI
)
/
3

A4155

._-@

NORTH WEST
READING

td 0
metres
5OO

---,(~---- Contour (m AO0)

Fig. 1. Study area location and topography.

course (Gibbard 1977, 1983). These events were mainly along joint fractures. As a result, solution features are
focused on the Vale of St Albans to the east, but they preferentially developed at the interface of the Chalk
also influenced deposition of the Hill Gravel in the surface with the overlying cover deposit. This is dis-
Reading area. cussed by West & Dumbleton (1974), Rigby-Jones et al.
Solution feature formation results from the progres- (1993) and McDowell & Poulsom (1996), and is illus-
sive downward infiltration of groundwater (under- trated schematically in Figure 3. The range of solution
saturated with respect to calcium carbonate) into the features formed includes solution pipes, swallow holes
Chalk surface, where it dissolves the rock gradually over and sinkholes (Edmonds 1983).
a long period of time (>10000 years). The process In northwest Reading it is suggested that solution
produces solution features that penetrate downwards features have formed below the Reading Beds and Hill
100 C.N. EDMONDS

SU69 SU70 SU71 SU72

I t" * ~ "

f.J
ADMINISTRATIVE
BOUNDARY TO )
) ~READING 7

/
/
. HG/RB

HG

UCh
All M155

VL

NORTH WEST
READING
o 5OO
r
su
74 met~

KEY NOTE
Drift Geology : The geological boundaries shown are
All Niuvium } based on the 1:10,560 scale geology
VL Valley Loam (Quaternary) map, County Series Sheet SU 29 SE,
HG Hill Grovel published by the British Geological Survey
Solid Geology :
RB Reading Beds Formation (Palaeogene)
UCh Upper Chalk (Cretaceous)

Fig. 2. Study area simplified geology.

Gravel at various times in response to climatic change, date from Anglian times onwards. It is expected that
depositional and erosional events. The most active solu- within the study area, solution features will be associa-
tion feature formation phases were probably associated ted with the Chalk/Reading Formation and Chalk/Hill
with the seasonal thaw of periglacial frozen ground Gravel interfaces. Slightly acidic groundwater collecting
(Higginbottom & Fookes 1970). It is suggested that upon these cover deposits will have had ample oppor-
many of the solution features present today possibly tunity to form concentrated flows able to percolate
 SUBSIDENCE HAZARD IN BERKSHIRE IN AREAS UNDERLAIN BY CHALK KARST 101

Fig. 3. Schematic diagram to illustrate range of natural solution features present on the Chalk (after Edmonds
1987). Solution feature types: swallow holes (1), (2); erosional remnants of solution pipes (3); solution pipes (4),
(5); solution pipe with upward migrating air-filled void (6); subsidence sinkhole over a solution pipe (7); solution
cavities connecting with solution-widened joints (8); collapse sinkhole over a solution cavity (9).

downwards into the chalk surface over a long period There is a deep-cut minor valley that extends in a
of time to produce solution features. This is shown N N W - S S E orientation on the west side of the Caver-
schematically in Figure 4. sham Heights area. Further east is the N N W - S S E
trending Hemdean Bottom valley, the floor of which is
relatively narrow, less than 100m wide in many places.
Mid-way along the east side of the Hemdean valley is
Techniques used and limitations a minor tributary valley that extends northeastwards,
Broad-scale geomorphological mapping has been car- dividing into two parts at the head of the valley.
ried out within the study area to determine the terrain Another minor tributary valley extends westwards into
and surface/subsurface drainage characteristics. The the Caversham Heights area a little further north along
mapping technique was applied following the principles the Hemdean valley.
of subsidence hazard mapping developed by Edmonds The land surface profile of the interfluves consists of
(1987, 2001). The mapped boundaries of terrain units a series of stepped surfaces that lead upwards from the
have not been set out by accurate survey, but have relied valley sides to a number of local topographic 'highs'.
on topographic base maps of various scales and obser- To the west of Hemdean Bottom the stepped surfaces
vation by walkover survey. The extensive coverage of climb up to a hill brow trending ENE from Blagrave
existing development has also influenced the accuracy Farm (see Fig. 1). On the east side of Hemdean the high
of mapping. points comprise Caversham Hill and another area to the
A review of historic maps has revealed a number of north centred on Highdown Hill and Caversham Grove.
old chalk quarries, clay pits and gravel workings within The stepped surfaces generally take the form of flat to
the study area. Their locations were noted during the slightly inclined (1 to 2 ~ bench levels which are boun-
mapping survey and their influence upon the local land- ded by breaks of slope. On the uphill side is a concave
form assessed. However, none of the man-made work- break of slope, while on the downhill side is a con-
ings were particularly extensive, obliterating the natural vex break of slope. Within the study area it appears that
land profiles. Therefore man's historical impact on the generally slope angles do not exceed 5~ between suc-
local terrain is considered not to have greatly affected cessive steps; however, locally steep slope angles (5 to
the survey results. 10~ or more) are present where the interfluves finally
step down into the valley floors of the Thames and Hem-
dean Bottom.

Study area morphology

The morphology of the northwest Reading study area is The conceptual model
shown in Figure 5. The land surface comprises a series of
valleys with tributaries that drain southwards towards Ground subsidence over solution features may be trig-
the River Thames. gered in many ways, e.g. by heavy rainfall, water flows
102 C.N. EDMONDS

STAGE 1 POST TERTIARY

LC
LC - LONDONCLAY LANDSURFACEDEVELOPSUPONLONDONCLAY.
RB - READINGBEDS :::;':.;::".":,;:-"i':~::::::'!; ~'::-:,;.':~":'~;~"::::~'"~ ;'.':;..::'4":~;::;'>.'~".';~:,'.:.:-:,:::'&!::'~-:?~. ,:.
::."::..:.;:~:..!.:..-:.,~:'-':-::-::,L"~.:.';.~'::,'.~:,.' RB.:.,,':-~,?::'~.:'"~.':<.,:,~..~~...-~!-;:-.!,';,:,~WL : THICKNESSAND PERMEABUWOF LONDONCLAY
OH - CHALK ~. , ~ . . : . ~ - : ' ~ . ~ . ~ ; , . ~ 3 . ~ . . : ~ : . ~ . . , ~ . ~ . . ~ --.7...:
RESTRICTSDOWNWARDDRAINAGETO
WL- WATERLEVEL ell READINGBEDSAND CHALK.

INITIATIONOF HEMDEAN
STAGE 2 ~,:.~ BOT~'OM ~ EROSIONOF LONDON CLAY
ji ~~ ;:,?.!f:!':-i! : . ~ ?:~i:.i.'~:;:.
: ~..,,..
~ . i ~ : ~:~:.' i~~.;':~.:~.'~.:!~.'~,.;.!.~.,.~..:.~:.:'~ I~UFFACER..OWTOWAFIDIS~NG B~:~.
I'~ : . ~ : : : : ; ! : ! ; : : i i , < ' ~ ~ i ] DOWNWARDSPI:::RCOI~TIONIN READINQBEDS,
! ~ . ~ . - : . ~ . : .:=-.~ ~ . ~ ~ . ~ . - : . .
,~_.CH ~/ V ,~ ~ \J L ~ .,,~W ~ INITIA'IIONOF ~LLrrION FEATURF~IN CHALK

SINKHOLES
STAGE 3 MAYFORM SWALLOWHOLES EROSION OF READING BEDS
MAYFORM

9:" . :. - ~~ SOLUTIONFEATURE8ENLARGEBELOWCOVEFL
HEMDF.J~BOTTOMDEEPENS.

~NAI_LOWHOLES
STAGE 4 MAY FORM DEPOSITION OF HILL GRAVEL
HG - HILLGRAVEL HEldDEANBOTTOMFORMSAN ACTIVETRIBUTARYTO
ALL- ALLUVIUM PROTO-'rHAMES.FLUVIALEROSIONPROGRE,~.~IVELY
REMOVESREADINGBED6A8 WATERLEVELFALLS.
80LUTION FEATUREBAND ERO61ONALREMNANT8
REMAIN.

SINKHOLES SINKHOLES
MAYFORM MAYFORM
STAGE 5 FROST ~ POST PERIGLACIALWEATHERING
SHATTERED , ut~

IN 80LUTIONFEATLIREB...qOLUTIONFEATURES,
REACTIVATEDFOLLOWINGPERIGLACIALCONDITIONS,
BECOMEENLARGEDANDTHE HILLGRAVEL
UNDERGOES80LUTIONSUBSIDENCEINTOTHE
FEATURESWITHTIME.GROUNDSUBSIDENCE
ONGOINGTO PRESENTDAY.

Fig. 4. Schematic illustration of solution feature formation in northwest Reading.

from soakaways, garden watering, leaking drains and church (1998). The instability that leads to subsidence
others. As the deposits which infill a solution feature problems is therefore associated with the destabilization
undergo solution subsidence they become weakened, of infilling deposits within pre-existing solution features
loosened and disturbed. When destabilized by any of the and not the formation of new features. This is further
above subsidence triggers, available subsurface void illustrated by Figure 3. In the case of very large, mature
space may be rapidly transmitted upwards through the features that are interconnected with solution-widened
infill to reach the surface and form a sinkhole. This is joints in the Chalk at depth, it is also possible, by ero-
caused by breakdown of metastable voids, internal sion at the base due to flowing water, for additional
piping, erosion and wash-through of fines within the upward-migrating voids to be introduced to cause sub-
infill. These mechanisms are widely reported by West & sidence at the surface.
Dumbleton (1972), Edmonds (1988), McDowell (1989), In northwest Reading, as illustrated by Figure 4 it is
McDowell & Poulsom (1996) and Rhodes & Mary- envisaged that solution features have been formed below
 S U B S I D E N C E H A Z A R D I N B E R K S H I R E IN A R E A S U N D E R L A I N BY C H A L K K A R S T 103

SU69 SU70 SU71

SU72 1

l. r - . - -

ADMINISTRATIVE
BOUNDARY TO ) \ I

-t-
9 \

X"

Mt55

NORTH WEST
READING
0 5OO
mc4res

-+- Fiat surface

v ~ v , v-- Convex break of slope
v v, ~ ~ v Concave break of slope
-*- -+- -*- ~ --,- -,- Co~cove valley floor axis
. . . . . . Convex change of slope
. . . . . . Concave change of slope
...... Slope direction of land surface

Fig. 5. Study area morphology.

both the Reading Beds and the more recent Hill Gravel at the Hill Gravel/Chalk interface. Fluvial erosion and
cover deposits. An earlier phase of solution feature the deposition of the Hill Gravel has contributed to the
development is thought to have occurred by downward erosion of the Reading Beds (Stage 4, Fig. 4). Conse-
infiltration of water through the sub-Palaeogene sur- quently, where the sub-Palaeogene surface has been
face (Stages 2 and 3, Fig. 4). A later phase of solution destroyed, solution features will also have been removed.
feature development is considered to have taken place However, locally in northwest Reading the difference in
104 C . N . EDMONDS

Fig. 6. Subsidence hazard map.

level between the sub-Palaeogene surface and the sub- features associated with the sub-Palaeogene surface
Hill Gravel surface appears to be only about 10 m or so. probably did not occur. Hence, erosional remnants
It is not unusual for solution pipes to extend more than of solution pipes infilled with Reading Beds could
10m in depth, and 20m or so is quite common. There- have been preserved below the Hill Gravel (Stage 4,
fore, it seems that total destruction of the solution Fig. 4). In later favourable, post-periglacial climatic
 SUBSIDENCE HAZARD IN BERKSHIRE IN AREAS UNDERLAIN BY CHALK KARST 105

circumstances, permitting resumption of the downward The original research database (Edmonds 1987), which
infiltration of water into the Chalk, it is proposed that was used to develop the hazard mapping technique, com-
the erosional pipe remnants became a focus for renewed prised 2226 solution features. When each of the solution
dissolution activity (Stage 5, Fig. 4). It is believed that feature locations in the database was subjected to hazard
the cumulative effect of this solution feature formation assessment and compared with the hazard categories, it
model begins to explain why solution features are preva- was found that the number of features per hazard cate-
lent in northwest Reading and why there is a continuing gory tended to increase with ascending hazard rating
ground subsidence problem. as shown in Table 1. Within the Reading study area the
Although at this stage the database of known subsid- relationship between hazard rating and solution feature
ence events is limited, a subsidence hazard mapping exer- occurrence is also shown in Table 1 for comparison.
cise has been carried out, using the techniques explained The research database suggests that the increase
elsewhere, to attempt to better define the likely occur- in solution feature occurrence and related subsidence
rence of future subsidence events. The subsidence hazard activity tends to be associated with hazard rating of
mapping techniques are explained in detail in Edmonds 'Moderate' and above. In the Reading study area, com-
(2001). The surface morphology, subsurface geology, parison of solution features and subsidence occurrence
geomorphological development and the hydrogeology of with hazard rating reflects the general findings of the
the study area have been utilized to prepare a subsidence previous research, in locating those areas where ground
hazard map. The hazard map is shown in Figure 6. subsidence is most likely to occur. Again, the areas
of ground which are prone to solution feature occur-
rence and subsidence are mostly rated as 'Moderate'
Results of the hazard mapping and above.
In the Reading study area the relationship between
During the original development of the hazard map- absolute numbers of recorded features, subsidence activ-
ping techniques, three study areas were chosen, each ity and increasing hazard ratings does not follow quite
100 km 2 in size. Within the study areas the number of the same pattern as exhibited by the original research
solution features was greatly enhanced by the interpre- database (Table 1). Factors such as study area size,
tation of aerial photographs (viewed stereoscopically) restricted solution feature data and locational bias
and airborne multispectral scanner imagery. The larger explain this in part. In addition, the proportional spatial
numbers of features recorded increased confidence that distribution of the various hazard zones will also influ-
representative cavity occurrence spatial patterns were ence matters. It is notable that this area lies south of one
being analysed and that locational bias was minimized. of the three 100 km 2 study areas used as part of the origi-
When applying the techniques to a smaller urban study nal research, mentioned above, where a larger represen-
area, certain difficulties arise. Firstly it is often the case tative set of data more closely reflected the predicted
that the numbers of solution features and subsidence pattern of solution feature occurrence beyond the urban
problems recorded are relatively small. Secondly loca- fringe of Reading.
tional bias is evident because of the nature of the surface In addition, within the section on 'The conceptual
development. Consequently the number of features and model' above, attention was drawn to the relative
subsidence problems recorded may not, on a local area proximity of the sub-HiU Gravel and sub-Palaeogene
basis, fully reflect the predictions of the hazard mapping surfaces in the study area. It was suggested that where the
techniques which have been developed from a much gravels overlay solution pipe remnants infilled with
larger database. Reading Beds (Fig. 4), they could act as a focus for

Table 1. Subsidence hazard classification

Subsidence hazard rating Numerical range of each Natural cavity occurrence Natural cavity occurrence based
hazard rating category (%) based on research on Reading study area
database
(%) (no.)

No anticipated hazard <55 0.1 0 0

Very low hazard 55-89 0.5 3 1
Low hazard 90-136 3 Not present
Moderately low hazard 137-200 4 3
Moderate hazard 201-300 7 32 9
Moderately high hazard 301-400 12 29 8
High hazard 401-600 23 14 4
Very high hazard >600 50.4 19 5
106 C. N. EDMONDS

reactivation of solution feature development beneath the on unstable land Annex 2: Subsidence and Planning.
gravels. Consequently below the gravels, particularly Consultation Paper PPG14, HMSO, London.
within the 'Moderate Hazard' rated zones, there may be a EDMONDS, C. N. 1983. Towards the prediction of subsidence
higher than average density of solution features liable to risk upon the Chalk outcrop. Quarterly Journal of Engi-
neering Geology, 16, 261-266.
lead to subsidence because of the favourable proximity
EDMONDS, C. N. 1987. The engineering geomorphology of
effects of the two dissolution weathering interface levels. karst development and the prediction of subsidence risk
upon the Chalk outcrop in England. PhD Thesis, University
of London.
Applications and uses EDMONDS, C. N. 1988. Induced subsurface movements asso-
ciated with the presence of natural and artificial under-
The hazard mapping techniques have wide application to ground openings in areas underlain by Cretaceous Chalk.
the Chalk outcrop. They can be used to determine areas In: BELL,F. G., CULSHAW,M. G., Cv.Ipps, J. C. & LOVELL,
prone to subsidence, as illustrated by the above example, M. A. (eds) Engineering Geology of Underground Move-
ments. Geological Society, London, Engineering Geology
to assist planners, developers and insurers. Planners can
Special Publications, 5, 205-214.
then make the identification and mitigation of subsidence EDMONDS,C. N. 2001. Predicting natural cavities in Chalk. This
a condition of planning in line with PPG 14 Development volume.
on unstable land (Department of the Environment 1989) EDMONDS, C. N., GREEN, C. P. & HIGGINBOTTOM,I. E. 1987.
and the new Annex 2: Subsidence and Planning (Depart- Subsidence hazard prediction for limestone terrains, as
ment of the Environment Transport and the Regions applied to the English Cretaceous Chalk. In: CULSHAW,
2000). New-build developers and their geotechnical M. G., BELL, F. G., CRIPPS, J. C. & O'HARA, M. (eds)
consultants can be alerted to the hazards and take Planning and Engineering Geology. Geological Society,
appropriate precautions to ensure that suitably designed London, Engineering Geology Special Publications, 4,
283-293.
foundations are installed. Insurers can review the existing
GIBBARD,P. L. 1977. The Pleistocene history of the Vale of St
and future potential for subsidence associated with areas Albans. Philosophical Transactions of the Royal Society of
such as northwest Reading, and the hazard map could be London, B280, 445-483.
used as a basis for decision-making when considering GIBBARD,P. L. 1983. Slade Oak Lane. In: Quaternary Research
their insurance liability for this form of geohazard. Association Field Guide for Annual Meeting, Hoddesdon,
A further use for the technique is to understand the "The Diversion of the Thames', 85-91.
contamination susceptibility of the Chalk aquifer. Solu- HIGGINBOTTOM, 1. E. • FOOKES, P. G. 1970. Engineering
tion features can form preferential pathways for contami- aspects of periglacial features in Britain. Quarterly Journal
nants to pass rapidly down into the aquifer without of Engineering Geology, 3, 85-117.
JONES, D. K. C. 1981. Southeast and Southern England. Uni-
attenuation. Hence it may be used to evaluate and check
versity Paperbacks, Methuen.
aquifer protection zoning around water supply boreholes, McDOWELL, P. W. 1989. Ground subsidence associated with
where it is normally difficult to take account of karst doline formation in chalk areas of southern England. In:
development by conventional hydrogeological model- BECK, B. F. (ed.) Engineering and Environmental Impacts
ling. The mapping might also be used as a basis to assess of Sinkholes and Karst. Proceedings of the Third Multi-
the possibility of environmental impact of certain poten- disciplinary Conference on Sinkholes, St Petersburg Beach,
tially contaminating activities (e.g. landfills, petrol filling Florida. Balkema, Rotterdam, 247-255.
stations, engineering works, etc.) upon the Chalk aquifer. McDOWELL, P. W. & POULSOM,A. J. 1996. Ground subsidence
Environmental risk is greatly enhanced where solution related to dissolution of Chalk in Southern England.
Ground Engineering, March, 29-33.
features are present.
RHODES, S. J. & MARYCHURCH, I. M. 1998. Chalk solution
features at three sites in southeast England: their formation
and treatment. In: MAUND,J. G. & EDDLESTON,M. (eds)
References Geohazards in Engineering Geology. Geological Society,
London, Engineering Geology Special Publication No. 15,
BLAKE, J. H. 1903. The geology of the country around Reading. 277-289.
Sheet 268, Memoirs of the Geological Survey, HMSO, RIGBY-JONES, J., CLAYTON,C. R. 1. & MATTHEWS,M. C. 1993.
London. Dissolution features in the Chalk: from hazard to risk. In:
DEPARTMENT OF THE ENVIRONMENT. 1989. Planning Policy INSTITUTIONOF CIVILENGINEERS(eds) Risk and Reliability
Guidance: Development on unstable land. PPG 14, HMSO, in Ground Engineering. Thomas Telford, London, 87-99.
London. WEST, G. & DUMBLETON, M. J. 1972. Some observations on
DEPARTMENT OF THE ENVIRONMENT, TRANSPORT AND THE Swallow Holes and Mines in the Chalk. Quarterly Journal
REGIONS. 2000. Planning Policy Guidance: Development of Engineering Geology, 5, 171-177.
Geological Society, London, Engineering Geology Special Publications
Creation of functional ground models in an urban area
R. J. G. Edwards

Geological Society, London, Engineering Geology Special Publications 2001; v. 18; p. 107-113
doi:10.1144/GSL.ENG.2001.018.01.16

© 2001 Geological Society of London

Creation of functional ground models in an urban area

R. J. G. Edwards
Earth Science Partnership, Leatherhead, Surrey, U K

Rationale The glacial and pre-glacial geology is only exposed at

the western and northern city margins. This sequence,
The ability to see the ground surface in an urban with the exception of the anthropogenic fills, is mirrored
environment is limited by access and the amount of in other identifiable areas in close proximity to Cardiff.
natural ground exposure. This is typically less than 10%
by area. The assessment of the probable ground struc-
ture and the behaviour of critical ground elements can
be achieved by comparison with non-urban environ- The technical problem
ments exhibiting similar landforms. This model can be
It was proposed to construct a barrage across Cardiff
verified by a detailed study of historic maps and a review
Bay to generate an attractive water frontage as an
of existing local subsurface information.
integral part of Cardiff Bay Development Corporation's
inner city regeneration programme that included the
derelict 'Tiger Bay' area of Cardiff Docks. The impact of
this structure on the behaviour of the existing ground-
The site water regime was recognized to represent a serious
potential risk to the city's infrastructure. Also, as a result
The City of Cardiff has largely been constructed in the
of a history of uncontrolled surface flooding within the
last 150 years. The area studied relates to the whole city
city, the potential increase in groundwater level implied
land area lying to the south of the A48T and between the
by the construction of the barrage was regarded as a
west banks of the rivers Ely and Rhymney.
matter of serious public concern.
In addition, the project was required by law to be
promoted by a Parliamentary Bill necessitating that it
be forensically examined in a series of Parliamentary
Area model Committee hearings at which objectors to the scheme
The area lies substantially on the Severn Levels, which were able to present their concerns and risk assessments.
comprise a recent tidal coastal margin zone of the A rational ground model was therefore an essential
Severn estuary. The area consists of a glaciated land element in assessing the occurrence and behaviour of
groundwater both before and after construction of the
surface comprising Mercia Mudstone and Dolomitic
Conglomerate draped by a thin and partly discontin- barrage. It was also required in order to generate
qualitative and quantitative risk criteria and as a vehicle
uous layer of glacial detritus that includes two identifi-
able layers of till and a consistent layer of fluvio-glacial for critical discussion. No such model existed and little
sands and gravels which forms the dominant aquifer. or no information regarding the distribution of strata
These strata reflect Devensian and Flandrian glacial and groundwater had been published or verified except
advance and recession. for Anderson & Blundell (1965).
To the south of the historic natural shoreline a series
of post-glacial foreshore alluvia are superimposed on the
late-glacial surface and form a confining layer to the Techniques used
underlying fluvio-glacial sand and gravel aquifer. These
alluvia represent a series of more recent marine trans- Comparable ground models of the basic 'solid' and 'drift'
gressions modified by sea defence construction works of geology had been developed for adjacent areas of the
Bronze Age to late mediaeval antiquity. Both the 'solid' Severn Levels using a combination of aerial photogra-
and 'drift' geology is substantially masked by a vari- phic interpretation, geomorphological mapping, ground
able thickness of waste and engineering fills of Recent survey and the review of historical maps and other
anthropogenic origin (almost exclusively post-1800). archival data. The results of these studies were published

From: GRIFFITHS, J. S. (ed.) Land Surface Evaluation for Engineering Practice. Geological Society, London, Engineering Geology
Special Publications, 18, 107-113. 0267-9914/01/$15.00 9 The Geological Society of London 2001.
108 R . J . G . EDWARDS

11 . .,...~

!,A
~'-,q
O
I . ,...~
+..a

9 I !
 CREATION OF FUNCTIONAL GROUND MODELS IN AN URBAN AREA 109

I4W Water Table Cordtff Boy $E

River Toff Toff Chore)at
{made ground/gravel) I

Recharge
t ~ lnfilfrafion to Se. . . .
\ Water Table
( made ground)

" "~ ~ ~ )-== Piez:omet/{c Su r f Q c e

1 i_ ALLUWOM

/ "'~'-L~.._L \ / ". ~ - GRAW~.

through Alluvium '
Dischorge through
- Exposed Gr~v~l - MERCIAMUDSTONE
_~ _ _

River Ely Leokeae through Ri~v" ;rvr!

Discharge tO AI~Jv'Tt'lm Dischargeto i
Shatlow Sewer \ Deep Sewer Y
\ \ \ Wote.r Table (rnadeoround}

2 Way Flow

9 ~-..- ~, " k-- . . . . ~ N . ~ _ , , . ~ . = ~ - "e"~-,=/'' - -..- . .....,.._~,~--).''",/.."'~ ~ " "--UA~ GROmD

~ I I i ! I '~ i ~ l "" ALLUWUM

-I - I- /-
~-"-.. 9 ,-- ,, . ~. , ~ ~ " ,. . ~ i o . ~ o ~

Groundwater Flow ~\
- - ~ -- _ . _ D;scharge to River

F i g . 2. C o n c e p t u a l m o d e l .
l l0 R . J . G . EDWARDS

Fig. 3. Engineering geological plan of the study area.

or in press at the time (Allen & Rae 1987; Hawkins 1984; boundaries (Fig. 1); and second, a model to demon-
Hornby et al. 1993). strate the behaviour of groundwater in three dimensions
Using this primary model format, a search was under- (Fig. 2).
taken of all the available map and borehole record data An engineering geological plan was generated for that
for the site. A conceptual model of the City of Cardiff area of the city where the main fluvio-glacial sand and
was generated that included geological, hydrogeological gravel aquifer is confined by alluvium (Fig. 3). This
and anthropogenic data and their functional relation- illustrated the distribution of strata in three dimensions
ship to the movement and time-variant behaviour of related to their time sequence deposition as shown on
surface and subsurface water bodies. This model was the key to the plan (Fig. 4). Using this information,
then verified and enhanced by carefully planned inves- together with carefully selected ground verification data,
tigatory borehole/piezometer work. From this model a it was possible to prepare a functional computer model
robust, logical, qualitative and quantitative risk assess- of the groundwater regime and to use this to examine
ment was derived for peer review (Edwards 1997). and verify its current time-variant behaviour. This com-
puterized model was then used to examine the potential
and probable impact of the barrage on this regime. This
The local ground model was achieved by applying critical parameter ranges to a
series of sensitivity model analyses using rational time
The local model was represented as idealized two- periods representing short-term response and steady-
dimensional sections in two forms: first, a model to show state conditions. The groundwater modelling program
the general sequence of strata and their geographic used was a site-specific modified version of Modflow.
C R E A T I O N OF F U N C T I O N A L G R O U N D MODELS IN A N U R B A N A R E A 111

0'3

. ,...~

O
112 R.J.G. EDWARDS

Use of ground models results were presented as part of the parliamentary

proceedings (Tables 1 to 4). The worst case and probable
The risk generated by the calculated impact of the case conditions were assessed. A cost estimate was made
barrage was examined using the developed and refined for the potential remedial works package required. This
hydrogeological ground model. A rational risk assess- risk analysis was developed from the evaluation of these
ment based on a logical scoring system related to nine models and was presented for discussion to the various
potential identifiable risk criteria was undertaken. The Parliamentary Committee hearings.

Table 1. Summary o f effects on infrastructure

Hazard Conclusion

(i) Flooding of low ground Discounted

(ii) Dampness/wetness or flooding of basements, tunnels and Some properties at risk and allowed for in budget costings
services
(iii) Solution of minerals or changes in movements of pollutants
- corrosion of construction materials Discounted except for limited risk of sulphate attack,
allowed for in budget costings
- gas displacement Effects limited to the impounding stage of barrage
construction
- salt leaching Discounted
- solution of pollutants Limited effects on infrastructure, allowed for in budget
costings
- dispersion of pollutants Discounted but a precautionary programme of monitoring
has been recommended
(iv) Reduction in bearing capacity under foundations and in skin Discounted
friction of piles
(v) Settlement or swelling of foundation soils Settlement discounted. Limited effects due to swelling of
foundation soils, allowed for in budget costings
(vi) Structural distress due to increased water pressures on Discounted
tunnels or basement walls
(vii) Increased difficulty with dewatering excavations during Discounted
construction
(viii) Surface blowouts if water in aquifer becomes artesian and Discounted
confining soil is too thin to resist uplift
(ix) Reduction of slope stability Discounted

Table 2. Effects o f local ground conditions on detrimental effects arising from hazard

Local ground conditions

Hazard* Fill over Estuarine alluvium over Gravel over Glacial till over

Estuarine Gravel Glacial Gravel Glacial Rock Estuarine Glacial Rock Gravel Rock
alluvium till till alluvium till

ii 2 3 1 2 1 1 2 3 3 3 1
iii 3 3 2 2 1 0 0 0 0 0 0
v 2 0 0 2 2 2 2 0 0 0 0

* Hazards: ii, dampness/wetness or flooding of basements, tunnels and services; iii, solution of minerals or changes in movements of
pollutants; v, settlement/swelling of foundation soils (settlement discounted as a result of Stage III investigations.
Key: 0, negligible risk of detrimental effects; 1, local conditions will tend to ameliorate detrimental effects; 2, local conditions will
tend to accentuate detrimental effects; 3, local conditions will definitely accentuate detrimental effects.
This classification has been developed specifically for conditions encountered in the study area and should not be applied elsewhere.
Local conditions addressed in the table encompass all soil profiles encountered in the study area.
 CREATION OF F U N C T I O N A L G R O U N D MODELS IN AN URBAN AREA 113

Table 3. R&ks with respect to anticipated groundwater change in o n g o i n g d a t a refinement. This c a n be d o n e by u p d a t i n g

study area zone the d a t a b a s e s in the light o f a d d i t i o n a l g r o u n d verifica-
tion. T h i s o u t p u t c a n be simply a d d e d to t h e p r i m a r y
Zone Anticipated r&e in ambient standing Hazard* computer database.
groundwater level (m)
It is an i m p o r t a n t principle t h a t by its very n a t u r e
ii iii v
such i n f o r m a t i o n systems s h o u l d be held in the public
A 0 to 0.5 0 0 0 d o m a i n a n d s h o u l d n o t be ' o w n e d or o p e r a t e d for profit'
B1 0.5 to 1.5 2 1 0 by a n y c o m m e r c i a l c o n c e r n o r o r g a n i z a t i o n . T h e y
B2 1.5 to 2.5 2 1 1 s h o u l d also be easily accessible a n d user friendly in
C 2.5 to 3.0 3 2 2 o r d e r to m a x i m i z e their value. It w o u l d be unrealistic to
D Zero because zone is currently tidal 0 1 0 a n t i c i p a t e t h a t they s h o u l d be accessible w i t h o u t charge,
b u t the cost o f access s h o u l d be m i n i m a l a n d involve a
* Hazards: ii, dampness/wetness or flooding of basements, tun-
quid pro quo w i t h respect to d a t a p r o v i d e r s in o r d e r to
nels and services; iii, solution of minerals or changes in move-
encourage participation and model development. The
ments of pollutants; v, settlement/swelling of foundation soils
(settlement discounted as a result of Stage III investigations). effective o n g o i n g o p e r a t i o n a n d u p d a t i n g o f this particu-
Key: 0, no risk (but existing problems may persist); 1, some lar d a t a m o d e l presents a m a n a g e m e n t p r o b l e m t h a t to
effects possible; 2, some effects probable; 3, perceptible effects date r e m a i n s unresolved.
will occur in places.
This classification has been developed specifically for condi-
tions encountered in the study area and should not be applied
elsewhere. References
New equilibrium levels will be below existing maxima in Zone D
and in this zone conditions will therefore be improved. ALLEN, J. R. L. & RAE, J. E. 1987. Late Flandrian Shoreline
Oscillations in the Severn Estuary: a geomorphological and
stratigraphic reconnaissance. Philosophical Transactions of
the Royal Society. London, B, CCCXV, 185-230.
In a d d i t i o n , the original g r o u n d a n d g r o u n d w a t e r ANDERSON, J. G. C. & BLUNDELL, C. K. 1965. The sub-drift
b e h a v i o u r m o d e l s , b o t h o f w h i c h c o m p r i s e d interro- rock surface and buried valleys of the Cardiff district.
gatable c o m p u t e r databases, were f u r t h e r refined in Proceedings of Geologists' Association, 76, 367-378.
response to c o n s t r u c t i v e technical criticism by parties EDWARDS, R. J. G. 1997. A review of the hydrogeological
o p p o s i n g the scheme. This w a s d o n e at the request o f studies for the Cardiff Bay Barrage. Quarterly Journal o f
P a r l i a m e n t as a p r e c o n d i t i o n o f passing the Bill for t h e Engineering Geology, 30, 49-61.
project p r o g r a m m e . HAWKINS, A. B. 1984. Depositional characteristics of estuarine
B o t h the engineering geological g r o u n d m o d e l a n d the alluvium: Some engineering implications. Quarterly Jour-
nal of Engineering Geology, 17, 219-234.
c o m p u t e r i z e d g r o u n d w a t e r b e h a v i o u r m o d e l are in the
HORNBY, R. P., EDWARDS, R. J. G., RICE, S. M. M., BENTLEY,
public d o m a i n a n d p r o v i d e a r o b u s t , logical a n d valu- S. P. & VININ6, P. 1993. The presentation of earth science
able technical archive t h a t c a n be i n t e r r o g a t e d for o t h e r information for planning development and conservation:
d e v e l o p m e n t projects or p u r p o s e s . S u c h m o d e l s u n q u e s - illustrated by a study of the Severn Levels. Department of
t i o n a b l y offer a simple a n d effective m e a n s o f achieving the Environment, Vol. 1.

Table 4. Infrastructure types and their sensitivity to potential hazards arising from the proposed barrage

Loeal ground conditions

Hazard* Light development Light developmentMixed Embank- Bridges Sports Buried Drains/ Buried Drains/
shallowfoundations deep foundations/developments ments and arenas services sewers services sewers
(low rise) basements (low rise)
(includinghigh viaducts at depths at depths at depths at depths
rise and >I.5m >l.5m <l.5m <l.5m
Residential Industrial Residential Industrial basements):
Residential/
Industrial

ii Ins Ins Det Det Det Ins Ins Po Pb Det Ins Ins
iii Po Po Pb Pb Pb Po Po Po Pb Det Ins Ins
v Po Po Pb Pb Pb Det Det Ins Po Po Ins Ins

* Hazards: ii, dampness/wetness or flooding of basements, tunnels and services; iii, solution of minerals or changes in movements of pollutants;
v, settlement/swelling of foundation soils (settlement discounted as a result of Stage III investigations).
Key: Ins, infrastructure is insensitive to hazard; Po, possibly some effects should the hazard arise; Pb, probably some effects should the hazard arise;
Det, detrimental effects expected should the hazard arise.
This classification has been developed specificallyfor conditions encountered in the study area and should not be applied elsewhere.
Geological Society, London, Engineering Geology Special Publications
Pipeline route selection and ground characterization, Algeria
P. G. Fookes, E. M. Lee and M. Sweeney

Geological Society, London, Engineering Geology Special Publications 2001; v. 18; p. 115-121
doi:10.1144/GSL.ENG.2001.018.01.17

© 2001 Geological Society of London

Pipeline route selection and ground characterization, Algeria

P. G. Fookes l, E. M. Lee 2 & M. Sweeney 3

1Consultant Engineering Geologist, Winchester, Hampshire, U K
2 Department of Marine Sciences and Coastal Management, University of Newcastle,
Newcastle-upon-Tyne, U K
3 BP-Amoco Exploration, London, U K

Purpose of the study consistent range of properties (the average terrain char-
acteristics and their potential variability), i.e. for the
A terrain evaluation of a remote area of hyperarid desert purposes of the exercise each unit can be considered as a
in central Algeria was undertaken as part of the front- single unit for route planning or costing.
end engineering design (FEED) studies for a gas field In most cases the criterion for identifying units and
development. The work was part of the identification drawing boundaries was a combination of topography
of suitable route options for a large buried pipeline be- and geology. In the study area the landscape could be
tween the gas fields and the existing pipeline network, easily subdivided into a simple sequence of plateaux
some 500 km to the north. The results of the study pro- bounded by marked scarps, low-lying depressions with
vided information to support construction costings and almost fiat plains and sand seas. Interpretation of
preliminary engineering design and construction evalua- near-surface materials, geology and geomorphological
tion of potential geohazards and geotechnical issues, processes was based on experience of landform-process-
notably trench excavatability and dune mobility, and to material relationships in desert environments, plus avail-
help subsequent selection of a specific alignment within able geological maps (see below) and information on the
the corridor. geomorphology of the area (particularly duricrusts and
sand dunes). Subsequent ground truthing in the field
(overflights and fieldwork) has enabled the preliminary
Techniques used evaluation to be confirmed or refined.
In the field drive over many of the terrain character-
The stages involved in the study are summarized in istics were quickly assessed using a 1-5 numerical scale
Figure 1, the work formed part of a broader programme to describe the relative significance between units and
of geological studies to characterize ground conditions subunits (Table 1). In addition, field assessments were
along a preferred pipeline route. The basis for the pre- also quickly made of other relevant engineering geo-
liminary terrain evaluation was a combination of a desk morphological (e.g. wadi dimensions, nature and extent
study review and interpretation of a 1 : 100 000 scale com- of instability, etc.) and engineering geological (e.g. rock/
posite (SPOT and Landsat TM) satellite image of the duricrust strength, fracture spacing, estimated silica con-
route corridor (approximately 100km either side of tent, etc.) features for most terrain units.
the proposed route). This evaluation was carried out in
the U K and was used to identify locations for a limited
number of boreholes and trenches. The results from the Terrain classification
boreholes and trenches were used, along with a limited
'ground truthing' programme in Algeria, to refine the Three main terrain models were identified: the Tademait
terrain evaluation and provide typical engineering geo- Plateaux (Fig. 2), the Grand Erg Occidental (Fig. 3) and
logical and geomorphological characteristics for each the Hassi R'Mel Plateau. Each model was subdivided into
terrain unit. a suite of terrain units and, in most cases, terrain subunits.
The terrain evaluation involved the identification of On the basis of the satellite image interpretation and the
generic terrain models (i.e. landscape types) each with a subsequent ground truthing, each terrain unit and sub-
characteristic assemblage of terrain units. In many areas unit was classified by:
terrain units were divided into subunits. The terrain units 1. Engineering geomorphology, subdivided into:
define areas within which certain predictable combina- 9 surface form;
tions of surface forms and their associated near-surface 9 geomorphological processes (i.e. water, wind, mass
materials and geohazards are likely to be found. At the movement, solution weathering, salt weathering,
scale of mapping, each terrain unit is assumed to have a etc.);

From: GRIFFITHS,J. S. (ed.) Land Surface Evaluation for Engineering Practice. Geological Society, London, Engineering Geology
Special Publications, 18, 115-121. 0267-9914/01/$15.00 9 The Geological Society of London 2001.
116 P . G . FOOKES E T AL.

Satellite Image Desk Study (available

Interpretation reports, scientific
papers etc.)

i
Preliminary Terrain I Plannlng and Design
Evaluation L of Ground
"l Investigation

~nee~g
l
Geomorphological
Field Mapping

l
Terrain Classification ii ....................
i .............................

l
Revised Terrain
Evaluation

Terrain Maps and

l
Assessment of Assessment of Borrow
Models Geohazards Suitability

l
Production of
Preferred Route
Alignment Sheets

Identification of Key Considerations along Preferred

Route
9 excavatability
9 dune mobility

Fig. 1. The work programme.

 PIPELINE ROUTE SELECTION AND GROUND CHARACTERIZATION, ALGERIA 117

Table 1. Terrain characteristic classes

Class 1 Class 2 Class 3 Class 4 Class 5

Undulations/relief Sensibly planar 0-2 m over 2-10 m over 5-50 m over 20-100m over
0.1-1 km 0.1-1 km 0.25-2 km 0.25-2 km
Fluvial systems Sheet flow and As 1 plus fluvial As 2 but channels As 3 but channels As 4 but channels
infiltration channels <5 m <20 m wide, < 100 m wide, >100m wide,
wide, 1 m deep 3 m deep 5 m deep 5 m deep
Flooding None Minor, not Likely, but minor Very likely, Very likely, best
significant realignment training works/ avoided
training works realignment
required required
Trafficability Reasonable road Tracks may exist, No tracks. All low As 3 but lots of Not sensibly viable
(cf. Landrover) usually no ratio 4 x 4, some bogging down
bogging down. bogging down and/or many
Some need for detours required
low ratio 4 x 4
Slope instability No instability seen No or very little As 2 with some Instability Very extensive
or expected instability seen, instability seen commonly seen chaotic
but in certain (avoid or take and expected instability.
circumstances the risk) (avoid or Frequent failure
could be engineer against) (avoid)
expected
Gypsum heave None seen or No or very little Heave seen, Heave common, As 4 but movement
expected instability seen, with vertical with vertical >2m
but in certain movement up to movement up to
circumstances 0.5 m 2m
could be
expected
Pavements and Sand sheet surface, One layer of gravel, One layer of gravel, River alluvium Fan deposits
surfaces no pavement particles not particles
touching touching
Duricrust None seen or Crust (i.e. BS5930 Floating crete Caprock Crete and caprock
expected up to MW, (i.e. BS5930 (i.e. surface-
e.g. gypcrete) >MW, e.g. enriched
calcrete, siliceous bedrock)
calcrete)

Gypsum heave can also be considered a 'salt aggression' index (1 = n o problem; 5 = severe problem), but salt aggression also
depends on moisture and other salts (e.g. chlorides).

9 hazards (e.g. flooding, instability a n d subsidence, 3. Identified principal constraints to route planning.
d u n e mobility, aggressive soils, etc.);
A n u m b e r of simple models were also developed to
9 superficial materials;
d e m o n s t r a t e the relationships between t o p o g r a p h y ,
9 duricrusts;
terrain units a n d particular terrain characteristics: flood
9 bedrock.
hazard, duricrusts, slope instability and g y p s u m mobility
and heave (e.g. Fig. 4a and b).
2. Engineering geology, subdivided into:
9 rock mass structure;
9 fracture spacing;
9 silica class;
9 rock strength; Main findings
9 potential excavatability;
9 suitability for b o r r o w materials; The principal geological a n d g e o m o r p h o l o g i c a l factors
9 trafficability. that influenced r o u t e selection a n d costing were:
118 P. G. FOOKES E T AL.

Fig. 3. Terrain model 2. Grand Erg Occidental.

the presence of very strong to extremely strong, mas- 2. the presence of mobile dunes, generally 5-15 m high,
sive bedrocks and duricrusts with high silica content within the sand sea. From world-wide dune migra-
in parts of the area. As a result there is potential for tion research (e.g. Cooke et al. 1992), it is expected
very slow trenching rates (blasting may be required that these dunes may move by around 10-15 m/year;
in some sections) and high pick (tooth) consumption 3. the presence of widespread instability on scarp
for rock cutting trenching machines; slopes;
 PIPELINE ROUTE SELECTION AND GROUND CHARACTERIZATION, ALGERIA 119

(a)

(b)

Fig. 4. (a) Simple duricrust model. Co) Simple fluvial systems model.

4. the potential for localized scour and exposure of the Excavatability

pipeline in wadi channels, during rare flood events;
5. the potential for subsidence or collapse over the cir- Trench excavatability is controlled by the nature of the
cular, silty sand filled depressions on many plateaux rock/duricrust (notably strength and fracture spacing)
surfaces (dayas); and the equipment used (size, type and method of work-
6. the potential for gypsum heave and aggressive ing). Two geotechnical-related issues are central to esti-
ground conditions; mating costs of trenching operations, namely the rate of
7. the widespread availability of potentially suitable progress and the wear of the cutting tools (bits/picks/
materials for trench backfill after processing (e.g. teeth). Characteristic rock/duricrust strength and frac-
stone pavements, fan deposits, etc.), road construc- ture spacing measurements were estimated for each
tion (e.g. gypsum-rich desert soils) and concretes. terrain unit and subunit as part of the fieldwork pro-
gramme. This information was supplemented by point
The most important factors, trench excavatability and load test data collected during the parallel borehole
dune mobility, are considered below. investigation. For each terrain unit, this information
120 P. G. FOOKES E T A L.

. . . . . . . . ~ ~ I

I
W ~ c , 5

~ ..=..=.q,

i
I
7- Envelope
t \ " of field assessments
\ %% "% "%% and point load tests9

i %'%%% . %%%%

',. I% xd "t, ",.

!w
9%
m tI.N
%
\
! \
\

~tO; , _ . _ . . _ = a . _ ~
0/I o.3 1.0 "~

Fig. 5. Excavatability chart for a terrain subunit.

was plotted on an excavatability chart (developed from 1988) formed during periods of extreme aridity and high
Pettifer & Fookes 1994) to determine the relative pro- wind speeds in the Pleistocene. Along the proposed
gress and potential mode of operation for particular pipeline route, a range of dune features can be observed
machines. An example of this is presented as Figure 5. at different spatial and temporal scales.
An estimate was made of the relative pick (i.e. bit/
tooth) consumption in different terrain units, based on 1. Ephemeral dunes (small, short-term features): prob-
the relative hardness of the rocks/duricrusts and the ably develop and decay in response to the annual
picks (this estimate is specifically for trenching machines wind climate or even daily wind climate. Predomi-
although it is broadly applicable for conventional rip- nantly transverse dunes. Actively mobile. Probably
ping operations as an indication of relative tine wear). <5m high.
Rocks/duricrusts comprising quartz (Vickers hardness 2. Secondary dunes (larger, medium-term features):
= 11 000 MPa), or harder minerals, will cause signiticant probably develop in response to 10-100 year varia-
abrasion of the picks (assumed to be tungsten carbide, tions in the wind climate. They can probably with-
Vickers hardness-- 10 000-18 000 MPa). Thus, trenching stand minor seasonal variability and are shaped by
in those areas with silica-rich materials (i.e. quartz-rich) extreme events. Predominantly transverse, barchan
will cause the highest relative rates of pick consumption. and star dunes. Probably medium sand. Episodically
active. Generally 5-15 m high.
3. Primary dunes (largest, long-term features): reflect
Dune mobility adjustments over very long timescales (e.g. 1000-
t0 000 years). Essentially relict linear features from
The Grand Erg Occidental is a complex sand sea, the arid phases in the Pleistoccne. These features
covering some 100000km 2 (see McKee 1979; Callot generally have sparsely vegetated plinths (low angled
 PIPELINE ROUTE SELECTION AND GROUND CHARACTERIZATION, ALGERIA 121

side-slopes) and are probably of coarser material image interpretation and terrain evaluation in desert
than the ephemeral and secondary dunes. They are environments.
immobile under current climatic conditions and When considering the value of the terrain evaluation
probably up to 100 m high. approach it is useful to compare it with other, perhaps
more traditional, approaches to pipeline route selection.
Amongst the issues relevant to the consideration of The two approaches considered for delivering informa-
route options through the sand sea are: tion on this pipeline route were:
9 the potential for pipe failure due to loading by blown 1. Approach A: a systematic borehole investigation,
sand; with boreholes planned, on average, every 1 km, sup-
9 the potential for exposure of the pipe through wind ported by a walk-over survey;
and/or water scour, leaving lengths of pipe unsup- 2. Approach B: a combination of terrain evaluation
ported; methods and a limited number of boreholes and
9 the identification of efficient routes through the dune trenches (around 50 in total) specifically located to
field to minimize cut-and-fill operations. test the terrain models.
The potential hazards associated with rapid dune move- The terrain evaluation approach (Approach B) was
ment were considered, using a series of empirical sand considered to be the only method that could deliver the
transport models. A helicopter-based reconnaissance of necessary information within the required timescale. It is
ground conditions was made within the sand sea to deter- estimated that it has provided information that is 'fit
mine whether it was possible to define potential routes for purpose' (i.e. supporting 'front-end' cost estimates)
which maximize the use ofinterdune flats and slacks, with at around one-tenth of the cost of a systematic bore-
crossing of active dunes limited to around 10-30% of hole investigation (Approach A). The terrain evaluation
selected alignment. approach also provides a structural view of the principal
cost drivers associated with pipeline construction through
the area and, hence, a framework for considering alter-
native route options, whereas the information provided
by a borehole investigation would be specific to a chosen
Lessons learnt alignment.
The objective of the study was to produce an overview of
the terrain conditions along the pipeline corridor, con-
centrating on potential geohazards and trench excavat- References
ability. The work was mainly based on satellite image
interpretation and less than 20 man-days field survey. CALLOT, Y. 1988. Evolution polyphas& d'un massif dunaire
Large lengths of the proposed route were only observed subtropical: Le Grand ErR Occidental (Alg6rie). Bulletin de
from vehicle or from an aircraft. The techniques involved la Socidtd gdologique de France, 4(4), 1073-1079.
COOKE, R. U., WARREN, A. ~ GOUDIE, A. S. 1992. Desert
are straightforward. However, the successful completion
Geomorphology. University College Press, London.
of the exercise within the significant logistical constraints McKEE, E. D. (ed.) 1979. A Study of Global Sand Seas. USGS
imposed by the security problems in Algeria and the very Professional Paper 1052.
tight timescale (routing information was required within PETTWER, G. S. & FOOKES, P. G. 1994. A revision of the
three months of the start of the studies) relied heavily on graphical method for assessing the excavatability of rock.
the judgement and experience of the authors in satellite Quarterly Journal of Engineering Geology, 27, 145-164.
Geological Society, London, Engineering Geology Special Publications
Building the geological model: case study of a rock tunnel in SW England
P. G. Fookes and D. T. Shilston

Geological Society, London, Engineering Geology Special Publications 2001; v. 18; p. 123-128
doi:10.1144/GSL.ENG.2001.018.01.18

© 2001 Geological Society of London

Building the geological model: case study of a rock tunnel
in S W England

P. G. Fookes j & D. T. Shilston 2

1Consultant Engineering Geologist, Winchester, Hampshire, UK
2 W. S. Atkins Consultants Ltd, Epsom, Surrey, UK

Objectives illustrates how the vagaries of geological exposure and

ground investigation programmes can be evaluated to
A 'geological model' is a representation of the geology give an understanding of the completeness and reliability
of a particular location. ' The form o f the model can vary of such data. This evaluation of data is called here the
widely and include written descriptions, two-dimensional 'determinability' of the geology.
sections or plans, block diagrams, or be slanted towards
some particular aspect such as groundwater or geomor-
phological processes, rock structures and so on' (Fookes
1997, p. 294). Formal creation of a geological model is The project
one of the fundamental processes by which geologists,
geomorphologists and other Earth scientists assemble an This case study concerns a large shallow road tunnel
understanding of the ground conditions at a site. It is a in southwest England, which was subject to arbitration
powerful and cost-effective vehicle for conveying this that was eventually settled out of court. The tunnel was
understanding, often in simplified form, to other disci- constructed through a variably weathered rockmass com-
plines such as civil and structural engineers and planners. prising steeply dipping slates which contained layers and
Geological models are not always easy or straight- irregular bodies of igneous rock.
forward to create. This is particularly so at the desk
study and field reconnaissance phases of site investiga
tion. However, it is during these early phases that a model
(or models) can be particularly useful by helping to set Techniques
out what is known, what is conjectured, and where
Four types of geological field data were available prior
significant gaps in knowledge may lie. Geological field-
to construction of the tunnel (Fig. 1):
work provides important information for the model, yet
much of the geological interpretation of such fieldwork is 9 coastal exposures;
necessarily subjective. The case study described here 9 railway cuttings;

Fig. 1. Geological section determined from exposures near line of tunnel.

From: GRIFFITHS, J. S. (ed.) Land Surface Evaluation for Engineering Practice. Geological Society, London, Engineering Geology
Special Publications, 18, 123-128. 0267-9914/01/$15.00 9 The Geological Society of London 2001.
 Fig. 2. Geology determined from actual exposures and ground investigations.

Table 1. Determinability of geology from different sources of information

Location Approx. Estimated Ability to Ability to Total out Overall When

age physical determine determine of 300% percentage of determinable?
(years) visibility of large-scale small-scale determinable
rock geology geology geology
(%) (geological (geological (%)
domains) styles)
estimate estimate
(%) (%)

Ground investigation
Boreholes - 10 30 5 45 Reasonable
at time of
tender
Trial adit - 75 75 90 240 80 Reasonable
at time of
tender
Surface exposures
Coastal exposures 2000 10 80 30 120 40 Reasonable
at time of
tender
Railway cutting 150 25 100 50 175 60 Unreasonable
at time of
tender
R o a d cutting 9 100 100 100 300 100 N o t available
at time of
tender
Relevance to Determines Determines
anticipated overall daily
construction methodology performance
BUILDING THE GEOLOGICAL MODEL: CASE STUDY 125

Fig. 3. Determinability of geology.

Fig. 4. Anticipated structural geological models.

126 P. G. FOOKES & D. T. SHILSTON

9 abandoned quarries; Results

9 pre-construction boreholes and a trial adit for the road
tunnel. Bedrock in the area of the tunnel consists of metamor-
phosed Upper Devonian marine sediments (predomi-
A fifth set of field data was available during and fol- nantly slates) with intercalated igneous rocks (tufts,
lowing construction (Fig. 1): dykes and lavas) which were folded and faulted mainly
during the Hercynian Orogeny. Ussher (1907) shows
9 road cutting (leading to one of the tunnel portals), thrust faults dipping to the south and roughly coaxial
about 10 years old. with the cleavage of the slates. More recent research has
highlighted the importance of thrust tectonics in the
For contractual reasons, the work described here was Hercynian deformation of the area (Coward & McClay
carried out following construction. Field data were 1983) and, whilst no local major thrust faults are marked
gathered by the authors from geological exposures using on the current 1 : 50 000 British Geological Survey map,
conventional engineering geological mapping techniques small-scale thrust faults are likely to be present. Also
appropriate for geological reconnaissance. The informa- in the area of the tunnel, Dearman (1963) recognized
tion was placed in a broader context using the maps and a zone of wrench faulting trending north-northwest to
memoir of the British Geological Survey (Ussher 1907) south-southeast, which could be extrapolated to the
and later publications (e.g. Coward & McClay 1983). tunnel location. However, evidence for this extrapola-
Logs of the pre-construction trial adit and boreholes tion is sparse. Rocks younger than the Devonian do
drilled for the tunnel project were reviewed, but the drill not outcrop in the area adjacent to the tunnel. The area
core was not relogged. was subjected to very long periods of subtropical weath-
Each type of exposure, the adit and the drill core ering during the Tertiary, followed by the periods of
had different characteristics that can be thought of as periglacial activity during the Quaternary; these con-
different samples of the geology of the area. It became ditions are particularly relevant as the tunnel was at
apparent during the work that, because of these char- shallow depth.
acteristics (such as age and visibility of the rock mass), Incompleteness of geological information at various
each would present different understandings (or models) exposures may not always be recognized and searching
of the geology. for exposures is often time-consuming and frequently

Fig. 5. Anticipated weathering models.

 BUILDING THE GEOLOGICAL MODEL: CASE STUDY 127

unrewarding. However, in this case not only were more was of critical importance for the tunnel's design and
exposures found than were anticipated by inspection of construction. Figure 3 shows some of this information
topographic and old geological maps, but also there was portrayed as a histogram to illustrate how the ability to
a relationship between their condition (determined make observations on discontinuities decreases with
largely by the age of the exposure) and the information increasing age of exposure.
which they yielded. Figures 4 and 5 show various systematic attempts at
Figure 1 is a simple section of the location and extent building three-dimensional (block diagram) geological
of exposure in the area around the tunnel and the geol- models of potential situations with all the information
ogy that was interpolated between the locations. Care available both from boreholes and from the mapping
was taken to ensure that interpolations were reason- fieldwork. Figure 5 is particularly important for a
able and took account of the distances between exposures shallow tunnel. It illustrates the effects of differential
and their location with respect to the structural grain of penetrative weathering along the very steeply dipping
the bedrock. Figure 2 shows the geology that could be cleavage, faults and igneous bodies. The problems of
determined from four specific types of exposure: investigating such rockmasses using boreholes without
having a clear geological model or models in mind can
9 unorientated vertical boreholes at about 50 m spacing
be readily appreciated.
and horizontal boreholes connecting the two portals
Figure 6 shows an attempt to integrate into a single
(from the tunnel site investigation proper);
model the various items of geological information
9 coastal exposures from behind a shingle beach, crudely
available prior to construction of the road tunnel,
estimated to be around 2000 years old;
excluding the boreholes and trial adit. It is reproduced
9 a railway cutting about 150 years old, now partly
from the Masters thesis of a student who studied this
vegetated and covered with slope wash debris;
case history after the conclusion of the arbitration
9 a new road works cutting in good condition.
proceedings (Anderson 1996). The student carried out
Figure 2 also shows estimates of the geological informa- his own independent desk study and reconnais-
tion gleaned by the geological fieldwork. Table 1 sum- sance fieldwork. His model illustrates the simple way
marizes these estimates and illustrates a simple concept in which geological conditions can be portrayed in
of the 'determinability' of the local geological conditions two dimensions, and in which geological hazards and
at the four different types of exposure. It shows how the gaps in knowledge can be highlighted, leading to
influence of weathering and erosion by the current cli- improved planning of ground investigation work (bore-
mate of southwest England clearly reduces the ability to holes, etc.).
collect data from natural and man-made exposures. In summary, our conclusions were as follows.
It also illustrates the difficulties in obtaining a complete
picture from boreholes (especially vertical boreholes) in 9 The tunnel was shallow, with a complex inter-
the cleaved and very steeply dipping Devonian age play of differential weathering and changes in atti-
rocks, particularly information on discontinuities which tude of cleavage; as an additional complexity, it

II I I II II Ill

Ground Leve~ Tunnel

x x South Ecst
North West ~9 ~
~ Port a I

~ _ _ . ~ - - . \ \ x ~ ~' . ~ , ~ ,~.,,,

( ~ ) LocolRies in the vicinity of the Sooth East Parrot (ie the road cutting and small quarries) consisted of slate=
with igneous intrusions and thrust faulting. Only slate was /'ound to the north west. The shape or size of the
igneous intrusions is not known at this stage,

Coastal exposures directly to the east revealed thrust foutting and o small amount of igneous rock.

C ~ ) i nIt is predicted that igneous rocks may be adjacent to zones of thrusting -

coastal exposures and in the railway cutting.
there appears to be a link between the two structures

(~ln coosto~ exposures on increase in weathering grade is seen around thrust faults due to zones of weakened slates being formed.
(~) Coastal exposures revealed vorTatlons Tn dip of slaty cleavage.
After Anderson(lag6)

Illll I I

Fig. 6. Model of geology along tunnel route as anticipated by Anderson (1996) from site investigation
information.
128 P. G. FOOKES & D. T. SHILSTON

had faults which were difficult to predict and thin engineering geologist, who worked on the original arbitration
igneous bodies which appeared effectively impossible case, and Mr Darren Anderson, whose MSc thesis re-evaluated
to predict. some of the original data.
9 It was apparent from the model-building exercises that
more borehole information would have been necessary
to portray a sufficiently accurate picture of the even-
tual tunnel conditions.
References
9 Without the development of such geological models, ANDERSON, D. 1996. The Application of Geology in Optimising
and an understanding of the determinability of the Site Investigation Layout Techniques. MSc Thesis, Queen
geology, the borehole information would have been Mary and Westfield College, University of London.
considered adequate. COWARD, M. P. & McCLAv, K. R. 1983. Thrust tectonics in S.
9 The modelled conditions strongly suggested that de- Devon. Journal of the Geological Society, London, 140,
spite otherwise reasonable borehole coverage at about 215-228.
50 m centres, together with long horizontal boreholes
DEARMAN, W. R. 1963. Wrench faulting in Cornwall and
South Devon. Proceedings of the Geologists' Association,
to improve the geological knowledge, more bore- 74, 265-287.
holes and/or a contract with reference conditions and FOOKES, P. G. 1997. The First Glossop Lecture: Geology
financial provision for the observational method for Engineers: the Geological Model, Prediction and Per-
would have been necessary. formance. Quarterly Journal of Engineering Geology, 30,
293-424.
Acknowledgements. The authors thank Balfour Beaty Civil USSMER, W. A. E. 1907. The Geology of the County Around
Engineering Ltd for permission to publish work carried out on Plymouth and Liskeard. Memoirs of the Geological Survey,
their behalf. We would also like to thank Mr Andy Lewis, HMSO, London.
Geological Society, London, Engineering Geology Special Publications
Development of a ground model for the UK Channel Tunnel portal
J. S. Griffiths

Geological Society, London, Engineering Geology Special Publications 2001; v. 18; p. 129-133
doi:10.1144/GSL.ENG.2001.018.01.19

© 2001 Geological Society of London

Development of a ground model for the U K Channel Tunnel portal

J. S. Griffiths
Department of Geological Sciences, University of Plymouth, Devon, UK

Purpose of survey The conceptual model

Engineering geomorphological mapping is primarily con- Castle Hill Landslide had a highly complex morphology
cerned with identifying and mapping the features on the (Fig. 1). The large-scale morphological mapping was fol-
ground surface. However, detailed mapping linked to lowed by a geomorphological interpretation that utilized
limited subsurface data can provide the first approxi- the results of a preliminary site investigation to provide
mation of a ground model in some situations. In this the basis for a subdivision of the landslide into five zones.
example mapping was carried out at the location of
1. Main landslide backscar with slopes in the range
the proposal Channel Tunnel portal at Castle Hill near
from 35 ~ to 40 ~ The extent of the backscar was
Folkestone, England, prior to any detailed ground
somewhat confused by a Norman defensive ditch
investigations with the aim of establishing the nature
which appeared to have been utilized and cleaned
and extent of the landslide complex that existed at the
out during the Second World War.
site. Full details of the study can be found in Griffiths
2. Benches with minor front scarps within the landslide
et al. 1995.
complex representing displaced blocks associated
with the main landslide movements. Five distinct
landslide blocks were identified reaching a max-
The site imum width of 20 m with their upslope extent clearly
The Channel Tunnel terminal and portal on the UK side indicated by a sharp concave break of slope.
is located immediately below the Etchinghill escarpment, 3. Three secondary landslides were identified within
a scarp slope developed in the Lower Chalk of the North the landslide complex with distinctive backscars and
Downs to the west of Folkestone. Whilst the tunnel itself downslope accumulation areas, the most apparent
was positioned within the lower permeability zone of the being a shallow translational debris slide formed on
Chalk Marl, the works for the terminal and portal the southeastern part of the main landslide.
required excavations in the Lower Chalk, the underlying 4. The landslide accumulation zone to the west of the
Gault Clay and through the solifluction cover of Castle Hill Road. Two separate terraces/benches
Coombe Rock. were visible with slopes of 3-6 ~ separated by a 40 m
The 1:10 560 scale BGS map identified six landslide wide scarp with a slope of 7-10 ~ A 20m wide scarp
complexes within the terminal and portal works area that had slopes of between 8 and 13~ also marked
(Aarons et al. 1977), although there was no information the front of the lower terrace/bench. The base of this
available on the form, depth and current stability of lower scarp was taken as the toe of the landslide on
these features. The design for the main tunnel portal, the BGS 1:10 560 scale maps of the area.
however, required it to enter the hillside through the 5. Made ground. The study of historical maps had
centre of one of the landslides at Castle Hill. shown that there was a significant amount of made
ground within the landslide complex.
This geomorphological interpretation of the morpho-
Techniques used logical form of the Castle Hill landslide is presented in
Figure 2.
Field mapping of the geomorphology was undertaken
at a scale of 1:500 using standard procedures for geo-
morphological survey work (Brunsden et al. 1975). The What the survey established
engineers were able to provide large-scale plans with
contours at 1 m intervals as a base. These plans made Identification of the various components of the land-
detailed mapping much easier and almost certainly slide unit provided a basis for examining both its cur-
increased the accuracy of the geomorphological bound- rent level of stability, the timing of the original and any
aries shown on the final maps. subsequent landslide movements and the likely form of

From: GRIFFITHS,J. S. (ed.) Land Surface Evaluation for Engineering Practice. Geological Society, London, Engineering Geology
Special Publications, 18, 129-133. 0267-9914/01/$15.00 9 The Geological Society of London 2001.
130 J.S. GRIFFITHS

reservoir

,eve,,ed
g r ~ ~ / ~'/
W levelled
I- ground

2;i J

Z, J4 J
q/ spoil /g d /~ # d I"
fl dump /~" 4 ~- i _-/.~--
I/,L. ~/.

9 a-~V." \ i. I z.c~"

5 < 1 W ..~,

W-7 7- 1
70 ..-~-.
.1, ~ , - - . T -~[ . ~u..~__..v--v..v._v.._v-.~-
, . -v-Z"r "~ F

~L

@'I
lOOm
, ~ , I . , ~ , I
contour interval 10 metres
LEGEND
N.
..... Convex break of slope Slope direction with angle

t
in degrees
r-~-'vw ~- Convex change of slope
Undulating slope unit
.... Concave break of slope
Jt ._v_x._ L Concave change of slope

, , , , , Convexo-concave break of slope

-r- n T Convexo -concave change of slope

Fig. 1. Morphology of Castle Hill landslide.

basal and any intermediate shear surfaces. In addition, The boreholes along the line of section contained
following the detailed geomorphological mapping of multiple shear surfaces, shear zones and disturbed
the site, existing borehole data were compiled and some lengths of Chalk Marl and Gault Clay. This clearly
trial pits in the accumulation area were excavated. The established the landslide as having multiple units and
intention was to provide a provisional interpretation of also that there had been more than one factor in the
the mode of failure of the main landslide prior to the development of shear surfaces. It was concluded that
commencement of the main ground investigations (see four separate types of shear surface existed within the
cross-section, Fig. 3). landslide complex.
 DEVELOPMENT OF A GROUND MODEL FOR THE UK CHANNEL TUNNEL PORTAL 131

Fig. 2. Geomorphological interpretation of Castle Hill landslide.

Type 1: deep-seated shear surfaces in the Gault Clay Type 2: the main surfaces associated with the
associated with both the tectonic history of the original Castle Hill landslide. This appeared to
deposit and the effects of erosion of the overlying have failed in the form of a multiple rotational unit
Chalk producing unloading shears which resulted in followed by subsequent block disruption, of the
lateral expansion and extrusion of the clay. form described by Brunsden & Jones (1976) on
132 J. S. G R I F F I T H S

~
m. A O D
o
o +o + m
T ~- T I I I I ?

E
+, + ,== + m
p, = -
=
-
,,, ~

e}.- m
9 .~.++

_8
_o+ ~9 "a
wom+-
!~/ell L FFHH, o_m

o~~~
IIIII
,,
i:.:l
J ++i
"'" /
Io ,
I' x
I
0..
E
<,.+
."m.

z -~
s ql.L I,--i 0
T- =

+- --~ ~
qs I.I. 0 x, ~

,i +
q~.l ql.l. 0 ' - - ~- 2

_o
_g
! '+'
et-I./ql.I. 0
=7 xlil
I,I.
x.
:1 =
Ill

l
.o <

~ + +

i </ i i i I i I ~ I

Z a o v -Lu
 DEVELOPMENT OF A GROUND MODEL FOR THE UK CHANNEL TUNNEL PORTAL 133

Stonebarrow in Dorset. The basal shear surface was subsurface data to be tied into specific boundaries in
estimated to be approximately 20 to 25 m below the geomorphological units in a way that has rarely been
present ground surface. possible elsewhere.
3. Type 3: shears in Coombe Rock gelifluction depos- For the Castle Hill landslide, because of the critical
its located towards the toe of the main landslide. nature of the engineering works, a highly detailed three-
The relationship between the downslope limits of dimensional model of the subsurface geometry was
the land-slide and Coombe Rock deposition was a needed. Whilst subsequent detailed, and very expen-
complex one. Subsequent logging of exposures dur- sive, ground investigations refined the understanding of
ing construction indicated that in places the Castle the landslide complex, the model developed from the
Hill landslide had been thrust over the Coombe author's original mapping and interpretation proved
Rock but then was buried by subsequent Coombe to be extremely robust and formed the basis for focus-
Rock deposits. This gave clear evidence of both the ing the subsequent geotechnical investigations into the
antiquity of the landslide movements and their critical areas.
repeated movement during the post-glacial period. This type of assessment has many applications in
4. Type 4: the shear surfaces associated with the detailed survey work for engineering. Any form of insta-
secondary landslides. These would be continuous bility which might affect engineering works could be
but much shallower than the basal shear surface subject to detailed mapping in order to provide both a
associated with the main landslide movement and framework for subsequent detailed ground investiga-
represent part of a general degradation process. tions and a preliminary ground model suitable for initial
design assessments.
This information and the postulated form of the land-
slide is presented in section in Figure 3. A trial shaft and Acknowledgements. The author would like to TML for per-
adit was excavated into the landslide and this confirmed mission to publish the work on the Channel Tunnel, and John
the multiple rotational form with the basal shear surface Abraham from the University of Plymouth for producing the
located at a maximum depth of 24 m below the present figures in this paper.
ground surface.
The actual distal limits of the landslide were not
finally confirmed until construction excavations took
place. These established that the scarp downslope of the References
lower terrace (Fig. 2) was the toe of the landslide.
AARONS, A., WEEKS, A. G. & PARKES, R. D. 1977. Site
investigation for the Channel Tunnel British Ferry
Terminal. Ground Engineering, May, 43-47.
Similar applications BRUNSDEN, D. & JONES, D. K. C. 1976. The evolution of
landslide slopes in Dorset. Philosophical Transactions of
Geomorphological mapping for civil engineering is not a The Royal Society of London, Series A, 283, 605-631.
new concept. However, the extent of the work carried BRUNSDEN, D., DOORNKAMP, J. C. FOOKES, P. G., JONES,
D. K. C. & KELLY, J. M. H. 1975. Large scale geomor-
out for the Channel Tunnel was far greater than just a
phological mapping for highway engineering. Quarterly
mapping exercise. Historical studies prior to the map- Journal of Engineering Geology, 8, 227-53.
ping provided a vast amount of information about man- GRIFFITHS, J. S., BRUNSDEN,D., JONES, D. K. C. & LEE, E. M.
made features in the landscape which, at the time of 1995. Geomorphological Investigations of the Channel
investigation, were highly degraded. The extent and Tunnel Terminal and Portal. The Geographical Journal,
detail of the geomorphological mapping also allowed 161, 275-284
Geological Society, London, Engineering Geology Special Publications
Terrain measurement using automated digital photogrammetry
J. H. Chandler

Geological Society, London, Engineering Geology Special Publications 2001; v. 18; p. 13-18
doi:10.1144/GSL.ENG.2001.018.01.02

© 2001 Geological Society of London

Terrain measurement using automated digital photogrammetry

J. H. Chandler
Department of Civil and Building Engineering, Loughborough University, Leicestershire, UK

Introduction represent the geometric characteristics of land and

terrain surfaces.
Photogrammetry has traditionally provided a means of
generating three-dimensional spatial data to represent
terrain surfaces, which complements traditional ground- Digital image acquisition
based surveying methods. Although techniques such as
airborne laser scanning (Lohr 1998) and synthetic aper- Digital photogrammetric methods rely upon the use of a
ture radar (Hogg et al. 1993; Vencatasawamy et al. 1998) digital image instead of the more traditional analogue
have developed, photogrammetry remains the primary contact diapositive. The digital image comprises a large
method of generating topographic maps (Wolf 1983; array of pixels, each representing a particular colour or
Capes 1998). One important advantage of photogram- grey-scale value to form the overall image. Space-borne
metry is the flexibility of scale that allows application imaging systems use push-broom sensors to create the
to imagery acquired from ground, air and space. Indeed, digital representation. For large and medium scales,
a new generation of high (i.e. 1 m) resolution satellite digital imagery can be obtained directly using a digital
sensors (Capes 1998) is likely to further increase the camera (Koh & Edwards 1996). Although digital camera
potential applications of photogrammetry. Despite many technology provides instant and appropriate imagery,
advantages, there have been several problems with the direct digital/vertical/aerial image acquisition is rarely
application of photogrammetry using traditional meth- used by the photogrammetric community (Maas & Ker-
ods. Most significantly, there was the requirement to use sten 1997). This is partly because digital cameras remain
an expensive and complex photogrammetric stereo- expensive, but, more crucially, such cameras only gener-
plotter. This ensured that the measurement process was ate images at comparatively low resolution (e.g. the
slow and generally required the skills of an experienced Kodak DCS460 camera costs s and provides an
operator, particularly if results of the highest accuracy image of only 3000 • 2000 pixels). The emulsions used in
were to be obtained. conventional aerial films represent an extremely efficient
Rapid developments in computing hardware and soft- means of storing image data at resolutions equivalent to
ware have allowed the science of photogrammetry to 50 000 • 50 000 pixels. Currently the most cost-effective
develop rapidly during the last ten years (Gruen 1994; means of obtaining digital imagery involves a hybrid
Atkinson 1996; Greve 1996). These developments have approach (Helava 1988) in which the conventional ana-
radically eased many of the problems and limitations logue photograph is converted into digital form using
associated with traditional analogue instrumentation. some form of scanning process. Scanning options and
Use of a purely numerical or analytical solution pro- costs vary widely, but to enable simplified processing
vides flexibility, which assists in two important ways. using full-format aerial images, a purpose-built geome-
Satellite imagery, oblique aerial photography and trically stable scanner should be used, (Warner et al.
ground-based imagery can be used, in addition to the 1996). These are expensive to buy, but bureau scanning
more traditional vertical aerial perspective. Similarly, services provide a cheaper alternative if production vol-
imagery acquired using a variety of non-photogram- ume is low.
metric cameras can be considered to be of value for It should be remembered that converting a photo-
spatial measurement. The most recent advance, known graph into digital form requires significant volumes of
as digital photogrammetry, now allows part of the file storage, which can cause serious system manage-
measurement process to be fully automated. This signifi- ment issues. For example, a normal black and white
cant development ensures that photogrammetry repre- aerial photograph (dimensions 230 • 230mm) scanned
sents an even more versatile and efficient method of at 25 micron resolution (1016dpi) with 256 grey levels
deriving dense digital elevation models (DEMs) to will generate an image consisting of 9200 x 9200 pixels,

From: GRIFFITHS,J. S. (ed.) Land Surface Evaluation for Engineering Practice. Geological Society, London, Engineering Geology
Special Publications, 18, 13-18. 0267-9914/01/$15.00 9 The Geological Society of London 2001.
 o

c~

o
o

o
ol~q

o
 TERRAIN MEASUREMENT USING AUTOMATED DIGITAL PHOTOGRAMMETRY 15

occupying 81 Mb of disk space. A minimum of two Application

images is needed to carry out three-dimensional mea-
surement and if colour images are scanned then these files Photogrammetric software packages, available commer-
need to increase in size by a factor of three. The manage- cially, have been developed for application using either
ment issues raised by working with digital imagery are stereo-satellite imagery (i.e. SPOT) or traditional vertical
described more fully by Colomer & Colomina (1994). aerial photography. One comparatively routine applica-
tion of automated digital photogrammetry is illustrated
by the work of Brunsden & Chandler (1996) who acquired
a new epoch of vertical aerial photography (photo scale
Image measurement and automatic 1:4000) of the Black Ven landslide, Dorset, UK, in
DEM generation March 1995. Automated methods were used to generate
a DEM consisting of 1 000 000 points within an area of
One of the important advantages of digital photogram- 1250 • 800m, a sampling density of one point every
metry is the replacement of the complex, specialized and metre (Fig. 1). Once a DEM had been created, it was pos-
expensive photogrammetric plotter with a more general, sible to create contours (Fig. 1), cross-sections and ortho-
and hence far cheaper, digital computer. Machines photos (Fig. 2). The orthophoto is particularly valuable
running the UNIX operating system have been widely for Earth scientists because it combines the interpretative
used in the recent past, but PC-based platforms are capabilities of the original aerial photograph with the
becoming ever more powerful and capable of manip- positional relevance of a map. The 1995 DEM of the
ulating the large images required. A competent UNIX Black Ven system was used to update a sequence of five
machine can be purchased for costs as little as s a lower resolution manually measured DEMs, which rep-
PC for less. Appropriate photogrammetric software is resented the morphology of the mudslides every ten years
required but competition for this market and the influ- since 1946 (Chandler & Cooper 1989). The improved
ence of software packages designed for remote sensing spatial resolution in combination with climatic and land-
applications have reduced costs significantly (e.g. Erdas slide incidence data allowed the revision of an evolu-
Imagine/OrthoMax and OrthoBase, PCI/EASI-PACE, tionary model (Brunsden & Chandler 1996).
R-WEL/Desktop Mapping System, VirtuoZo). Prices Although application using vertical aerial photo-
vary widely, currently within a range between s graphs will always remain most important, it is signifi-
and s 000. cant that automated digital photogrammetry can be
The second and perhaps most significant advantage of applied to both ground-based and terrestrial imagery. To
digital photogrammetry is the potential to automate ensure that the automated DEM extraction software
various aspects of the measurement process. The essen- remains successful, it is necessary to introduce an extra
tial capability relevant for land surface measurement is stage in the photogrammetric processing and although
the automated measurement of digital elevation models this is not documented in software user manuals, it
from an overlapping stereo-pair, which is now both is possible (Chandler 1999). One application of this
practicable and in an advanced state of development. approach is reported by Pyle et al. (1997) and involved
Such automation is based upon sophisticated image the creation of DEMs representing riverbanks using
correlation or image matching techniques that automa- ground-based oblique photography. The objective was to
tically identify and measure common image patches compare successive DEMs in order to map the spatio-
appearing on two overlapping digital images. Once temporal pattern of bank erosion. The precision of auto-
matched, these two image measurements are transformed matically generated DEMs was + 1 2 m m and, signifi-
into object coordinates using established photogram- cantly, the locations where individual clasts had been
metric methods, and the process repeated. With appro- removed from the gravel riverbank could be identified.
priate hardware and software, this cycle can recur at The same methods have been used to quantify the three-
speeds in excess of 100 points per second and so very dimensional form of exposed and subaerial riverbed
dense and consequently accurate DEMs can be gener- gravels, both in natural riverbed gravels (Butler et al.
ated. In most packages, the derived DEM is in the form of 1998) and in a flume (Stojic et al. 1998).
a regular grid of elevation estimates draped over the
desired area. This ability to measure very dense and
regular grid DEMs has instigated a return to grid-based Limitations
methods of manipulating and presenting height informa-
tion. This will perhaps reverse the trend towards using When considering using digital photogrammetric tech-
DEM processing methods based upon the Delauney niques, it is important to consider some of the require-
triangulation (Petrie & Kennie 1990). This algorithm ments and limitations associated with the method.
efficiently creates a surface from a limited sample of Although not theoretically essential, introducing photo-
terrain elevations in which the break-line assumes great control points into the object space eases photogram-
significance. metric processing significantly and should always be
 oo

~m

o
0
O~
o
,.~

0
r4
 TERRAIN MEASUREMENT USING AUTOMATED DIGITAL PHOTOGRAMMETRY 17

considered. Photo-control points are simply recognizable Developments in software implementing digital photo-
features that appear on the photography, and are situated grammetric methods are significant and readers are
at known geographical locations. Such points provide the encouraged to consider using such packages. The soft-
means to establish the transformation from measure- ware is also 'user-friendly', which allows the inex-
ments on the image, to positions in the desired ground perienced novice user to obtain results. Despite this, it is
coordinate system. The technology available to carry out perhaps important to conclude with several caution-
photo-control surveys has advanced in recent years (Lane ary comments. First, it is necessary to become famil-
et al. 1998). For surveys of small areas using terrestrial iar with the procedures briefly outlined in this paper.
photogrammetry, the modern total station is most effec- More guidance is provided by photogrammetry text-
tive, particularly if combined with an intersection method books (Wolf 1983; Atkinson 1996; Greve 1996) and
of surveying. Over larger areas and using vertical aerial other papers (i.e. Chandler 1999). It is easy to become
imagery, the Global Positioning System (GPS) (Schofield overly ambitious in terms of the size and number of
1994) is particularly appropriate, mainly because GPS DEMs that can be generated in a project. It is prudent to
does not rely upon maintaining 'line of sight' between start with a small area and extend only when experience
points. has been gained. Finally, it must be recognized that it
Although automation afforded by digital photogram- takes considerable care and some expertise to generate
metry provides a distinct advantage, such automation truly accurate surfaces on a routine basis. It is always
can also be interpreted as a weakness. The software will essential to assess the accuracy of generated surfaces
always produce some form of surface, but this may not using independent methods.
necessarily represent the surface that the user requires.
If, for example, the parameters used to control the deriva-
tion of the D E M are inappropriate, then erroneous sur-
faces will be generated (Smith et al. 1996). In vegetated References
areas, the surface generated automatically will represent
ATKINSON, K. B. 1996. Close Range Photogrammetry and
the tops of all visible vegetation and not the underlying
Machine Vision. Whittles, Caithness.
ground surface, which the user may perhaps require. BRtrNSDEN, D. & CHANDLER,J. H. 1996. The development of
Automated D E M acquisition is possibly best suited to an episodic landform change model based on the Black
situations in which there is either a lack of vegetation or Ven mudslide 1946-95. In: ANDERSON, M. J. & BROOKS,
where vegetation is homogenous in terms of foliage S. M. (eds) Advances in Hillslope Processes. John Wiley &
height above the true ground surface. Sharp disconti- Sons, Chichester, 2, 869-898.
nuities in the land surface can also downgrade final BUTLER, J., LANE, S. N. & CHANDLER, J. H. 1998. DEM
surface representation, particularly where resolution of Quality assessment for surface roughness characterisa-
the sampling points is low. It may then become neces- tion using close-range photogrammetry. Photogrammetric
Record, 16(92), 271-291.
sary to measure the three-dimensional characteristics of
CAPES, R. 1998. Developments in earth observation - how new
such discontinuities directly, using break-lines (Petrie & space-borne sensors will affect mapping. Surveying World,
Kennie 1990). 6(6), 24-27.
Many packages provide the opportunity to assess the CHANDLER, J. H. 1999. Effective application of automated
accuracy of points measured automatically using stereo- digital photogrammetry for geomorphological research.
superimposition (Smith et al. 1996). This practice is Earth Surface Processes and Landforms, 24, 51-63.
valuable but although such packages provide the oppor- CHANDLER,J. H. & COOPER, M. A. R. 1989. The extraction of
tunity to edit and correct erroneous height estimates, positional data from historical photographs and their
this process can become impracticable if the density of application in geomorphology, Photogrammetric Record,
generated data is high. 13(73), 69-78.
COLOMER, J. L. 8~;COLOMINA,I. 1994. Digital photogrammetry
at the Institut Cartogr~ific de Catalunya. Photogrammetric
Record, 14(84), 943-956.
GREVE, C. 1996. Digital Photogrammetry: an addendum to the
Conclusion manual of photogrammetry. American Society of Photo-
grammetry and Remote Sensing, Bethesda.
Digital elevation models can now be generated auto- GRUEN, A. 1994. Digital close-range photogrammetry- progress
matically using digital photogrammetry. Such automa- through automation. International Archives of Photogram-
tion has allowed the density of data used to represent metry and Remote Sensing, 5, 122-135.
terrain morphology to be increased by a factor of over HELAVA,U. V. 1988. On system concepts for digital automation.
International Archives of Photogrammetry and Remote
100 compared to manual photogrammetric measure- Sensing, 27/2, Kyoto, 171-190.
ment methods. This development is of distinct value to HOGG, J., MCCORMACK, J. E., ROBERTS, S. m. & GAHEGAN,
terrain measurement because it replaces the tedious and CHANDLER, J. H.M.N. 1993. Automated derivation of
expensive traditional manual procedures and generates a stream channel networks and selected catchment charac-
far higher resolution DEM. teristics from digital elevation models. In: P. M. MATTER
18 J . H . CHANDLER

(ed.) Geographical Information Handling- Research and PYLE, C. J., RICHARDS,K. S. & CHANDLER,J. H. 1997. Digital
Applications. John Wiley & Sons, Chichester, 207-235. photogrammetric monitoring of river bank erosion. Photo-
KOH, A. & EDWARDS, E. 1996. Integrating GPS data with Fly- grammetric Record, 15(89), 753-763.
on-demand digital imagery for coastal zone management. SCHOFIELD, W. 1994. Engineering Surveying. Butterworth-
Association of Geographic Information 1996 Conference Heinemann, Oxford.
Proceedings, 6.1.1-6.1.5. SMITH, M. J., SMITH, D. G. & WALDRAM, D. A. 1996.
LANE, S. N., CHANDLER, J. H. & RICHARDS, K. S. 1998. Experiences with analytical and digital stereoplotters.
Landform monitoring, modelling and analysis: landform Photogrammetric Record, 15(88), 519-526.
in geomorphological research. In: LANE, S. N., RICHARDS, STOJIC, M., CHANDLER, J. H., ASHMORE, P. & LUCE, J. 1998.
K. S. & CHANDLER, J. H. (eds) Landform Monitoring, The assessment of sediment transport rates by automated
Modelling and Analysis. John Wiley & Sons, Chichester, digital photogrammetry. Photogrammetric Engineering and
1-17. Remote Sensing, 645, 387-395.
LOHR, U. 1998. Digital elevation models by laser scanning, VENCATASAWAMY,C. P-, CLARK, C. D. & MARTIN, R. J. 1998.
Photogrammetric Record, 16(91), 105-109. Landform and lineament mapping using radar remote
MAAS, H.-G. & KERSTEN, T. 1997. Aero-triangulation and sensing. In: LANE, S. N., RICHARDS, K. S. & CHANDLER,
DEM/Orthophoto generation from high-resolution still- J. H. (eds) Landform Monitoring, Modelling and Analysis.
video imagery. On the potential of digital cameras onboard John Wiley & Sons, Chichester, 165-194.
an aircraft. Photogrammetric Engineering and Remote WARNER, W. S., GRAHAM, R. W. & READ, R. E. 1996. Small
Sensing, 63(9), 1079-1084. Format Aerial Photography. Whittles, Caithness.
PETRIE, G. & KENNIE, T. J. M. (eds). 1990. Terrain Modelling in WOLF, P. R. 1983. Elements of Photogrammetry. McGraw Hill,
Surveying and Civil Engineering. Whittles, Caithness. Singapore.
Geological Society, London, Engineering Geology Special Publications
Low-cost road construction and rehabilitation in unstable mountain areas
G. J. Hearn

Geological Society, London, Engineering Geology Special Publications 2001; v. 18; p. 135-141
doi:10.1144/GSL.ENG.2001.018.01.20

© 2001 Geological Society of London

Low-cost road construction and rehabilitation in unstable
mountain areas

G. J. Hearn
Scott Wilson Kirkpatrick & Co Ltd, Basingstoke, Hampshire, U K

Rationale Construction design of the

When designing the construction or rehabilitation of
Arun Access Road, Nepal
roads in unstable mountain regions subjected to floods, The area model
landslides, erosion and earthworks failures, it has become
usual practice to employ a selection of the techniques The Arun Access Road was designed to provide access
described in Section 2. The need for adequate land through the remote interior of the Middle Himalaya to
surface evaluation in this context is obvious: (i) the infor- the site of a proposed hydropower installation. The Pro-
mation so produced is essential to many if not most ject Area (here defined as the area containing all plau-
design processes; and (ii) proper evaluation of topo- sible alignment alternatives) is approximately 5000 km 2,
graphy, materials and geohazard is central to the per- and is located entirely in the drainage basin of the Arun
formance of the road construction, and its maintenance River in east Nepal; this major river is one of five that
in the longer term. These concepts are embodied in have cut through the Himalayas, thus maintaining their
Fookes et al. (1985) and in Overseas Road Note 16 southerly course as the mountain ranges have risen.
(Transport Research Laboratory 1997) which deals spe- To illustrate the extreme difficulty that the topography
cifically with geohazards and road design in unstable poses to road alignment, one of the two main alignment
mountain areas. corridors considered was forced to make a cumulative
The selection of the techniques, and the manner in rise and fall of 8800 m, equivalent to the height of M o u n t
which they are applied, will depend on project area con- Everest above sea level, over an alignment distance of
ditions, the availability of the necessary data or docu- 223 km. Slopes are usually steep and underlain by weath-
ments, and the nature of the engineering scheme being ered and intensely fractured rock. Erosion and slope
proposed. The techniques are usually applied in a pro- instability, predictably, constitute significant hazards.
gressive manner, with the need to refine and detail
ground conditions more closely as the site selection and
design procedure takes place. The conclusion to this
Desk study identification of alignment corridors
process is the usual inevitability that final design will not
be established until ground conditions are fully defined Published 1 : 25 000 and 1 : 50 000 scale topographical
during construction. maps, in conjunction with the interpretation of existing
In the case of road rehabilitation projects, existing 1 : 20 000 scale black-and-white aerial photographs, were
cut slope exposures provide a more or less continuous used to define broad route corridors. This was achieved
record of soil and rock conditions above the road, by identifying acceptable topography from an alignment
while the stability of slopes and the observed distress design point of view, and by locating favourable facets
to the existing road pavement provides a useful or features in the landscape, such as high level river
overview of stability, and the reaction of slopes and terraces, potential river crossings, preferred locations for
drainage channels to road works in general. These crossing ridges, and slopes suitable for climbing sections
benefits obviously do not accrue in the case of new (stacked hairpins). Unfavourable features, such as cliffs,
road construction. complex and deeply incised topography, landslides,
The design case studies for a proposed new road areas of slope erosion, eroding river terraces and flood-
in Nepal and the rehabilitation of an existing road in prone areas, were also identified on the aerial photo-
the Philippines are described below to illustrate the graphs. Because it is rare to find an alignment corridor
value of land surface evaluation techniques in each which is devoid of these features, a careful assessment of
case. risk potential is required through ground reconnaissance.

From: GRIFFITHS, J. S. (ed.) Land Surface Evaluation for Engineering Practice. Geological Society, London, Engineering Geology
Special Publications, 18, 135-141. 0267-9914/01/$15.00 9 The Geological Society of London 2001.
136 G.J. HEARN

Ground reconnaissance of potential route corridors maps and later summarized according to the land surface
classifications shown in Figure 1.
Each of the potential route corridors identified from the This combination of intensive desk study and field
desk study was examined in the field. Soil and rock types, mapping techniques provided maximum familiarity with
slope angles, landslides, colluvial deposits, cliffs, river the terrain at the earliest possible stage in the route
terraces, etc. were mapped and plotted onto published corridor selection process. The approach did not rely on
1:25 000 scale contour maps. The alignment corridor landslide hazard mapping in a formal sense, as this
options were plotted onto the contour maps and onto technique is not generally compatible with the need for
enlarged paper prints of the aerial photographs, and rapid and reliable engineering assessment of route
verified in the field using hand-held levelling equipment. practicability, stability and relative cost.
Field notes describing geology, soils, drainage, geomor- The collected data were converted and tabulated in
phology and geohazard were recorded on the contour the office. Each alignment corridor was summarized

o 1 2 ~,Km
l I t . . . . . . .

JITPUR

DIYAI

SIDHUWA

PAKHRIBAS

HILE

TO DHANKUTA
AND DHARAN Alignment options for the Arun Access Road
Ridge top or spur divide
Convex break in slope (slope steepens)
Concave break in slope (slope slackens)
~o Average slope angle (degrees)
~% : Shallow creep or recent failure (marginally stable) Rock type and general foliation dip direction
, Old landslide or erosion catchment Sc-schist (Ca Sc - Calc Schist)
Landslide or erosion scarp Ph-Phyllite
Potentially unstable reck cliffs (shown in colour on original) Gn-Gneiss
: Bedrock exposures Q-Quartzite
[~ Bedrock cliffs F -- Fault or shear zone
: Soil cliff Principal gullies and rivers
(~ : Shallow residual soils (less than 2m) on rock Active floodplain or channel sediment
~' ;=~ : Shallow transported soil (less than 1m) on rock I--'-3 Low level terrace (t0 year flood?)(shown in colour on original)
: Shallow transported residual soils r---i Intermediate and high level terraces (25 year flood?)
~) ~Deep residual soil (greater than 2m) (shown in colour on original)
%%o ,,Shallow colluvium (less than 2m en masse transported soils on rock) Active erosion or shallow slope instability (shown in red on original)
%*= ~Deep colluvium Active gully erosion
A~ Talus/taluvium Active slope failure

Fig. 1. Comparison of route corridor options, originally at 1 : 25 000 scale.

 LOW-COST ROAD CONSTRUCTION AND REHABILITATION IN UNSTABLE MOUNTAIN AREAS 137

Fig. 2. Summary map showing potential slope and drainage hazards at a scale of 1: 75 000 for client
familiarization.

according to estimated length, side slope angles, centre-line was then set out on the ground. Plan and ver-
estimated proportions of soil and rock in excavations, tical profile drawings were prepared with cross-sections
and the length of alignment encountering existing and printed at 20 m intervals.
potential slope instability and erosion (potential instabil-
ity was assessed in a qualitative way by reference to
recorded rock mass characteristics, soil types, drainage Field assessment
and slope angle). Broad cost comparisons were made A highway engineer and an engineering geomorpholo-
between different alignments using estimated quantities gist undertook a systematic assessment of the designed
of rock and soil cut, numbers of culverts and lengths of alignment. Shade, tree cover and the inevitable scale
bridge spans, anticipated lengths and heights of retain- distortion on photographs taken in such steep terrain led
ing wall based on side slope angle and preferred cross- to contour inaccuracies in some areas. In one location,
section, and provisional sums for erosion protection and for instance, a ravine approximately 50 m deep and 50 m
slope stabilization. Summary maps were prepared for wide was unseen on the aerial photographs. It took
the client showing the extent of anticipated slope and almost a week of survey through dense jungle to 'find' a
drainage hazards along the alignment options at a scale suitable alignment that avoided this feature. Less radical
of approximately 1:75 000 (Fig. 2). decisions to shift the centre-line to the right or to the left
were based on the need to avoid unstable or eroding
slopes, or to negotiate difficult topography,
Field assessment of ground conditions often led to
Alignment design the conclusion that it would be preferable to construct the
Colour and black-and-white aerial photographs were road either wholly in cut or wholly on retaining wall,
taken along the chosen alignment corridor in associa- depending upon the anticipated strength, stability and
tion with the Transport Research Laboratory, UK. The volume of excavated materials, or the suitability of the
photographs were produced at a scale of approximately topography and foundations below the road for wall con-
1 : 5000 and covered an area generally 2 km either side of struction. The centre-line was shifted accordingly, but
the proposed alignment. The horizontal and vertical usually within 20 m of the designed alignment, and fre-
alignments were optimized using a photogrammetrically quently less. Engineering geomorphological plans were
derived ground model corrected for distortion by ground prepared as part of this exercise and aerial photographs
control. The suitability of the alignment was checked were used in the field to assist in geomorphological
using aerial photograph interpretation. The resultant interpretation, especially in areas of minor realignment.
138 G.J. HEARN

Output review mapping classified the numerous erosion and landslide

embayments within the route corridor according to
Engineering geomorphological mapping, even at the
mechanism of failure and perceived level of threat
time of detailed alignment review, was therefore under-
posed to the road. The mapping formed a useful checklist
taken mostly as a means of recording features and
of possible landslides and slope failure locations.
ground conditions. This work led to a decision to realign
the centre-line and to provide information on landform,
stability of slopes and drainage channels, preferred
cross-section and the general suitability of materials for Desk study and ground reconnaissance
construction. In most cases, the degree of geomorpho- Land surface evaluation techniques applied to the feasi-
logical detail portrayed on the maps was limited to the bility design comprised aerial photograph interpretation,
delineation of specific features only; there was consid- tabulated inventory and mapping of engineering geol-
ered to be little point in carrying out detailed mor- ogy, geomorphology, landslides and erosion at 1:2000
phological or geomorphological mapping once a design scale. The extent of detail portrayed at this stage was
decision had been made and the requirements of the such that it allowed:
geomorphological assessment had been satisfied. Com-
ments were made on the need for erosion protection, 9 the preferred solution at each site to be defined and its
slope drainage and slope stabilization works. These feasibility to be confirmed;
observations were included in the design report and 9 outline survey and calculation of the likely quantities
ancillary documentation, and the recommendations for of retaining structures, slope stabilization and erosion
remedial works were incorporated as lump sum items in protection works associated with each preferred
the Bill of Quantities. solution.
Slope stability analyses and outline design of stabiliza-
tion measures were performed at high-risk sites using
input parameters derived from the field mapping and
Rehabilitation design of the sensitivity analyses for the range of materials, antici-
Halsema Highway, Philippines pated groundwater and failure types encountered.
A limited amount of subsurface drilling and more
The area model detailed topographical surveys were undertaken at sites
where reconstruction feasibility was uncertain, or where
The Halsema Highway follows the spine of the steeply the derivation of a reliable cost estimate required more
sloping Central Cordillera in north Luzon. The topo- detailed information and interpretation.
graphy is composed of intrusive and extrusive volcanic
rocks, mostly diorites, basalts and breccias, that are
highly fractured and altered, together with limestones,
Detailed design
shales and conglomerates. Three main active fault struc-
tures cross the alignment and form part of the Philippine This was based largely on the confirmation and develop-
Fault which defines one of the most active plate margins ment of feasibility schedules through the preparation
in the world. In July 1990 an earthquake which regis- of engineering geological/geomorphological mapping at
tered 7.8 on the Richter Scale caused major damage to scales of 1: 1000 and 1 : 2000, together with stereonet
the road and its adjacent slopes, as well as structural analysis and a programme of drilling, trial pitting and
damage and loss of life in urban areas. The earthquake cone penetrometer probing. Trial pitting, with close
was followed by a succession of typhoons that resulted control on material descriptions and classification,
in slope failures, erosion of slopes and inevitable road proved particularly valuable from a geological, stability
loss or undermining in several locations. Erosion has and design point of view. The interpretation of mapping
continued with every subsequent typhoon season to the and ground investigation data at each site was then used
point that slopes regress back on an annual basis, to develop slope stability analysis and the scheduling of
requiring further realignment into the hillsides. Erosion road reinstatement and remedial measures. Elsewhere,
scars immediately beneath the carriageway can plunge where slope problems were considered to be less severe,
almost vertically for 20m or more, creating highly design schedules were based on field judgement using
dangerous conditions for traffic. qualitative assessment of engineering geological mapping
and ground investigation data alone.

Initial hazard assessment

Output review
Under a previous contract, landslide hazard mapping
had been undertaken using aerial photographs and The design output comprised 1:2000 scale drawings
ground truthing at a scale of 1:10000. The hazard showing the locations of proposed structures and related
LOW-COST ROAD CONSTRUCTION AND REHABILITATION IN UNSTABLE MOUNTAIN AREAS 139

Fig. 3. Original plan (1:2000 scale) showing principal geomorphological features.

Fig. 4. Original plan (1 : 2000 scale) showing proposed outline remedial works.
140 G.J. HEARN

work, together with detailed schedules and annotated cross-section and the detailed location of the centre-line
site photographs. Figures 3 and 4 illustrate the devel- with respect to cut, fill, topography, foundations and
opment of geomorphological and engineering geological stability was essential to the stability of the formation,
mapping into a general arrangement of works scheduled its associated structures and the slopes upon which it
to reinstate the road formation, improve the stability of was founded. Land surface evaluation at the detailed
adjacent slopes and protect slopes and drainage chan- level was therefore essential.
nels from further erosion. In both the Nepal and Philippines studies, the question
Large-scale engineering geological and geomorpholo- remained as to how much information should be
gical mapping proved to be the most valuable and mostly recorded if design decisions are to be made more or
widely applied of the land surface evaluation techniques less instantaneously, as part of a multidisciplinary engi-
used. The aerial photographs, while useful in examining neering team. There is certainly a point where the recor-
the wider geomorphology of the road corridor, were in ding of too much detail becomes cost-ineffective, and
places so adversely affected by scale distortion and shade deciding on the cut-off point is a matter of judgement,
that they became difficult to use. When compared to combined with a knowledge of what level of detail is
engineering geological mapping and trial pitting, sub- required for design. This comes with experience, but the
surface drilling investigations provided relatively little examples given in Figures 1 and 3 serve as a useful illus-
added value to the geotechnical interpretation of the tration from the point of view of a low cost road scheme.
observable shallow surface instability affecting the road. Having engineering geological and geotechnical person-
However, drilling did confirm that the low strength of nel on the project throughout design and implementa-
surface materials did not improve significantly with tion allows slopes to be reappraised and more data to
depth, and this general lack of foundation stability was be collected, especially as slopes are excavated and the
an important factor in the geotechnical decision not to performance of the design can be evaluated. This staf-
deploy large structural schemes in the approach to road fing arrangement is essential when low-cost engineering
reinstatement and slope stabilization. works are expected to perform under adverse topogra-
phical and stability conditions, when the perception of
hazard, and a design that leads to an acceptable level
Summary of risk, can vary from one wet season to the next.
The geomorphological mapping component used is
In the case of the design of new road construction, that described by Brunsden et al. (1975a, b). The degree
engineering geological and geomorphological evaluation of detail collected and the manner in which it was
can and should play a major role in the development of applied to design represent a development illustrating
design, from initial desk study and route corridor how engineering geomorphology can and should be
feasibility assessment through to detailed design and incorporated into the design process.
the preparation of contract documentation. The use of
aerial photograph interpretation and engineering geo-
logical/geomorphological mapping through the deploy- Did the outcome match the prediction?
ment of relevant personnel directly in the engineering
design team generates this facility. In the Nepal study, a Although neither of the road construction and rehabi-
different set of documents and inputs to decision making litation schemes had been implemented at the time of
and the design itself would have arisen had land surface writing, a large rock avalanche in Nepal provides an
evaluation been undertaken by specialists who were not interesting comparison between outcome and predic-
directly involved in design. The advantages of the tion. Figure 2 summarizes the geotechnical hazards
integrated approach include more cost-effective use of recorded or predicted along one section of the Arun
mapping techniques, a closer familiarization with the Access Road river route. At Tumbetar a combination of
terrain and its implications for design, and an immediate rock fall and toe erosion was described. The 1:2000
input of geological and geomorphological data into scale engineering geomorphological plans prepared for
important design decisions as the project proceeds. this section of the alignment recorded rock fall deposits,
In the case of road rehabilitation, the design relied far adverse jointing and dilation in the cliffs above, and
more on detailed site appraisal, and used a range of consequently a horizontal alignment was plotted that
methods for recording this information. Engineering provided a safety margin between the road and the toe
geological and geomorphological mapping at scales of of the cliffs. A channel was also to be cut in the ter-
1 : 2000 and larger provided an essential means whereby race between the road and the cliff to provide drainage
slopes were evaluated qualitatively and a decision made and extra rock trap capacity. Conventional survey tech-
through team discussion over the relative merits of niques were used to monitor rock dilation in the cliffs
local realignment, widening into the cut or widening adjacent to the road.
on the outside using retaining wall, earthworks design In July 1996 a rock avalanche took place with an
and spoil disposal, etc. The geotechnical design of each estimated volume of 750 000 m 3 and covered the terrace
 LOW-COST ROAD CONSTRUCTION AND REHABILITATION IN UNSTABLE MOUNTAIN AREAS 141

at this location. The trigger mechanism for this ava- References

lanche is unknown, although the earlier rock dilation was
thought to have arisen through seismic disturbance BRUNSDEN, D., DOORNKAMP, J. C., FOOKES, P. G., JONES,
during an earthquake in 1988. The occurrence of the D. K. C. & KELLY, J. M. H. 1975a. Large scale geo-
rock avalanche in this location illustrates the value of morphological mapping and highway engineering design.
land surface evaluation in identifying potential slope Quarterly Journal of Engineering Geology, 8, 227-253.
hazards, but their magnitude and timing, or frequency of BRUNSDEN, D., DOORNKAMP, J. C. FOOKES, P. G., JONES,
D. K. C. & KELLY, J. M. H. 1975b. Geomorphological
recurrence, remain extremely difficult to judge prior to
mapping techniques in highway engineering. Journal of the
the event. Institution of Highway Engineer, 22, 12, 35-Q1.
FOOKES,P. G., SWEENEY,M. MANBY,C. N. D. & MARTIN, R. P.
Acknowledgements. The author would like to thank the 1985. Geological and geotechnical engineering aspects of
Department of Roads, His Majesty's Government of Nepal, low-cost roads in mountainous terrain. Engineering Geol-
and the Department of Public Works and Highways, Republic ogy, 21, 1-152.
of the Philippines for permission to publish this paper. Bob TRANSPORT RESEARCH LABORATORY. 1997. Principles of low
Weekes of Scott Wilson provided the engineering input to field cost road engineering in mountainous regions. Overseas
survey and design in Nepal, while Jonathan Hart, also of Scott Road Note 16, Transport Research Laboratory, Crow-
Wilson, provided some of the geological information from the thorne, UK.
Philippines.
Geological Society, London, Engineering Geology Special Publications
Terrain hazard around the Ok Tedi copper mine, Papua New Guinea
G. J. Hearn, R. Blong and G. Humphreys

Geological Society, London, Engineering Geology Special Publications 2001; v. 18; p. 143-149
doi:10.1144/GSL.ENG.2001.018.01.21

© 2001 Geological Society of London

Terrain hazard around the O k Tedi copper mine, Papua New Guinea

G. J. Hearn 1, R. Blong 2 & G. Humphreys 3

1Scott Wilson Kirkpatrick & Co Ltd, Basingstoke, Hampshire, U K
2 Natural Hazards, Research Centre, MacQuarie University, Australia
3 School of Earth Sciences, Macquarie University, Australia

Rationale The area model

Located in the remote Star Mountains of Western Prov- A regional geomorphological map with a scale of
ince, Papua New Guinea, the Ok Tedi open-cut copper 1:10000 covering an area of 120kin 2 was prepared
mine operates under conditions of particular topogra- using published and project-derived geological map-
phical and geotechnical adversity. Slopes are frequently ping, aerial photograph interpretation and substantial
steep, rocks are highly fractured and weakened due to ground checking in accessible areas. One of the main
tectonics and weathering, and rainfall is world record- outputs from this mapping exercise was the differentia-
breaking, with annual totals exceeding 10m in the tion between areas and landforms dominated by struc-
vicinity of the mine (Jones & Maconochie 1990). The tural control and those areas where large landslides and
mine is planned to occupy an area of 2 km diameter with colluvial deposits predominated. Individual landform
a depth of 500 m at an altitude of 1800 m a.s.l. The mine features identified on these maps included ridge and
supply corridor, along which power, food and other ravine topography, karst landforms and cavity collapse,
essential materials are brought in and copper concen- landslide scars and failure deposits (including rock
trate is piped out, is 160 km long, and is itself at risk avalanche flow structures), igneous dykes, gully erosion,
from slope, flooding and erosional hazards. sackung rock relaxation features and fault lineations.
The threat posed by these hazards to the operation of The geomorphological map enabled patterns of land-
the mine and the viability of its supply corridor has been form development, drainage and slope stability to be
recognized and evaluated by the geotechnical depart- recorded and interpreted in conjunction with maps and
ment of Ok Tedi Mining Ltd (OTML) since the project's cross-sections showing the underlying geology prepared
inception. However, the severity and proximity of this by others in the multidisciplinary study team. This
risk were perhaps not fully realized until 1989 when a mapping programme led to the conclusion that most of
rock avalanche involving perhaps as much as 70• the valley floor landforms are less than 10000 years
106 m 3 of material occurred within a very short distance in age - youthful even in comparison with other areas of
of the mine office. The collapse of the mountainside was the Papua New Guinea highlands. Furthermore, it was
dramatic and was considered to have been responsible for concluded that the most important influence on recent
the seismic event recorded at the time of failure in Port geomorphological evolution in the mine area had been
Moresby, approximately 800km away. Although the the occurrence of large rock avalanches, a conclusion
infrastructure of the mine was not directly affected by this that clearly had important implications for the geotech-
failure, it served as a catalyst for a multidisciplinary nical management of the mine and its infrastructure.
geotechnical study of slope hazards in the mine opera- Thus, the l:10 000 scale regional geomorphological and
tional area, which focused on techniques of land surface geological mapping became the area model and pro-
evaluation rather than intensive subsurface investiga- vided a framework within which other geomorphologi-
tion. These techniques are reviewed below and included cal techniques were applied, and specific land surface
regional geological and geomorphological mapping, haz- risk and management issues were addressed.
ard and risk mapping, engineering geological inventory,
detailed geomorphological mapping, and risk assessment
of landslide and avalanche runout. As such, this study Techniques used
probably represents one of the most intensive and wide
Hazard and risk mapping
ranging applications of land surface evaluation techni-
ques for engineering schemes in remote mountain ter- The findings of the regional mapping study, combined
rain, and is perhaps unique in the mining sector. with the observable slope failures around the mine, led

From: GRIFFITHS,J. S. (ed.) Land Surface Evaluation for Engineering Practice. Geological Society, London, Engineering Geology
Special Publications, 18, 143-149. 0267-9914/01/$15.00 9 The Geological Society of London 2001.
144 G . J . HEARN E T AL.

to the decision by O T M L to fund a landslide hazard and The practical outcome of this study was the develop-
risk mapping study (Fookes 1997). This study was split ment of a model of geomorphological risk posed by
into two parts: the first was at a scale of 1:100 000 and landslides and rock avalanches in the mine area with the
investigated the magnitude and frequency of large land- conclusion that the evolution of the landscape over the
slides; the second was at 1 : 10 000 scale and concentrated last 15 000 years has been dominated by catastrophic
on individual slopes and drainage channels around the slope failures followed by periods of 'recovery' from
mine and the service corridor. The former was based on them. The observation that an avalanche with a volume
aerial photograph interpretation, the regional mapping similar to that at Vancouver Ridge in 1989 can be ex-
and the development of a database of landslide events, pected to occur every 30 years on average in the mapped
including mechanism, geology, volume and approximate area is one of the most significant outcomes in terms of
date or age. The latter was based more on field inventory mine planning. The comparatively high frequency of
using conventional engineering geological methods and small- to medium-sized landslides (up to 100 • 106 m 3)
hazard and risk assessment. has clear implications for the day-to-day running of the
mine, and emphasized the need for a more detailed
review of the mine area and its service corridor.
Magnitude-frequency assessment
Combining historical records of landslides and ava- Hazard and risk mapping
lanches, together with landslide features mapped from
aerial photographs taken since 1969, a database of 79 The establishment of a hazard and risk categorization for
large landslides was established. Volumes were esti- the entire mine operating area was accomplished through
mated from aerial photographs and field survey, and an systematic field inventory assisted by the regional geo-
age was assigned to each failure on the basis of historical logical mapping and a review of drilling logs and other
and community knowledge, landform freshness and geotechnical records held at the mine. Aerial photo-
vegetation patterns, aerial photograph evidence for the graphs at 1:30000 scale were examined during this
more recent failures and 14C dating of organic matter study, but the dense forest cover over much of the area
retrieved from prehistoric avalanches. In addition, the obscured ground detail except in the immediate vicinity
consequences of these failures were categorized accord- of the mine and parts of the service corridor where it had
ing to the geomorphological effects described in Table 1. been cleared.
The largest rock avalanche in the area occurred approx- The study area was divided into 245 geomorphological
imately 8800 years ago with a volume of 7 • 109 m 3 and, units of average size 40 000 m 2 and shown at a scale of
on the basis of landforms mapped from aerial photo- 1:10000. Rock slope and soil slope inventories were
graphs and topographical maps, deposited material to a completed for each unit, detailing engineering geological,
depth of 360 m in the Ok Tedi river. One of the smaller slope geometry, drainage and stability data. These data
avalanches mapped was only 20 x 106 m 3 by compar- were used to compile hazard and risk classifications for
ison and occurred in AD 1977, raising the level of the Ok each unit according to the procedure described in Hearn
Tedi river bed by only 2-3 m, at a distance of 30 km (1995a). The principal geomorphology of the study area
from the avalanche source. was mapped from aerial photographs and field survey,
and was combined with the regional geological mapping
to define geomorphological domains in which hazard
and risk were summarized for easier review by OTML.
Table 1. Magnitude of future landslides estimated from the 100 In addition to providing the basis for the mapping
year landslide record procedure, the inventories and the associated documen-
tation provided recommendations for remedial mea-
Recurrence interval (years) 30 100 1000
sures, further investigation or monitoring schemes, and
Probability of occurrence in a 0.64 0.25 0.03 the timescales over which they should be applied.
30 year period
Volume (• 10 6 m 3) 70 120 >200?
Area (km2) 3.2 4.5 >5? Large-scale geomorphological mapping
Geomorphological consequences* 5 7 8?
Following the failure of Vancouver Ridge, there were
* 1, Cliff or slope retreat; aggradation of lower hillslopes; two main sites of geotechnical concern. The first was
2, movement/erosion/removal of older colluvium; 3, stream Vancouver Ridge itself, where the confirmation of the
bed aggradation and/or stream bed incision; 4, stream bank
trimming and landslide initiation; 5, channel migration; causes, mechanisms and sequence of failure, and an
6, creation of landslide dams, blockage of rivers; 7, truncation assessment of the stability of the remainder of the ridge,
of streams and/or ridges, drainage derangement; 8, catchment were considered essential. The second was New York
boundaries modified by landslides. Ridge, immediately to the east.
 TERRAIN HAZARD AROUND THE OK TEDI COPPER MINE, PAPUA NEW GUINEA 145

The structural geological mapping and three-dimen- With the aid of drillhole information, the geological
sional modelling of Vancouver Ridge remains unpub- mapping led to the conclusion that the thrust which had
lished, though an overview of these studies is provided underlain Vancouver Ridge had been partly intruded by
by Read & Maconochie (1992). Instability on New York the diorite that forms the bulk of New York Ridge, and
Ridge had been monitored for some time by OTML, the thus the potential for widespread failure was lower.
concern being that further deterioration and eventual On the basis of slope steepness, stream undercutting,
slope failure could jeopardize the stability of the pri- past failure morphology and visible signs of recent or
mary crusher, the ore processing facilities, the mine active movement, the geomorphological mapping estab-
waste dump and the mine offices, all of which are located lished a classification of slopes in terms of failure poten-
in the vicinity. tial and potential failure volume, and allowed important
conclusions regarding the stability of critical areas to be
made. However, even with detailed 1 : 500 scale geomor-
phological mapping, augmented by additional drilling
Site 1: Vancouver Ridge and geotechnical analysis, it was not possible to say that
Methods of land surface evaluation applied to the geo- slope failure would not progress to the point that it
morphological evaluation of this area included conven- would eventually compromise the stability of the nearby
tional stereo aerial photograph interpretation (Fig. 1), primary crusher within 20 years. An alternative crusher
sequential plotting of slope erosion, slope instability and site was therefore investigated by OTML.
channel incision assessment (from photography taken
by helicopter at various intervals since the commence-
ment of the project), and 1:2500 and 1:10000 scale Further mapping applications
geomorphological field mapping. The extent to which
field mapping could safely and effectively be undertaken Despite the inability of the multidisciplinary study to
was limited by the precipitous and highly disturbed provide a clear conclusion over the stability of the
slopes that formed the avalanche scarp, and the crusher, it was decided by OTML that a combination of
unpredictability of dense cloud and heavy rain that structural geological and geomorphological mapping,
could very rapidly create extremely dangerous condi- augmented by drillhole information, was the best way
tions for personnel and air rescue. forward in the assessment of slope stability and risk, both
The cause of the avalanche could be explained by in the mine operational area as a whole, and in other key
reference to the following geological factors: an adversely areas. Land surface evaluation techniques were, there-
orientated underlying thrust plane, a hanging wall drop fore, further utilized with the following objectives:
fault which could have formed the avalanche release
surface, and the low strength of the crushed limestone 1. to assess the overall stability of a proposed site for a
that constituted the ridge itself. The triggering mechan- large in-pit mobile crusher;
ism was stream incision beneath the waste dump. 2. to assess the stability of the mine access road in
It was critical to the geomorphological interpretation critical areas;
of the sequence of failure that large-scale slope dilation 3. to determine the risk posed by potential landslides
was detectable from site photographs taken prior to the on the slopes above the mining township.
collapse of the ridge. A combination of aerial photo-
graphic evidence together with geomorphological map- Although aerial photographs were used to provide
ping of the scarp, failed blocks and avalanche flow general interpretative overviews of landform patterns,
structures, enabled the sequence of slope failure and the density of rainforest canopy throughout much of
deposition to be defined. Furthermore, this mapping the study area, outside the immediate vicinity of the
information provided the basis for evaluating the flow mine and its supply corridor, precluded detailed ground
dynamics of the avalanche: most notably, a travel interpretation. Detailed geomorphological studies had
distance of 3500m at a velocity of 70-80 km/h. to rely almost entirely, therefore, on field mapping. This
observation tends to confirm the conclusions of Fookes
et al. (1991) regarding earlier aerial photograph inter-
pretations of a landslide site in the study area.
Site 2: New York Ridge The first two assessments combined 1:2500 scale
structural geological and geomorphological mapping
By contrast, the geotechnical evaluation of New York with drillhole information to develop conclusions that
Ridge relied more heavily on structural geological map- formed the basis of geotechnical decision making (Hearn
ping and detailed geomorphological field mapping over a 1995b). An illustration of the mapping output is shown
much larger area (Hearn 1995b). Field mapping was in Figure 2. The assessment of risk posed by landslide
undertaken at a scale of 1 : 2500 over the entire 4 km 2, and runout onto the terrace occupied by the mining town-
locally at 1:500 in the vicinity of the primary crusher. ship is also reported in Hearn (1995b) and comprised
146 G. J. H E A R N E T AL.

+....

0
11) o)
.~_
o ~" E m

9~ o
i- ._

~ , ~ . . .~ . e ' ~ ~ og " o ..

I=

% ::<>o t :

oo
.,._~

D~_--~ .I ,.7 /
"x-.,\ "6 E~
~'~- ~, 8// o

--,,- -~

Io d,
=2~ .~

>, = I ,

...::...~:

~ ~-~.-
2
o

.:...
9..~ . . . : .
N
<

..~
~ o
"~@ |
i _ c ; >~a=
 T E R R A I N H A Z A R D A R O U N D THE OK TEDI COPPER MINE, P A P U A NEW G U I N E A 147

,.

"-
E
i
~ _,
~
-~ .~
o
,,_
~ ~ ~" ~ .~ ~,
~o

i ~I ~ I ~ ~ 1 7 4 1 7 6[]" o

o
~ o

| ~ ~ | ~ ~ ,-

~o -~'~o ~ l~_ ~ ~ ~ ~ ~ ~ ~o ~
"~

~ ~~ ~ ~ ~ ~, ~ .~ ~ ~
.- "o ~ o - "~
o ,. .o

o~ o o ~ o ~ ~ ~, ~ ~ ~ ~ ~

,~I ,,If ~ ~ ~ ~ ~ U o If, ~~ ~~,~~176 ,

148 G . J . H E A R N E T A L.

,---k

t'q

[..-,
9

o
,.Q

..~
 TERRAIN HAZARD AROUND THE OK TEDI COPPER MINE, PAPUA NEW GUINEA 149

detailed geomorphological mapping, landslide hazard Engineering Services Roads Section and the O T M L
assessment and runout modelling over an area of 3 km 2 Mine Engineering-Geotechnical Section are developing
(Fig. 3). With ground detail obscured by dense rainforest contingency plans for the mine access road.
on aerial photographs, the approach adopted relied
entirely on 1:2500 scale geomorphological mapping to Acknowledgements. The authors would like to thank Ok Tedi
define slope morphology and identify failure landforms. Mining Limited (OTML) for the opportunity to undertake the
The mapping was also used to collect data for a formal study and for permission to publish this paper. The contribu-
tion to the discussion by Mr T. N. Little (current Super-
hazard mapping analysis, together with the development
intendent Geotechnical at OTML) is gratefully acknowledged.
of empirical runout equations to assess the potential for Professor P. G. Fookes acted as review consultant to OTML
future failures to extend onto the township terrace. for the Multidisciplinary Geotechnical Study. Dr R. Mason of
CSIRO, Melbourne, Australia, was responsible for the geolo-
gical mapping described in this paper.
Output
From the consultants' perspective, the land surface References
evaluation and related techniques applied at Ok Tedi
provided both summary and detailed assessment of FOOKES, P. G., DALE, S. G. & LAND, J. M. 1991. Some
hazard and risk posed by landslides and related observations on a comparative aerial photography inter-
phenomena to mine facilities and operations, and in pretation of a landslipped area. Quarterly Journal of
particular the sites of high value housing and processing Engineering Geology, 24, 249-265.
FOOKES, P. G. 1997. Geology for engineers: the geological
plant. This was appreciated not only by the O T M L
model, prediction and performance. Quarterly Journal of
technical staff but also by senior management as a result Engineering Geology, 30, 293-424.
of a series of in-house seminars. These highlighted the HEARN, G. J. 1995a. Landslide and erosion hazard mapping
findings in terms of risk awareness, feasible risk mini- at Ok Tedi copper mine, Papua New Guinea. Quarterly
mization options, and an improved understanding of the Journal of Engineering Geology, 28, 47-60.
way the natural environment behaves and reacts to HEARN, G. J. 1995b. Engineering geomorphological mapping
major slope failures and engineering interference. and open-cast mining in unstable mountains. A case study.
Eight years on, from O T M L ' s current Geotechnical Transactions of the Institution of Mining and Metallurgy,
Superintendent's perspective, 'the multidisciplinary geo- Section A: Minerals Industry, 104, AI-A17.
JONES, T. R. P. & MACONOCHIE,A. P. 1990. Twenty five million
technical study and the associated geohazards work has
tonnes of ore and ten metres of rain. Mine Geologists'
proved a good investment.' The geohazards work is Conference, Mount Isa, 2-5 October, 159-165.
referenced whenever stability issues occur within the READ, J. R. & MACONOCHIE, A. P. 1992. The Vancouver
mapped region, and recently the geohazards studies Ridge landslide, Ok Tedi Mine, Papua New Guinea. Sixth
have formed the basis for geotechnical risk assessment in International Symposium on Landslides, Christchurch, New
the operational area of the mine. Currently, O T M L ' s Zealand, 1317-1321.
Geological Society, London, Engineering Geology Special Publications
GIS-based landslide hazard mapping in the Scotland District, Barbados
G. J. Hearn, I. Hodgson and S. Woddy

Geological Society, London, Engineering Geology Special Publications 2001; v. 18; p. 151-157
doi:10.1144/GSL.ENG.2001.018.01.22

© 2001 Geological Society of London

GIS-based landslide hazard mapping in the Scotland District, Barbados

G. J. Hearn, I. Hodgson & S. Woddy

Scott Wilson Kirkpatrick & Co Ltd, Basingstoke, Hampshire, U K

Rationale interbeds, clays and conglomerates. Folding and faulting

are widespread, and consequently local lithological dip
The Scotland District occupies an area of approximately orientations are complex and multiple. According to the
60km 2 forming the northeast coastline of Barbados published geology (Barker & Poole undated) there is a
(Fig. 1). The district is geologically and topographically general consistent northeast to east-northeast trend in
distinct from the rest of Barbados in that the coral lime- the larger faults and folds, indicating that the sediments
stone cap that covers the remainder of Barbados is have been deformed by northwesterly orientated com-
absent. The underlying Tertiary rocks exposed in the pressive forces. Major displacements have taken place
Scotland District are soft and erodible, and are them- with isoclinal folding, overfolding, thrust-faulting, shear-
selves prone to landsliding and erosion. Land use in the ing and intrusion of Joe's River Beds into the country
district comprises forest, sugar cane, fruit and vegetable rocks as a result of Andean tectonic compression during
growing and pasture, and the district has high landscape the Miocene-Pliocene (Barker & Poole undated). These
value and eco-tourism potential. Low rise, low density beds are structureless, distorted and slickensided, often
housing has developed over the decades, predominantly containing oil and rafts or fragments of the earlier rocks
along cart tracks and roads with alignments that tend (see description in Senn (1940) and Fan et al. (1996)).
to follow ridge and spur lines in an attempt to avoid The geomorphology of the Scotland District is defined
potentially unstable side-long ground. However, there by the delimiting coral limestone escarpment almost
are numerous locations where erosion and ground move- everywhere to the west, and drainage basin topography
ments have caused significant distress to road pave- developed on the Tertiary rocks and recent sediments to
ments, and where foundations to residential buildings the east. The coral limestone escarpment is approxi-
have been undermined to the extent that some areas mately 33 km long and up to 90 m high. The topography
have been abandoned altogether. to the west of the cliff slopes gently towards the west,
The Soil Conservation Unit (Ministry of Agriculture i.e. away from the Scotland District. The gradual reces-
and Rural Development) has been tackling slope insta- sion of the cliff westwards has meant that many valleys
bility and erosion problems in the district for over half a draining westward have become progressively beheaded.
century. A number of useful projects have been commis- On the Scotland side of the cliff, and along the cliff itself,
sioned to examine and remediate areas of known ground there is much evidence of slope failure, with large and
instability, and several schemes involving drainage and numerous detached blocks of limestone deposited in
earthworks have proved successful. However, realizing front of the cliff-line and, in some cases, for considerable
that there were several geological, topographical, drai- distances downslope. Tension cracks are also observed
nage and land use factors distributed throughout the behind the cliff-line and are indicative of stress-relief
district that had a role to play in the pattern of ground and loss of support as the escarpment has retreated.
instability and erosion, the Soil Conservation Unit com- Springs and seepages are obviously important controls
missioned a Geographical Information System (GIS)- on ground stability.
based project to use remotely sensed and desk study data
sources to prepare maps of land degradation. The
approach adopted to develop this GIS-based registra- Techniques used
tion, analysis and mapping is described below.
Methodology
Figure 2 shows the method used to derive maps showing
The area model landslide susceptibility, landslide hazard and planning
guidance in the Scotland District. Initially, a high speci-
The geology of the Scotland District is underlain by fication PC with GIS and image interpretation software
marls, siliceous mudstones, sandstones with clay-shale was established in the offices of the Soil Conservation

From: GRIFFITHS,J. S. (ed.) Land Surface Evaluation for Engineering Practice. Geological Society, London, Engineering Geology
Special Publications, 18, 151-157. 0267-9914/01/$15.00 9 The Geological Society of London 2001.
152 G . J . H E A R N E T AL .

Fig. 1. Study area location.

GIS-BASED LANDSLIDE H A Z A R D M A P P I N G IN THE SCOTLAND DISTRICT, BARBADOS 153

o~.

m
9

=
S

I~ g

",...~

.,..~

N
o

~ :
+.
o
g.

e<
M
..~

1 I
-1
154 G . J . HEARN E T AL.

Unit. Topographical mapping, geological outcrop pat- ment of several geological units, as a result of failure
tern, agricultural soil types, land use and drainage data through underlying strata. Their distribution is therefore
sets were obtained from government sources, predomi- difficult to analyse in any susceptibility mapping as they
nantly in GIS format, and these were co-registered into may bear no simple relationship between stratum, slope
the Soil Conservation Unit GIS. One of the key activities angle and failure.
in this process was the conversion of the different spatial
datasets to conform with the underlying National Grid of
Barbados, thus allowing compatibility in later spatial Field observation
analyses. Having established these factor mapping layers,
a dataset of landslide events and recorded damage to In addition to the field validation exercises required for
roads was established from government records. While the aerial photograph interpretation, observations were
these damage records were of value in describing the his- made of natural slope angles at points of ground failure
torical record of ground movements, they were insuffi- or marginal ground stability in order to identify relation-
cient for hazard mapping purposes. Aerial photographs, ships between material types and limiting angles of
together with field validation and field observation, stability. The lowest threshold slopes were found to
became the principal means by which landslide, erosion range between 11 ~ for clays, 15~ for marls, and 22-28 ~
and drainage data were gathered, interpreted and digi- for sandstones. This range was found to be approxi-
tized into the GIS for susceptibility analysis and hazard mately coincident with relationships derived from the
mapping purposes (Fig. 2). GIS factor analysis described below.

Spatial correlation analysis

Aerial photograph interpretation
The data processing and analysis were mainly carried out
Black-and-white and colour aerial photographs flown in using the Geoprocessing Wizard and Spatial Analyst
1951 and 1991 respectively, and both at approximate extensions within GIS. The correlation analysis used is
scales of 1:10 000, were used systematically to interpret based on the chi-squared statistic that examines the
and map landslide and erosion scars throughout the observed versus the expected distribution of frequencies
Scotland District. Areas of landslide, areas of creep, and according to factor mapping categories. Each GIS factor
areas of disturbed ground likely to be associated with layer, such as slope or geology, was composed of a
landslide deposits were recorded onto overlay transpar- number of classes, such as 0-10 ~ 11-20 ~ etc., or strata
ency. Field validation was progressed in parallel with the type. The area occupied by each of the classes was
aerial photograph interpretation. A total of 253 land- expressed as a percentage of the Scotland District area,
slides and 313 erosion areas were mapped in this way. or the percentage of the area occupied by the sum of the
M a n y of the larger old failures and areas considered to classes, if this did not cover the entire Scotland District.
be disturbed by old landslides and landslide materials are The observed number of landslide areas for the entire
located beneath the coral limestone escarpment. Some of Scotland District was then multiplied by each percentage
these failures appear to have been deep-seated in that to obtain the Expected number of events for each factor
they have involved the detachment of large blocks of class. However, because a given landslide may occupy
limestone escarpment. In fact, many of these larger, more than one strata type or more than one land use type
deep-seated landslides probably involved the detach- for instance, a separate parcel (digitized areal unit) had

Table 1. Distribution of observed and expected landslide/failure frequencies and areas according to susceptibility categories based on
geology and slope

Landslide Landslide Total % Expected Observed O/E (O-E)2/E Expected Observed O/E
susceptibility susceptibility area (m 2) Area no. of no of area (m 2) area (m 2)
combined failure failure
rank areas areas

0-3 Very low 14317274.89 23.6 221 172 0.78 10.47 1 857455.59 1 586878.65 0.85
4-6 Low 14812122.52 24.4 228 186 0.82 7.74 1 921 654.78 1 142516.68 0.59
7-9 Moderate 11 079088.45 18.2 171 165 0.96 0 . 2 1 1437348.58 1 538674 1.07
10-16 High 8226618.11 13.5 127 142 1.12 1.77 1067282.56 1 549589.55 1.45
>16 Very high 12296875.42 20.2 190 271 1.43 34.53 1 595338.51 2061421.15 1.29
Total 936 936 54.72 7879080.03 7879080.03
GIS-BASED LANDSLIDE H A Z A R D MAPPING IN THE SCOTLAND DISTRICT, BARBADOS 155

L~
o

. ,....,

'2

o
156 G. J. HEARN ET AL.

to be created for each. Therefore, one landslide may have None of the other factors analysed, namely soil type,
been included in the spatial analysis as three parcels in proximity to a spring or drainage channel, or land use
order to account for the fact that it spans three geologies. category, showed any meaningful relationship with the
The Observed number of events on each factor class was landslide distribution, and therefore these were not
then counted and the value of Observed/Expected (O/E) included in the susceptibility analysis.
calculated. This provided a direct indication of the ten-
dency of landslides to cluster on different factor classes.
However, this statistic alone did not indicate how sig- Outputs
nificant the control of factor class was on landslide
distribution. To determine significance at different levels Figure 3 shows the landslide susceptibility map prepared
of probability it was necessary to calculate ( O - E)Z/E for the Scotland District, based on the geology and slope
for each factor class and then sum these values to arrive factor analysis described above. Relative susceptibility
at a chi-squared value. mapping was then combined with the distribution of
Despite the need to use frequency data in the statis- mapped landslide hazards from the aerial photographs.
tical analysis, it is often the areal coverage of landsliding A three-fold categorization of land capability and plan-
or erosion, rather than the number of events, which ning guidance was developed from this mapping output.
provides the best indication of the relative susceptibility Similar maps were produced for erosion susceptibility
of a given factor class to failure, as one large failure on a and erosion hazard using the same method. Recommen-
given factor class is considered to be more significant dations for risk assessment and risk management were
than a large number of very small failures on the same made in the accompanying report.
class. Therefore, once correlation significance had been
established on the basis of frequency distribution, it was
the O/E values for landslide area that were used in most Discussion
of the susceptibility ranking of factor classes.
The development of a GIS has facilitated the storage, co-
registration and analysis of spatial data sets. This has
enabled a project-derived landslide distribution to be
Results examined in terms of its relationship with potential
landslide-controlling factors. The mapping outputs from
Geology and slope combined were highly correlated this study provide a Scotland District-wide overview of
with the distribution of landslide areas. However, the relative landslide susceptibility, rather than an indication
relative susceptibility of different rock types was found of absolute instability at any one location. It is important
to be, to an extent, dependent on whether they had been to bear in mind that the study has been based essentially
analysed in terms of landslide frequency or landslide on pre-existing desk study data sets together with
area coverage. Susceptibility based on landslide fre- project-derived landslide and erosion distributions based
quency is likely to be biased towards those rock types on aerial photograph interpretation. The analysis and
that are associated more with a large number of small mapping outputs are reliant on the accuracy and resolu-
slope failures than a small number of large failures. tion of these data sources, and the GIS application
Susceptibility based on area distribution is more repre- technology should not be allowed to give a false sense
sentative of single large failures, and this is reflected in of detail to the user of the maps. This issue can only
the results obtained. The rock types that outcrop close be addressed realistically when more detailed datasets
to the coral cap escarpment were found to be associated become available, usually with the assistance of field
with the larger failures, while a larger number of smaller observation and monitoring.
failures were recorded on steeper sections of slope else- Although essentially a desk study, the procedure has
where. Because of this, the relative susceptibilities based incorporated field observation wherever possible, in
on both analyses were combined into the susceptibil- order to combine and validate desk study interpretation
ity mapping. with observational data. The added value of the GIS
This distribution of susceptibility was then tested approach to this process is that data can be added and
against the mapped distribution of landslide parcels and analysed as they become available, and the land degra-
the chi-squared data shown in Table 1 were obtained. dational model updated and enhanced accordingly.
For landslide frequency there is a sensible progression of
O/E values for increasing susceptibility, as would be Acknowledgements. The authors would like to thank the
expected, and the chi-squared statistic is 54.72, which is Ministry of Agriculture and Rural Development for the
opportunity to undertake this study. Desk study data sets
highly significant. It should be noted that this value is in
were obtained from various government ministries and depart-
excess of that obtained for landslides on geology alone ments through the Ministry of Agriculture, and the authors
even though the latter had more than twice the number would like to thank the Department of Lands and Surveys and
of degrees of freedom. the Coastal Zone Management Unit (Ministry of Environment)
 GIS-BASED LANDSLIDE H A Z A R D MAPPING IN THE SCOTLAND DISTRICT, BARBADOS 157

in particular for their assistance. Mr Glenn Marshall and Mr FAN, CHEN-HUI, ALLISON, R. J. & JONES, M. E., 1996. Weath-
John Warner, both of the Soil Conservation Unit, provided ering effects on the geotechnical properties of argillaceous
invaluable assistance as project counterparts. sediments in tropical environments and their geomorpho-
logical implications. Earth Surface Processes and Land-
forms, 21, 49-66.
References SENN, A. 1940. Palaeogene of Barbados and its bearing on the
history and structure of the Antillean-Caribbean Region.
BARKER, L. H. & POOLE, E. G. Undated. The geology and min- Bulletin of the American Association of Petroleum Geolo-
eral resource assessment of the island of Barbados. Part 1: gists, 24, 1548-1610.
The geology of Barbados.
Geological Society, London, Engineering Geology Special Publications
Ground conditions and hazards: Suez City development, Egypt
D. K. C. Jones

Geological Society, London, Engineering Geology Special Publications 2001; v. 18; p. 159-169
doi:10.1144/GSL.ENG.2001.018.01.23

© 2001 Geological Society of London

Ground conditions and hazards: Suez City development, Egypt

D. K. C. Jones

Purpose (3) the early identification of problematic locations

requiring detailed investigation;
Rapid engineering geological/geomorphological surveys (4) the preliminary identification and delimitation of
are sometimes required for extensive urban and indus- surface material resources;
trial developments on tracts of land for which little or (5) establishing a basis for preliminary assessments of
no background topographic or geotechnical information ground hazard potential.
is available. In areas unobscured by vegetation, geo- The classic example of such requirements arose in the
morphological mapping based on air-photo interpreta- proposal to redevelop and expand dramatically the size
tion augmented by field checking and detailed ground of Suez Town (El Suweis) in the aftermath of the
survey, can quickly provide a robust framework of ter- October 1973 Yom Kippur War. The proposed new
rain units of value in: Suez City was to be developed on 85km 2 of desert
(1) planning cost-effective site investigations; surface that had previously been little investigated
(2) interpreting and extrapolating subsurface point and was presumed initially to be relatively uniform and
data provided by trial pits/boreholes; non-problematic (Fig. 1). However, preliminary site

Fig. 1. Maps of Suez, Egypt, showing (a) topography and extent of urban area in 1976 and (b) proposed layout
for a much enlarged Suez City. The sewage works is the square-shaped built-up area immediately to the west of
the town; the fertilizer factory is the oblong area further to the west.

From: GRIFFITHS, J. S. (ed.) Land Surface Evaluationfor Engineering Practice. Geological Society, London, Engineering Geology
Special Publications, 18, 159-169. 0267-9914/01/$15.00 ~) The Geological Society of London 2001.
160 D . K . C . JONES

investigations revealed unexpected lateral variations in In this case a 1:25 000 scale geomorphological map of
near-surface materials, thereby necessitating a rapid the whole of the Suez City site was prepared by
geomorphological survey to establish terrain condi- interpretation of generally poor quality monochromatic
tions and to identify the potential for aggressive soils air-photography onto acetate overlays. These were sub-
and flash-flood hazard. This was achieved by a team of sequently checked and added to in the field, using pro-
five specialists in three weeks; details are to be found cedures and symbolic notation developed in the earlier
in Doornkamp et al. (1979), Bush et al. (1980) and Bahrain Surface Materials Resources Survey (Door-
Cooke et al. (1982). nkamp et al. 1979, 1980; Cooke et al. 1982) (Fig. 2B).
Available air-photography included 1 : 40 000 scale cover
for the whole area and extending northwards and west-
The site wards beyond the boundaries of the proposed develop-
ment; c. 1:20 000 scale for the area north of the present
Suez Town is located at the northern extremity of the Red town; and some 1 : 60 000 scale enlargements for the coast
Sea where the Suez Canal enters the Gulf of Suez. The and town. All were of some antiquity (1956), which was
proposed urban development was to be on gently sloping an advantage as they showed ground form prior to
and slightly dissected ground (Fig. 1) rising to the west extensive disturbance caused by military activities and
and northwest towards the base of the precipitous Gebel more recent urban developments (the value of old maps
Ataka (871 m) some 3 km beyond the western boundary and photography should never be underestimated). The
of the proposed development (Fig. 2A). The relatively checked boundaries were then transferred onto a base
uniform surface coloration and insignificant topography map created by modifying pre-existing poor quality
were typically deceptive, concealing variations in ground 1 : 25 000 scale maps of the area. The resulting materials-
conditions that only began to emerge with the commence- orientated geomorphological map and accompanying
ment of exploratory subsurface investigations. descriptions of identified units and surface materials
could be used at the local level (as in Fig. 2B) and as a
synoptic map for the whole development area (Fig. 3).
Techniques used The geomorphological map was then reinterpreted
to yield a surface materials map (Fig. 2C and Fig. 4)
Scale and detail of mapping are functions of operator which evolved through the incorporation of subsurface
skills, time and the quality of available base materials. data as they became available from the ongoing site

A. GEOMORPHOLOGICAL SETTING AND LEGEND ,J

Atoka
Steep mountain front providing
drainage direct to fans
'-'--- Fault zones
...........
~ M ..... ~ ~ Figs
t ,B&C
~ ~
Major wadis
~x. Main channels

~/a71 "" " 9 ":?SUEZ: Source area of piedmont sediments

Piedmont slopes
Direction of sediment supply to
piedmont
I

I
Extent of fans
Emerged marine features and sediments
~ ~ ~ ~ 0 5km ers. Height in metres
~ " ~ 7" ~,32"",o,:

Fig. 2. Rationalization of borehole and trial pit information through a geomorphological and geological
interpretation of landforms (Doornkamp et al. 1979).
 GROUND CONDITIONS AND HAZARDS: SUEZ CITY DEVELOPMENT, EGYPT 161

LEGEND
Piedmont Drainage
~'~ Older (higher) ~ Fluvial terrace

Younger (lower) ~ Channel

.... Wadi margin anc
Bedrock 2~.B~ sediments
Exposed or thinly
veneered pediment [~ Wadi outwash
Marine
Miscellaneous
Degraded cliff
Q Quarry
Raised platform
.... Pronounced scarf
Foreshore and
beach
Off- shore bar 0 lkm
i | t

Inter- bar area

Silty, sandy gravel
with shells

Fig. 2. (continued).

investigation. Full details of surface geology and super- of aggressive ground conditions. For this, information
ficial deposits are to be found in Bush et al. (1980). on surface sediments was combined with:
The geomorphological map subsequently formed the
basis for a preliminary assessment of flood hazard poten- (i) further air-photo interpretation to distinguish
tial, where it was combined with analysis of 1 : 40 000 air variations in darkness (dampness) tones in order
photography of the area between Suez City and Gebel to establish the capillary fringe limit (Fig. 6);
Ataka to yield information of wadi catchment areas (ii) ground survey of the distribution of damp ground,
(Fig. 5). The geomorphological and surface materials 'puffy soils' (saline rich), salt efflorescence and
maps were also used as the basis for an initial assessment depth to water table;
162 D . K . C . JONES

C. BEDROCK AND COVER :i:!:!:!:!:!:?:?:!!!

...,.. - - ..,.,.-..

(---

!]ii!iiiiiiiiiii!iiiii!~

t
Sandstone and sandy limestone
- at surface

- beneath O ' l - 2 - O m soils

~ -beneath 2.O-5-0m soils

Silty mudstone

iI i [l~[1
- at surface

- beneath 0 " l - 2 - O m soils

- beneath 2-O-5"0m soils

Bedrock boundary

Cover boundary

J~J Trial pit

Bore hole

Fig. 2. (continued).

(iii) a survey of building damage (Tables 1-3); Conceptual framework

(iv) water sampling for simple conductivity testing
using a portable meter. The programme of investigation was underpinned by
several fundamental concepts.
This yielded a preliminary map of salt weathering poten- 1. Near-coastal desert environments often contain
tial (Fig. 7) that could serve as a basis for a detailed a surprisingly diverse range of surface conditions,
programme of monitoring and chemical analysis. the existence of which is frequently obscured by the
 GROUND CONDITIONS AND HAZARDS: SUEZ CITY DEVELOPMENT, EGYPT 163

not
surveyed

,SUEZ

~ Stormbeach
Foreshoreand beach
Offshorebars
RAISEDMARINE Interbararea
~ Shallowwaterdeposits
Salt pan

f ~~Olderfan
Youngerfan
~ Higherpiedmont
FLUVIAL ~ Lowerpiedmont
~ Wadideposits
l Mudflood
{ ~ Wave- trimmed bedrock
~ Pediment
BEDROCK ~ Riverterrace
0 ki/ometres 3 Reworkedmiocene
I i J I

Quarry ~ Dumpsof wasteor refuse

Fig. 3. Geomorphological map of the Suez area produced by air-photo interpretation and ground mapping.

relatively uniform surface coloration and shallow drifts tive age and the spatial extent of distinctive suites of
of sand and silt. Morphological/geomorphological map- surficial materials. Although such information can be
ping carried out by experienced personnel facilitates built up gradually from site-specific investigations (trial
the identification and delimitation of landforms which, pits/boreholes), it is more efficient to obtain a general
in turn, provide invaluable information on genesis, rela- appreciation at an early stage of a project through
164 D . K . C . JONES

[
50 6G

20--

!iii!i!i!i!~i~i#i!ii:ii:i~!:i:!:i:~

15--

Proqi|e ~ocation
(Figure S)

• Built-upareas

~Beach & raised marine deposits

~ Offshore bar & bar fringes

~ Othercoastaldeposits

~ Fluvialdeposits

f f - ~ Miocene bedrock

0 2kin O5--
I 1 I

45 55 6O
1 t

Fig. 4. Main landform units of the Suez area (Bush et al. 1980).

geomorphological mapping. The established framework provide the basis for examining the specific (location/
of landforms and deposits can then be used as the basis problem) is invariably preferable to using the results of
for planning costly site investigations and extrapolating specific studies to establish the general picture, most
subsurface data obtained in boreholes, trenches, pits especially because potentially problematic sites can often
and natural exposures. Thus using the general picture to be identified at an early stage.
 GROUND CONDITIONS AND HAZARDS: SUEZ CITY DEVELOPMENT, EGYPT 165

Fig. 5. Flood hazard potential, Suez City.

2. Establishing a general (synoptic) appreciation deposits, historical records and personal accounts are all
of terrain conditions (Fig. 3) allows the feasibility of invaluable sources of information that need to be inte-
generalized plans to be evaluated at an early stage. grated together to create a realistic appraisal of hazard
3. The identification of ground hazards requires the potential.
careful assessment of all available types of evidence. 4. The potential for hazard impact at a particular site
For example, morphological evidence of flash flooding due to dynamic or transport hazard (e.g. flood, debris
(i.e. wadi appearance) is often deceptive, as signs of the flow, avalanche, ash fall, etc.) cannot be ascertained by
infrequent occurrence of major discharges can be rapidly reference to the observed conditions at the site alone
obscured by aeolian sediments. Scrutiny of sedimentary but must involve consideration of all areas that could
166 D. K. C. JONES

LIGHT TONES ON AIR PHOTOGRAPHS DARK TONES ON AIR PHOTOGRAPHS

ZONE I ZONE II ZONE III ) ZONE IV

No hazard as Limit Foundations affected Foundations and Foundations emplaced

of Capillary Fringe by Capillary Fringe superstructure at risk in water (water-table
below founda tions < 0.5m ) capillary rise
in buildings up to 2m
above ground level

SABKHA

Lira,to, f
Capillary Fringe "l'['rrwrl'rrvrt~...t,., ............. ~ I Potential L m t of Capillary Fringe

Water-table LLLLLLLI_LIIl_ILLIJl!_JJlI J3!!3!3;!qrjr13i1

-- . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Fig. 6. Methodology used for estimating aggressive ground hazard, Suez City.

yield material to threaten the site in question. This The materials-orientated geomorphological map and
crucial relationship between site and situation means accompanying sediment descriptions readily explained
that site-specific or localized studies are often inadequate the observed variations in site investigation findings and
determinants of hazard potential. What are required allowed the redesign of the site investigation programme
are broader investigations of all areas that could con- to increase cost-effectiveness.
tribute to the threat potential of a specific location: in The preliminary assessment of flood hazard was
this instance, all wadi systems flowing through the accomplished by examining wadi channel form, channel
development zone up to their catchment boundaries on sediments, channel slope, drainage area and reported
Gebel Ataka. occurrence of flash-floods. Although records of recent
flash-floods were limited to two occurrences (in 1970 and
1971), the existence of extensive flood bunds (Fig. 5)
pointed to the recognition of the hazard and both local
Survey results inhabitants and the army were aware of the threat. The
identification of fine-grained sediments, scour zones,
The mapping programme, despite being hampered by
bank erosion and gullying all indicated the occurrence of
the lack of good quality aerial photographs and restric-
flood events, but the crucial evidence with respect to
tions on access arising from military activity, quickly
scale was the recognition of two small areas of remnant
revealed that the apparently uniform desert surface
mudflood deposits (Fig. 3). These had survived from an
actually consisted of three distinct groups of landforms
event of sufficient volume/velocity to rise over a 2m
(Fig. 4):
terrace bluff onto a higher surface. The resulting quali-
(i) low relief features developed on Miocene bedrock tative assessment of flood hazard into low, medium and
(sandstones and siltstones) or bedrock thinly high relative threat (Fig. 5) was based on catchment
veneered with surficial sediments; areas identified on 1:40 000 scale air-photography, and
(ii) emergent marine deposits (beaches, bars, shell clearly indicated that the original development plan
banks, deeper water sediments) with an inland had failed to take sufficient note of flash-flood hazard
cliffed boundary at 17.2m (dated at c. 44000Bp potential, thereby necessitating some redesign and the
(Bush et al. 1980)) representing intermittently fal- incorporation of flood protection measures. The map
ling late Quaternary sea-levels in the Gulf of Suez; (Fig. 5) served as an excellent planning base for such
(iii) low-angled fans and piedmont slopes shallowly developments and also provided an ideal framework for
dissected by wadis, all produced by drainage off subsequent monitoring and modelling programmes.
Gebel Ataka. The survey divided both fans and The rapid investigation of the groundwater-salt sys-
piedmont slopes into two groups on the basis of tem yielded maps of depth to water table, groundwater
age and found evidence for the faulting of bedrock contours and conductivity values, which were combined
against recent gravels, thereby indicating the area to yield a groundwater hazard intensity map (Fig. 7)
to be seismically active. using a four-fold zonal division based on the relationship
 G R O U N D C O N D I T I O N S A N D HAZARDS: SUEZ CITY D E V E L O P M E N T , EGYPT 167

Table 1. Classifications used in survey o f building damage by Table 2. Relation o f degree of weathering to age and type o f
weathering." Suez, Egypt construction." Suez, Egypt

Category or Description Age and Degree o f weathering Totals

measure type of
construction Low High
I. Type of construction - 1 2 3 4 5 6 7 8 9 10
A Load-bearing stone walls throughout sometimes
rendered, often to 75 cm above ground Over 15 years
B Load-bearing stone under brick walls often Stone - 1 6 4 4 5 2 3 3 1 - 29
rendered throughout, sometimes with concrete Brick, stone base - 5 13 13 7 13 7 4 1 63
upper floors through walls. Stone extends to Brick, no stone 3 4 2 4 2 15
3 m above ground base
C Concrete frame, brick infill, usually on stone Concrete frame 1 3 3 - 3 5 - 1 1 17
base 0-0.5 m above ground Suezi 1 6 5 7 2 17 5 1 2 1 - 47
D Suezi type, commonly mixed stone and brick Mixed 2 3 - 2 2 - 1 10
load-bearing walls with a varying proportion Subtotal 2 15 32 31 18 46 18 9 8 2 - 181
of walling as clay-bonded rubble within light
timber cages 3 to 15 years
E Mixed construction other than Suezi Stone 1 1 2 5
Brick, stone base 4 8 14 16 16 12 11 1 1 83
II. Degree of weathering Brick, no stone 3 4 3 4 1 4 2 21
1 Barely perceptible efflorescence base
2 Distinct discoloration but no visible surface Concrete frame 2 4 6 10 7 3 4 36
damage Mixed 1 1
3-4 Heavy discoloration combined with inception of Subtotal 10 17 25 30 24 21 15 1 3 - - 146
decay by spalling of rendering along lower
edges, slight retraction of mortar in joints and Less than 3 years
the like Brick, stone base - 1 1
5-6 Widespread loss of rendering, 25% missing and Concrete frame 5 7 6 1 19
some loosened. Substantial mortar loss in Subtotal 5 8 6 1 20
joints or spalling away of base stone and
brickwork Totals 17 40 63 62 42 67 33 10 11 2 - 347
7-8 Rendering largely gone, mortar lost from lower
joints, about 1 cm of base stone and brickwork
spalled away
9-10 Structural breakdown of wall leading towards
settlement of living areas sufficient to curtail
their utility
Table 3. Relation o f degree o f weathering to state of construction:
III. State of construction Suez, Egypt
0 Uninhabitable for structural reasons
1-2 Long-neglected crumbling structure, probably Age and Degree of weathering Totals
best demolished type of
3-4 Neglected and generally weakened structure, construction Low High
probably capable or restoration - 1 2 3 4 5 6 7 8 9 10
5-6 Stable and generally serviceable structure in
Derelict- - 4 4 5 - 3 2 - 1 19
average state of repair and decoration
1 1 2 2 8 - 14 6 2 1 36
7-8 High quality specification solid and well finished
2 - 3 3 6 3 6 6 4 4 35
structure with apparent indefinite life
3 5 5 13 15 16 18 4 2 3 81
9-10 High quality specification design specifically for
4 - 3 9 5 7 10 10 44
Qalzam conditions (a fully effective damp
5 8 13 22 17 15 14 4 2 1 2 - 98
proof course would be obligatory)
6 2 8 5 4 - 1 1 - 1 22
7 - 2 3 1 - 1 7
8 1 - 2 1 1 5
9
Sound 10

Summary <3 1 9 9 19 3 23 14 6 6 90
3to5 13 21 44 37 38 42 18 4 4 2 - 223
>5 3 10 10 6 1 2 1 - 1 34
Totals 17 40 63 62 42 67 33 10 11 2 - 347
168 D. K. C. JONES

/ ...... m.~ ~.~'~.~'~.

/ .'., .- . .. , .- .... 9 .- 9 .- . .- .. 9 - ,

/
~ . .......................
..... -. -... - ..., - -... 9 -. 9 -. ". -.... ". -. " /~

f. J.."J''.....'J'."'L"
..',,,.,..~"
....~.'-~' "~-".~..-'... .'".."....................
.'- ."' . . . ii. . . . . .
/i : :: i :: i i i : ........
/
I:

...........
' ........... ii !i -" -
~:::::::::::::::::: :::i ~ m
\::::::::
~ . ". " , . ". ". "." l
........... .......

\::::~~

I
] Salt Water Table Electrical Conductivity
Weathering (metres) (;J mhos/cm)
I Zone ~ >100,000
I III & IV <2.0-2.5
[ 50,000-100,000
I <50,000

;
\
\. II c. 2.0-5-0
~ >1o0,ooo
50,ooo-loo,ooo
\ % < 50,000
\
I >5-0 ti.ilJ 3o,ooo-loo,ooo

\ 0 kilometres 3
" L L I I

Fig. 7. Salt weathering (aggressive ground) hazard map, Suez City (electrical conductivity: 1 gmho = l siemen (S)).
 GROUND CONDITIONS AND HAZARDS: SUEZ CITY DEVELOPMENT, EGYPT 169

between the topographic surface, the capillary fringe and vary with the scale of the project, quality of available
the water table (Fig. 6) (see Cooke et al. (1982) for full information, nature of terrain, extent and character of
discussion). The important points to emerge from the vegetation cover, quality and suitability of available
study were: imagery, skill of operators and the time and finance
(i) the extremely shallow depth of the water table available. Dramatic results can be achieved quickly in
along the coastal belt and associated potential for desert environments whereas surveys of difficult forested
capillary rise in buildings; terrain are much more labour intensive. Either way,
(ii) the huge range of conductivity values recorded the resulting synoptic picture of surface conditions
(4.6-388x 103~tScm -1) and the high values in provides a sound framework for subsequent detailed
strongly evaporating zones; investigations, especially when used in conjunction with
(iii) extensive areas of relatively low conductivity (less topographic and geological information. It is invariably
than Red Sea values, i.e. 55-60 x 103 ~tScm -1) in better to start with the generalized picture of surface
the west (near the fertilizer factory), in the centre conditions (i.e. the context) and then to focus in on
of the area focusing on the sewage works and in specific sites/problems, than to attempt progressively
the east beneath Suez Town, adjacent to the Sweet to generalize a context by extrapolation away from
Water Canal and under Feisal City (the north- scattered points of detailed knowledge as they become
western extension of Suez Town) (Fig. 7). available. The same argument has even more force in the
case of the assessment of dynamic or transport hazards,
These last mentioned freshwater caps or mounds are the where establishing the context of a site (i.e. its situation)
products of influent seepage from water pipes, drains, is vital to the assessment of hazard potential (threat).
effluent channels and surface irrigation and clearly point
to the dynamic nature of the water-salt system following
urbanization, thereby indicating the need for monitoring
and modelling during the development of near-coastal References
locations. BUSH, P., COOKE,R. U., BRUNSDEN,D., DOORNKAMP,J. C. &
JONES,D. K. C. 1980. Geology and geomorphology of the
Suez city region, Egypt. Journal of Arid Environments, 3,
265-281.
Similar applications COOKE, R. U., BRUNSDEN, D., DOORNKAMP,J. C. & JONES,
D. K. C. 1982. Urban Geomorphology in Drylands. Oxford
Extensive geomorphological mapping based on the University Press, Oxford.
interpretation of remotely sensed imagery (satellite or DOORNKAMP, J. C., BRUNSDEN,D., JONES, D. K. C., COOKE,
R. U. & BUSH, P. R. 1979. Rapid geomorphological
airborne) with checking/extension by ground survey assessments for engineering. Quarterly Journal of Engineer-
undertaken by suitably skilled personnel, has wide appli- ing Geology, 12, 189-214.
cation where there exists the need quickly to understand DOORNKAMP, J. C., BRUNSDEN, D. & JONES, D. K. C. 1980.
the spatial distribution of surface materials and poten- Geology, Geomorphology and Pedology of Bahrain. Geo-
tial ground hazards. The level of detail achieved will books, Norwich.
Geological Society, London, Engineering Geology Special Publications
Blowing sand and dust hazard, Tabuk, Saudi Arabia
D. K. C. Jones

Geological Society, London, Engineering Geology Special Publications 2001; v. 18; p. 171-180
doi:10.1144/GSL.ENG.2001.018.01.24

© 2001 Geological Society of London

Blowing sand and dust hazard, Tabuk, Saudi Arabia

D. K. C. Jones

Rationale to factors/processes arising from elsewhere (situation);

full details can be found in Jones et al. (1986).
Blowing sand and dust are increasingly common hazards
associated with dryland economic development, largely
because of disruption of desert soils by human activity
(Cooke et al. 1982; Cooke & Doornkamp 1990). Blowing
The site
sand has long been recognized as a problem; it has clearly Tabuk is located on the western side of a sparsely
identifiable sources (blowouts, scour zones) and sinks vegetated basin of interior drainage, the lowest parts of
(drifts, dunes) and is amenable to a range of manage- which contain several playas. The basin has a complex
ment techniques once the problem has been assessed low-relief terrain of bedrock pediments, thinly veneered
objectively (Cooke et al. 1982, 1993). The 'dustification' bedrock surfaces and extensive stone pavements at vary-
problem, by contrast, has come to prominence more ing elevations, separated by minor scarps and bluffs and
recently. It is a more complex process involving greater cut by shallowly incised wadis. Towards the margins of
transport distances, and results in a very wide range the basin the amounts of exposed bedrock progressively
of adverse impacts including abrasion to paintwork, increase, so that pavements give way to rounded and
damage to machinery, contamination of houses, food pinnacled hills cut in near-horizontal sandstone, which
and water, electrical short-circuits, disruption to radio become organized into blocks separated by large, sandy-
signals, killing of plants, suffocation of livestock, the floored wadis. At the western edge of the basin these
spreading of disease, disruption to transport, and gen- blocks become higher and more extensive, with a relative
eral reduction in environmental quality and amenity. relief of over 100m, separated by well-defined wadis
The different modes of t r a n s p o r t - saltation for sand partly filled by sand.
and suspension for d u s t - means that while defensive
measures may be employed against sand, they are less
effective in the case of dust where the emphasis must
be placed on curbing the production of fine-grained The conceptual model
sediment from identified source areas.
The fundamental model underpinning the Tabuk study
was that the assessment of the blowing sand and dust
hazard could be achieved by relating estimations of the
The problem erosivity of the wind to identified patterns of erodibility
of the desert surface. Thus the strength and duration of
The strategic town of Tabuk in NW Saudi Arabia was winds capable of transporting sand and/or dust over
experiencing sand and dust movements of increasing Tabuk had to be related to the nature of the surfaces
magnitude and frequency. These were having adverse over which the winds blow, in terms of their potential
affects on the quality of urban life and military opera- to contribute sediment capable of being entrained. The
tions because of two fundamental processes: (i) rapid transformation of wind roses into sand and dust roses
urban expansion in an extremely arid (28.4mm annual therefore merely represents the first basic step in the
rainfall) environment; and (ii) the increasingly wide- assessment; the potential for aeolian transport then has
spread disturbance of the desert surface around the town to be related to the actual landscape in order to estimate
due to vehicle movements, materials extraction, con- the relative combined strength of erosivity and erod-
struction and agriculture. In order to draw up an effective ibility. This process not only provides a clear indication
management plan to curb the growing impact of these of the directions from which aeolian transport is most
hazards, a rapid geomorphological investigation (three pronounced, thereby providing a basis for planning
specialists for three weeks) was undertaken in 1982 to defensive measures, but also allows the identification of
provide an objective assessment of the problem. The those landform units acting as actual or potential major
study focused on the classic geomorphological task of contributors to sediment transport. This provides a basis
relating adverse conditions at a particular location (site) for hazard mitigation through land management.

From: GRIFFITHS,J. S. (ed.) Land Surface Evaluation for Engineering Practice. Geological Society, London, Engineering Geology
Special Publications, 18, 171-180. 0267-9914/01/$15.00 9 The Geological Society of London 2001.
172 D. K. C. JONES

Techniques used In this particular case only the first three were possible
because of the brief duration of the study.
The objective assessment of sand and dust hazard at a
specific location requires the integration of four main
lines of enquiry:
Model production
9 evidence for the movement of sand and dust;
9 historical data on sand and dust transport episodes; Analysis of remotely sensed imagery is important
9 identification of conditions suitable for the transport because it can give a broad-scale or regional perspective,
of sand and dust; as well as provide a basis for mapping. In the Tabuk
9 identification of source areas for sand and dust. study, a L A N D S A T colour composite of bands 4, 5 and
Once sediment source and aeolian transport parameters 7 was used to establish the regional setting of the town
have been established, it should be possible to predict and to determine the general (prevailing) direction of
the magnitude and frequency characteristics of the sand drift. In addition, a 1 : 50 000 scale monochromatic
hazards and to propose targeted management strategies. air-photo mosaic was used to develop a 1:50 000 scale
To achieve these objectives ideally requires the inte- geomorphological map covering an area of 150km 2
gration of four main methods of data acquisition centred on the town (Fig. 1). This map was checked in
(Cooke e t al. 1982): the field and concentrated on the identification and
delimitation of landform units rather than morphologi-
9 analysis of remotely sensed imagery; cal features, especially in areas of exposed bedrock.
9 meteorological records; Geomorphological survey was then used to establish
9 geomorphological survey; evidence for sand and dust transport, in terms of ero-
9 process monitoring. sional and depositional features, both within the town

Fig. 1. Reproduction of 1: 50 000 scale geomorphological map of Tabuk (from Jones et al. 1986).
 BLOWING SAND AND DUST HAZARD, TABUK, SAUDI ARABIA 173

and in the surrounding desert, and to prove the geo- displaying the greatest vulnerability and highest potential
morphological map. The key refinement was the quali- to yield entrainable sediments, and Type III the greatest
tative assessment of all landform units in terms of their stability (Fig. 1). The final map was then reinterpreted in
sand and dust drift potential, i.e. the extent to which these terms of the sand and dust drift potentials of the recog-
units would generate sand/dust if disturbed by human nized units (Figs 2 and 3).
activity. This was achieved by disturbing the surface of The analysis of limited available meteorological data
units by walking, digging and driving, and resulted in a sought to identify diurnal and monthly patterns of wind
five-fold classification of stone pavements, with Type I speed, and the relationship of higher wind speeds (above

Fig. 2. Map of estimated sand drift potential (Jones et al. 1986).

174 D . K . C . JONES

Fig. 3. Map of estimated dust drift potential (Jones et al. 1986).

11 knots which is the entrainment velocity for sand) with where: Q =proportional amount of sand drift, V=
wind direction and with reduced visibility. These data average wind velocity at 10m, vt=impact threshold
were then used to calculate the potential drift of sand and wind velocity and t = duration of wind.
dust by month and by direction, leading to the produc-
to estimate sand movement, assuming a threshold
tion of 'sand rose' and 'dust rose' diagrams (Fig. 4).
entrainment velocity of 12 knots. It is preferable to use
The production of the 'sand rose' diagram was based
a long run of data but in the Tabuk study only two years
on Fryberger's (1979) method which uses the formula:
(1980 and 1981) of records were available. It should be
noted that the resulting sand rose assumes unlimited
Q ( v - vt)V2t supply of entrainable sand.
 BLOWING SAND AND DUST HAZARD, TABUK, SAUDI ARABIA 175

SAND ROSE
Re;ultant2 4 ~
1
,.o
,-0

.:.:.~:~
~ ' :~-:...$..::~.............-.::~::::.:.:....:
' " ~ [ 0.9

~ i f ~~':'---~ 7.,,

2~.~ @~

DUST ROSE Resultant2 4 . ~

2.3 3.9

"9 ~~.i!iiN'.."~!Ni.:...'::i::i~:.."-:.."~i~
:::::::::::::::::::::::::::::::::
2 8"~ j ~ ~;~::::-~' :::'i:.:.':.:~:u
~"~:~!:':"'-":':':~:':':':': .... 8 - 6
::::::::::::::::::::::::::::::::::::::::::::::::
~~'..:.....:..:::~:.. ~..................
~~;:!:.::i:!:~:~
J ~ ~"::~:":i!:"i:'~:~':':':::::::":
::::~-..'.-.:::~
2s~
Fig. 4. Sand rose and dust rose for Tabuk Airport using data for 1980 and 1981 (Jones et al. 1986).

Calculation of a dust rose is more problematic as the Inspection of the L A N D S A T image revealed an exten-
entrainment of dust can occur at a range of wind speeds sive source of mobile sand some 40 km to the southwest
above 12 knots depending on ground surface character- of Tabuk, from which emanated well-developed belts
istics and the presence or absence of the ballistic impact of sand streaming northwestwards across the topo-
of saltating sand. This problem was overcome by relat- graphic grain of the bedrock uplands towards the Tabuk
ing dust transport to wind data associated with poor basin. Analysis of the orientation of these stream-
visibility (<3000m), once it was ascertained that mist lines and major linear dunes (Fig. 5) revealed dominant
and fog were rare. The vast majority of low visibility sand drift directions of 241.5 ~ across Tabuk and 244.5 ~
conditions were found to be associated with wind speeds slightly to the north of the town, before backing to
of over 20 knots and the resulting 'dust rose' was gen- between 216.5 ~ and 220.4 ~ in the north. The results gave
erally similar to that constructed for sand. substance to the evidence gained from observations of
176 D. K. C. JONES

0 miles 10
i ]
i i
0 km 10
N

1
dune

/
/ Medina Road
107~ N
/ sand
drift
Jebel Rais 241"5~

( • ~ 5 , ~%:...-
S,C y/5 ~and drift from
~. , gap to south of
/~rifts /~' Jebel Rais
and dunes
~ dunesin ,//
lee of hills

Fig. 5. Major dunes and sand drift directions as identified on LANDSAT imagery (Jones et al. 1986).

sand drifts and abrasion features within the town and liminary assessment of available wind data that revealed
its immediate environs. the dominance of winds from the SW quadrant. A quali-
Surprisingly, the 1:50000 air-photo mosaics did not tative three-fold classification of low, intermediate and
show the existence of the major streamlines, thereby high was employed, based on field examination of units.
emphasizing the benefits of employing remotely sensed However, in many cases it was found that the poten-
imagery of different types and scales. The aerial photo- tial for landform units to supply sand or dust varied
graphs did, however, provide a clear indication of the dramatically depending on the nature and extent of
extent of surface sand and the existence of dunes and human disturbance, resulting in the three intermediate
drifts, as well as providing an invaluable base for geo- categories low-medium, medium-high and low-high.
morphological mapping (the existence of stereoscopic These maps (Figs 2 and 3) clearly reveal the existence
pairs would have greatly assisted this process). The of significant contributory areas close to the town and
resultant map (Fig. 1) was based on standard methods therefore vulnerable to disturbance.
developed previously for use in desert areas (Cooke e t al. Meteorological data were only available for Tabuk
1982) but adapted to the particular needs of the Airport, immediately southeast of the town, and were of
investigation, most especially in terms of the division limited duration. Nevertheless, it proved possible to
into 18 surface types and the special focus on pavement show that significant winds (>12 knots or about the
type and stability. entrainment velocity of sand) occurred for about 8% of
The reinterpretation of the 1: 50 000 scale geomorpho- the time, mainly in the period 10am to 2am with a
logical map to identify potential contributory sources pronounced peak in the late afternoon (3pro to 9pm),
of aeolian sand and dust was focused on the western and that the windiest period was January to May in
side of the town. This followed the evidence of sand terms of both wind strength and the duration of strong
streams on LANDSAT imagery (Fig. 5) and a pre- winds. A wind rose was constructed for the airport and
 BLOWING SAND AND DUST HAZARD, TABUK, SAUDI ARABIA 177

then recalculated to produce a sand rose or sand move- The production of a dust rose was more problematic
ment rose (Fig. 4), whose arms are proportional to the (see earlier) and had to be based on wind data associ-
amount of sand that could be moved by winds from ated with visibility recordings of less than 3000m.
different directions using Fryberger's (1979) method (see The results (Fig. 4) were remarkably similar to those
earlier). The resulting 16-arm sand rose based on the for sand, with 88.2% of dust drift potential from the
two years of available data (1980 and 1981) revealed that south and west (169-281 ~ and a resultant of 241.5 ~
83.5% of sand drift potential was achieved by winds The next step was to apply the sand and dust roses to
from 169-281 ~ with easterly winds (011-169 ~ contri- the outline of the Tabuk urban perimeter (Fig. 6) in
buting a meagre 4.2%. The resultant drift direction of order to yield a map showing the directions from which
243.5 ~ was remarkably similar to the 241.5 ~ revealed by most sand and dust probably reaches the town. This
the examination of LANDSAT. clearly showed that defences against sand should be

Fig. 6. Sand and dust drift potentials for Tabuk (Jones et al. 1986).
178 D.K.C. JONES

focused on the southwestern and western margins of potential, resulted in a map showing areas most promi-
the town. nent in the generation of dust if disturbed (Fig. 8).
The assessment of dust hazard was more difficult in
the absence of an established procedure and the study
had to develop a methodology. The rays of the dust rose
(Fig. 4) were applied to the urban perimeter and over- Use of the model
lapping segments added together to yield a dust source
significance map with values recorded as percentages The survey concluded that the careful management of
(Fig. 7). These were indicative of the proportion of areas with high values (certainly those greater than 100
aeolian transported dust from a site likely to affect the and preferably those greater than 50; see Fig. 8) should
town (i.e. the higher the value, the greater the likelihood result in a decrease of the dust problem. Management
that dust generated by surface disturbance will be blown strategies for such areas could include the provision of
on to the town). This map clearly emphasized the impor- metalled roads, the planting of tree belts, restrictions
tance of source areas immediately to the southwest of access onto particularly vulnerable areas, irrigation,
(upwind) of the town. Superimposing this map onto the restriction of building activity to the southwest, etc.
map of estimated dust drift potential (Fig. 3) and The general principle of relating site and situation
multiplying the percentage values (Fig. 7) by 1 (low), underpins a wide range of engineering geomorphological
189 2 (medium), 289 or 3 (high), depending on dust drift investigations, so the approach adopted in this study has

Fig. 7. Dust source potential map for Tabuk (Jones et al. 1986).
 BLOWING SAND AND DUST HAZARD, TABUK, SAUDI ARABIA 179

Fig. 8. Dust source significance map for Tabuk showing areas likely to contribute much dust to the town and
therefore requiring management (Jones et al. 1986).

wide applicability. Similarly, the integration of different and evidence for changing levels of aeolian transport
kinds of remote sensing, ground survey and other forms over time; monitoring of actual aeolian transport epi-
of data analysis is of general relevance. sodes using sand and dust traps; monitoring of the
The methodology described is widely applicable to the behaviour of landform units under different wind condi-
investigation of sand and dust problems in arid and tions; study of the conditions leading to the formation
semi-arid environments. However, it would benefit from of duststorms; and refinement of the methodology for
the following: the availability of stereoscopic pairs of ground surface classification with respect to the poten-
aerial photographs, longer runs of meteorological data tial for moving sand and dust.
180 D.K.C. JONES

References COOKE, R. U., WARREN, A. & GOUDIE, A. S. 1993. Desert

Geomorphology. UCL Press, London.
FRYBERGER, S. G. 1979 Dune forms and wind regime. In:
COOKE, R. U. & DOORNKAMP, J. C. 1990. Geomorphology in MCKEE, E. D. (ed.) A Study of Global Sand Seas. US
Environmental Management. Oxford University Press, Geological Survey Professional Paper 1052, 137-169.
Oxford. JONES, D. K. C., COOKE, R. U. & WARREN, A. 1986. Geo-
COOKE, R. U., BRUNSDEN, D., DOORNKAMP, J. C. & JONES, morphological investigation, for engineering purposes,
D. K. C. 1982. Urban Geomorphology in Dry Lands. of blowing sand and dust hazard. Quarterly Journal of
Oxford University Press, Oxford. Engineering Geology, 19, 251-270.
Geological Society, London, Engineering Geology Special Publications
Sediment budget analysis for coastal management, west Dorset
E. M. Lee and D. Brunsden

Geological Society, London, Engineering Geology Special Publications 2001; v. 18; p. 181-187
doi:10.1144/GSL.ENG.2001.018.01.25

© 2001 Geological Society of London

Sediment budget analysis for coastal management, west Dorset

E. M. Lee 1 & D. Brunsden2

1Department of Marine Sciences and Coastal Management, University of Newcastle,
Newcastle-upon-Tyne, UK
2 Emeritus Professor of Physical Geography, King's College, London, UK

Background Purpose of the study

The identification and characterization of geomorpho- Awareness has been growing of the cumulative effects of
logical units or systems is particularly important in the coastal engineering works on the environment. Particu-
coastal zone. Indeed, the consequences of failure to lar problems have arisen as a result of the disruption of
appreciate the physical environment can be more acute longshore coarse sediment transport by structures such
on the coast, as rapid, major changes are a reality for as harbour breakwaters, piers and groynes. This has
land use planning and development. On the coast, it is led to decline in beach levels and degradation of some
often more useful to map the landscape in terms of sedi- sites of Earth science or nature conservation value.
ment 'cells' (i.e. process units) rather than terrain units It has been recognized, therefore, that there is a need
(i.e. landform units) as an understanding of the supply to evaluate the potential effects of coastal engineering
and transport of sediment (e.g. sand and shingle) is works on the adjacent coast by assessing the signifi-
fundamental to dealing with many shoreline problems. cance of sediment supply and transport to the sustain-
Sediment is circulated in what often can be regarded, for ability of nearby conservation sites and beaches (Rendel
practical purposes, as almost closed cells that are sep- Geotechnics 1995, 1997a).
arated by boundaries across which little beach material The following example describes the development of a
is transferred. Each cell can be characterized in terms of shingle budget and longshore transport model for the
the inputs, outputs, stores and sinks of sediment. eastern part of Lyme Bay, England. The study was
For example, to understand the development of a undertaken to determine the geomorphological impacts
beach it is useful to consider it as a store of shingle or of the current proposed options for coastal defence
sand supplied from source areas on the adjacent coast- improvements at West Bay, Dorset on Chesil Beach and
line or offshore (Fig. 1). Beach building material might the Fleet (Rendel Geotechnics 1997b).
be supplied from the seabed, moved onshore by wave
energy, or from rivers and eroding cliffs. This material is
then redistributed along the shoreline by waves ('long-
shore drift'), unless prevented by barriers such as head- The Lyme Bay coast
lands or breakwaters. Although longshore drift might be
prevented by these barriers, some of the material can The shoreline of the eastern part of Lyme Bay (Fig. 2)
still be 'lost' to the system around the seaward end of the comprises a series of discontinuous shingle beaches,
barriers or offshore, particularly during large storms. which show marked changes in character either side of
Sediment inputs and longshore drift are not necessarily West Bay. To the west, pocket beaches occur at Char-
constant over time and so it is important to consider the mouth, Seatown, Eype and West Beach (West Bay),
current beach behaviour within the context of the changes separated by headlands and backed by high eroding
that might have occurred over the period of the historical cliffs. To the east, the continuous shingle ridge of Chesil
record. Over time, the balance between sediment inputs Beach extends for 28 km from East Beach (West Bay) to
and outputs (i.e. the sediment budget) within the system the Isle of Portland. Most of eastern Lyme Bay is an
will determine whether the beach experiences growth, de- unprotected coastline, although there are notable excep-
cline or has remained constant in overall size. tions: the seawall at Chiswell, the coastal defences and
The sediment cell concept is now used as a shoreline piers at West Bay and the coastal defences and The Cobb
management tool in the UK (e.g. MAFF 1995), as it at Lyme Regis.
provides a framework for assessing the nature of the Chesil Beach is a landform of unique scientific impor-
interdependence of coastal landforms and assessing the tance, both nationally and internationally (Carr &
impact of engineering works. Blackley 1973, 1974). The beach also performs a critical

From: GRIFFITHS,J. S. (ed.) Land Surface Evaluationfor Engineering Practice. Geological Society, London, Engineering Geology
Special Publications, 18, 181-187. 0267-9914/01/$15.00 9 The Geological Society of London 2001.
182 E. M. LEE & D. BRUNSDEN

Fig. 1. Sediment transfers on a typical beach.

coast protection and sea defence role for communities Techniques used
such as Chesilton, the Portland Naval base and the
important lagoon habitat of the Fleet. It is important, The study was based on aerial photograph interpretation
therefore, that proposed coastal defence improvements and field mapping of the coastal landforms, together
at nearby West Bay should not affect the operation of with a review of the available published scientific litera-
coastal processes to the extent that the level of erosion or ture (e.g. PhD theses, consultancy reports, historical
flood risk along Chesil Beach is increased, or lead to the charts held at the Hydrographic Office, Taunton) and
degradation of the conservation value of the coastline. discussions with research scientists who have an expert
 SEDIMENT BUDGET ANALYSIS FOR COASTAL MANAGEMENT, WEST DORSET 183

Fig. 2. Shingle transport cells of Lyme Bay.

knowledge of Chesil Beach and sediment movement in sources. Important outputs include beach mining
Lyme Bay. The study focused on the development of a operations, with possible volumes of shingle extrac-
simple geomorphological model of the coastline between tion identified from previous studies (e.g. Carr 1985).
Lyme Regis and the Isle of Portland. This model pro- 9 The contemporary evolution of the shingle transport
vided the framework for understanding historical and cells and the historical impact of coastal engineer-
contemporary changes and for predicting potential future ing works (such as the West Bay piers) was deter-
changes associated with coastal defence works at West mined from detailed inspection of historical charts
Bay. The model development involved the following. and topographic maps dating back to the end of the
eighteenth century.
9 The identification and characterization of shingle
transport cells along the coastline. Cell boundaries
were defined in the field from known or suspected The geomorphological model
sediment barriers (e.g. natural headlands or man-
made structures). Shingle transport cells
9 The development of shingle budgets for the cells, This coastline comprises a sequence of discrete shingle
based on a review of the available literature (notably transport 'cells' separated by natural headlands and
Bray 1996) and field inspection. Sediment inputs from coastal engineering structures (Fig. 2)"
cliff recession had been previously quantified from a
detailed assessment of the geology and geomorphol- 9 cell 1 - Lyme Regis to Golden Cap (the western end is
ogy of the landslide systems along this coastline (Bray defined by The Cobb);
1996). Inputs from rivers had been considered by 9 cell 2 - Golden Cap to Doghouse Hill;
Rendel Geotechnics (1996). Seabed sediment maps 9 cell 3 - Doghouse Hill to the West Bay piers;
were inspected to identify potential offshore shingle 9 cell 4 - the West Bay piers to the Isle of Portland.
184 E. M. LEE & D. BRUNSDEN

Pre 1740
IL~OK YEN ~ ~BLACK)
LYE REGIS l i e u OOtJEN CR DD6HBJ~ HaL
_ _IB . . . . .
~e------e~ ~e -- -------------~ e---D. II I

1756 - Connection o f The Coblo ~o the Foreshore (1756)

- Construcr o~ West B~y Piers (1744)

~.J ~.~ ~._J . .

EXCHMNGES

c 1850 - CLosure o f Doghouse I-I,I. - Thorncoebe Beo.con Cell Bo~dery

m . U ~ . I L . . T .

Ii- -E .....
EXCHANC~S EXCt,,~
$CUUR

c 1990 - Construction o f b0~tlon Qt Btctck Rock (1983)

]~ST'JI~
NI--L---------.LJ

II ...... FIR
NO '~'--J .a
EXCHANC~$ ND ID(DeUe~ L'XCI'kMeZ EXO.IN~is
19f~

Key, DOnlnant U t t o r ~ l drift direction

S e c o n ~ r y l i t t o r a l dr-let direction

! ! ShingLe supply

Fig. 3. A schematic model of the fragmentation of coastal cells on the West Dorset coast.

Shingle is moved both eastwards and westwards along Historical chart and photograph evidence suggests
this coastline, following the prevailing wind and wave that the four discrete shingle transport cells are probably
direction at the time. However, as there is a long-term the remnants of what once may have been a continu-
prevalence for strong southwesterly and westerly winds, ous shingle beach extending from Lyme Regis to the
the overall resultant drift is eastwards. Periods of domi- Isle of Portland. Prior to the construction of The Cobb
nant easterly winds do occur, leading to net drift reversals this beach may have even extended westwards into East
when the dominant shingle movement is to the west. Devon. However, the long-term effects of differential
SEDIMENT BUDGET ANALYSIS FOR COASTAL MANAGEMENT, WEST DORSET 185

G ~
9~ ~'~'E-',
~.~ ~.~
r OO r t~
o'~ "R. r ~ r
"~: Cq e,~ r-- V'-

' ' 7 7

OO OO OO

.~&~ o o o o
.~ ~
~ ~.~

I I f

Z Z ei Z N

d o o o o

~.~ q-~ h ~

s
o
IN 0 ~,
O
,,s cs Z Z ,6

8
a Z Z Z Z Z
E

("4

ID

r ~ r',,1

oo o~ I----
Po~ o Z Z ,6

~N

~s
186 E. M. LEE & D. BRUNSDEN

erosion of the more resistant headlands (Golden Cap, only episodically linked by shingle transfers. The frag-
Doghouse Hill to Thorncombe Beacon) and the softer mentation of the continuous beach into discrete pocket
rocks in the intervening bays has gradually created beaches may have been a relatively recent phenomenon
the present-day sequence of pocket beaches, which are (Fig. 3).

L c 1750
VEST ] ~ V ~ H O ~ ' T IEJ~CH CHKSlL E A C H PORTLAND

i217
I_ A
II 9
9
9
Ion 0 t e r ~
rmor Int~-F~ence
b4r ocross
net drift

hsrbour
A ---,-B
wrth d r Y *
rmut~h
by pin's
V C

9 sh~g~ supply F~m t h e w e s t Oe o p e n s y s t e m )

9 onsho.-~-~rFshore exc~nOes
2. c l~0s VeSt Boy piers Inl~ted

NET LOSS NET CVbIN

! 84 11 11
R[BUCL'9 SUPPLY
9
-
9
Ion9 1 ~ - ~ t o r t d r ~ t
drift ~
n~t tronsFer
9 tort loss Fron Vest
A ----3
b y mnFIlled p m - s
F r o n A' - - - - - F
)each
V c

9 reduced ~ _m_~J~ry F r o R ~ west

3. c 1850

NET LOSS N ( T 6Mm

. . . . . .

i A
NO SUPPLY
ll ii A" B*

9
9
9 nO shingle s u p p l y f ) ~ R
sys'(e~)
ton9 ~
the

n e t d r t P t 3" - - - - - 3
m n o r n e t * r t m s F m - D' - - ~ A "
west (If closed

Qnd A" ------)"

V c

9 n e t l o s s F r o ~ Ves"t ];each
4. 1990"s

N I T LI]SS NET LOS~

i I 9 long t m - R n e t d r l ~ * ] r - - ~ B
V
A A' B" 9 ~ "u-AnsFer ]l" ~ A"
NO SUPPLY C
- net loss Fron West Beach and
K a s t )eoA:h
9 onshore-oFFshore exchanges at
5, The Future ? (ost ~ch

NET LOSS NET LOSS

9 I o n o ~ e r ~ n e t d r l F ~ 3" - - - ~ ] t
& A' 3" - no ~ - o n s F e r s ~ r o ~ B" ~ A"
141] SUPPLY C
9 ne~t l o s s f r o n ~#est l i n c h and
[as* )eo~:h

Fig. 4. A schematic model of the evolution of coastal cells either side of the West Bay piers.
 SEDIMENT BUDGET ANALYSIS FOR COASTAL MANAGEMENT, WEST DORSET 187

Shingle budgets to 4 largely unfeasible. This depletion has been

compounded by beach mining operations and
A contemporary shingle budget for cells 1-4 is pre- thelack of sediment inputs from cell 2 because of
sented in Table 1. The inputs into the Lyme Bay shingle the mudslide lobes and boulder aprons below
transport cells are limited to cliff recession inputs, Doghouse Hill, and a rock bastion built at the
especially from Black Ven and Stonebarrow. Although west end of West Beach.
the net drift direction is eastwards, the presence of head- (v) The Future. The net eastward drift in cell 4 will
lands at Golden Cap and Doghouse Hill - Thorncombe lead to the long-term gradual depletion of East
Beacon restricts the longshore movement of shingle. Beach. Periods of westerly drift will no longer lead
Indeed, it is believed that there has been no shingle to the major build-up of material against the east
exchange between the cells since 1820-1850 (cells 2 and 3) pier and a time will arrive when the westward
and 1962 (cells 1 and 2). The net eastward drift of shingle transfer of material from cell 4 to cell 3 becomes
has been further restricted by the construction and unfeasible.
infilling of the West Bay piers, since between 1744 and
1820. Beach mining of gravels in all the cells would have
had a significant effect on the shingle budgets until
extraction finally ceased in 1987.
Potential impacts to Chesil Beach
The shingle budgets and transport models described
Shingle transport, West Bay above provided a framework for a preliminary assess-
ment of the potential impacts on coastal defence
Figure 4 provides a schematic representation of the improvements at West Bay on Chesil Beach. It was
sediment dynamics in cells 3 and 4 highlighting the concluded that Chesil Beach is a relict feature with no
changes that have occurred since the construction of significant shingle exchanges with West Beach. It is con-
the West Bay piers in 1744. The 'stages' in this evolu- sidered unlikely, therefore, that coastal defence improve-
tionary model are as follows. ment works in cell 3 would have a significant impact on
the sediment budget and morphology of Chesil Beach.
(i) Pre-1750. A shingle bar across the mouth of the Such works could, however, have significant local im-
piers allows the free transfer of material between pacts which would need to be addressed when consider-
cells 3 and 4. The continuous shingle beach which ing the suitability of different scheme options.
probably from Lyme Regis to Portland at this time
ensures that the system is open, receiving shingle
supplies from the eroding cliffs of East Devon
(until The Cobb was connected to the foreshore in References
1756), Black Ven and Stonebarrow (in westerly
BRAY, M. J. 1996. Beach budget analysis and shingle transport
storms) and from Chesil (in easterly storms). dynamics in West Dorset. PhD Thesis, LSE, London.
(ii) Around 1820. The piers were infilled, further CARR,A. P. 1985. Gravel extraction at Cogden Beach. Report to
restricting the transfer of material between cells 3 Dorset County Council, evidence submitted to the Cogden
and 4. However, because of the shingle bar at the Beach Public Planning Inquiry.
pier mouth and the high beach levels in cell 3 there CARR, A. P. & BLACKLEY,M. W. L. 1973. Investigations bear-
remains a net loss of material from cell 3 around ing on the age and development of Chesil Beach, Dorset
the pier heads to cell 4 because of the predominant and the associated area. Transactions of the Institute of
easterly drift. British Geographers, 58, 99-111.
(iii) Around 1850. The loss of the shingle bar at the pier CARR, A. P. & BLACKLEY,M. W. L. 1974. Ideas on the origin
and development of Chesil Beach, Dorset. Proceedings of
mouth and the depletion of West Beach between Dorset Natural History and Archaeology Society, 95, 9-17.
1820 and 1850 reduces the potential for east- MINISTRYOF AGRICULTURE,FISHERIESANDFOOD. 1995. Shore-
ward exchange of sediment between the cells. This line Management Plans: a guide for operating authorities.
condition is exacerbated by the closure of the MAFF Publication PB2197.
shingle transport link between cells 2 and 3 with RENDEL GEOTECHMCS. 1995. Coastal Planning and Manage-
the development of the Doghouse H i l l - Thorn- ment: a review of earth science information needs. HMSO,
combe Beacon headland as a sediment barrier London.
between 1787 and 1850. However, westward ex- RENDEL GEOTECHNICS. 1996. Sediment Inputs from Rivers.
changes between cells 4 and 3 remain possible Unpublished Report to SCOPAC.
RENDEL GEOTECHNICS. 1997a. The Investigation and Manage-
because of high shingle levels on East Beach. ment of Soft Rock Cliffs in England and Wales. Report to
(iv) Late 20th century. Scour around the west pier and MAFF.
in front of the seawall (built in the late 19th RENDEL GEOTECHNICS. 1997b. West Bay Geomorphological
century) had led to severe depletion and set-back Study. Unpublished Report to West Dorset District
of West Beach making further transfers from cell 3 Council.
Geological Society, London, Engineering Geology Special Publications
Land use planning in unstable areas: Ventnor, Isle of Wight
E. M. Lee and R. Moore

Geological Society, London, Engineering Geology Special Publications 2001; v. 18; p. 189-192
doi:10.1144/GSL.ENG.2001.018.01.26

© 2001 Geological Society of London

Land use planning in unstable areas: Ventnor, Isle of Wight

E. M. Lee I & R. Moore 2

1Department of Marine Sciences and Coastal Management, University of Newcastle,
Newcastle-upon-Tyne, UK
2 Sir William Halcrow and Partners, Birmingham, U K

Purpose of the study of Gault Clay, then massive cherty sandstones (Upper
Greensand) and Chalk. Of particular note is the pres-
Recent U K Government guidance has emphasized the ence of thin argillaceous layers within the Sandrock of
need to take account of landslide problems in the land the Lower Greensand, which together with the Gault
use planning process (DoE 1990, 1996). To assist the Clay have a very important influence on the stability and
implementation of this policy, the then Department of hydrogeology of the area. The geological structure of the
the Environment (DOE) commissioned a number of dem- Undercliff is relatively simple, with the strata dipping
onstration projects to develop approaches to assess the seaward at around 1.5-2 ~ south-southeast. In addition,
potential for landsliding and to identify the best ways of a NNW-SSE trending synclinal structure is superim-
incorporating this information in the planning process. posed on the general dip.
In the U K there are many situations where historic
development has resulted in the concentration of urban
development and infrastructure on unstable ground.
Techniques used
This is often on a scale such that total avoidance or aban- The Ventnor study involved assessing the ground behav-
donment are out of the question, as is recourse to large- iour from field mapping and desk study sources alone.
scale and inordinately expensive engineering solutions The study was directed towards:
(e.g. the Bath area, Lyme Regis, the South Wales valleys,
etc. (Jones & Lee 1994)). Under these circumstances, (a) determining the nature and extent of the landslide
detailed knowledge of slope instability is required so that complex;
pragmatic policies can be developed to assist commu- (b) understanding the past behaviour of separate parts
nities to reduce risk. This approach has been pioneered of the landslide system;
by the detailed study of the Undercliff at Ventnor, on the (c) formulating a range of management strategies to
south coast of the Isle of Wight, England (Lee & Moore reduce the impact of future movement.
1991; Lee et al. 1991a, b, c; Moore et al. 1991). The methodology used in this study is shown in Figure 1
and highlights the importance of geomorphological
mapping to the whole project. The approach involved a
The study area thorough review of available records, reports and docu-
ments relating to instability, followed by a detailed field
The situation at Ventnor is unusual in that the whole investigation comprising geomorphological and geologi-
town lies within an ancient landslide complex. The cal mapping, photogrammetric analyses, a survey of
Undercliff has generally been subjected to slow, deep- structural damage caused by ground movement, a land
seated ground movements or creep with less frequent epi- use survey and a review of local building practice (Fig. 1).
sodic periods of more active movements. Consequently A search through historical documents, local newspapers
most of the developed area has been affected, resulting from 1855 to 1989, local authority records and published
in cumulative damage to buildings, roads and services scientific research revealed nearly 200 individual inci-
(Lee & Moore 1991; Rendel Geotechnics 1995). Over the dents of ground movement over the last two centuries.
last 100 years at least 50 buildings have had to be demol- The various forms of instability that have occurred are
ished in Ventnor due to ground movement. The planning summarized in Figure 2.
problems are, therefore, related to the control of develop- The results of these investigations provided an under-
ment in those parts of the town which have been shown standing of the nature and extent of the landslide
to be particularly susceptible to ground movement. system, together with the type, size and frequency of
The landslides within the Undercliff are developed in contemporary movements and their impact on the
Lower and Upper Cretaceous rocks. These consist of local community. This detailed understanding of ground
sandstones (Lower Greensand) overlain by over 40m behaviour was used, in conjunction with knowledge

From: GRIFFITHS,J. S. (ed.) Land Surface Evaluation for Engineering Practice. Geological Society, London, Engineering Geology
Special Publications, 18, 189-192. 0267-9914/01/$15.00 9 The Geological Society of London 2001.
190 E. M. LEE & R. MOORE

* From nearby s/tes

I Review of available records,
reports and documents
' I
T

1
. . . . . .

ofbydamage
groundl [ Suwey ofcurrent
II==I
Survey
investigation* mapping land use
movement
i

v ~
L-. I

of landslide
systems
l
J
I

contemporary
movement
contemporary
movement
conlemporary
movemefll
1
II I1
I Impact of
contemporary
movement
Nature of land
use at dsk
J structures to
] ground
L,,rnovement

Information d

System

J I
Factors Factors
influencing influencing
the the frequency
distribution of of
contemporary contemporary
movements movements
i

Ground
Behavlour
Map

I management r
strategies
L
Fig. 1. The programme of work for the Ventnor study (from Lee & Moore 1991).

of the vulnerability to movement of different types of 9 in Upper Ventnor, a graben-like feature occurs land-
construction and the spatial distribution of property at ward of the zone of multiple rotational slides, compris-
risk, to formulate a range of management strategies e ing a 20 m wide subsiding block bounded by parallel
designed to reduce the impact of future movements. fissures, and extends parallel to the coast for over
Geomorphological mapping at 1:2500 scale revealed 500m: this unit exhibits the most serious ground
the extent and complexity of the landslides. The Under- movements recently experienced in the town;
cliff lies immediately below the Chalk Downs and Upper 9 mudslides have developed on the coast where dis-
Greensand bench. From the surface evidence and sub- placed Gault Clay is exposed.
surface work elsewhere in the Undercliff (e.g. Hutch-
Once the framework of landslide units had been estab-
inson et al. 1991) the following main features were
lished, it was possible to relate building damage and
distinguished:
movement rates to units with known dimensions.
9 a sequence of compound slides which occupy a zone of
similar breadth in the lower part of the Undercliff;
9 multiple rotational slides occupy a broad zone in the Ground behaviour mapping
upper parts of the Undercliff, giving rise to linear
benches separated by intermediate scarps: these units One of the main problems with many landslide hazard
mainly comprise back-tilted blocks of Upper Green- assessments is that subjective comparisons are made
sand and Chalk; between magnitude and frequency of different processes
 LAND USE PLANNING IN UNSTABLE AREAS: VENTNOR, ISLE OF WIGHT 191

Fig. 2. Types of contemporary ground movement in the Ventnor Undercliff (from Rendel Geotechnics 1995).

within the confines of a simple scale of hazard. For 9 the nature and extent of individual landslide units
example, Chandler & Hutchinson (1984) devised a pre- which together form the mosaic of landslide features
liminary zonation for Ventnor, using four classes of known as the Undercliff at Ventnor;
probability (negligible, low, moderate and large) of future 9 the different landslide processes which have operated
movements. The use of such arbitrary subjective scales within the town over the last 200 years;
can cause serious difficulties, as perceptions of what 9 the location of ground movement events recorded in
actually constitutes a 'high' probability or a 'small' prob- the last 200 years;
ability will vary considerably. This can lead to misunder- 9 the recorded rates of ground movement, over the last
standings and unnecessary alarm amongst those the maps 30-100 years;
are intended to help, namely the general public. For these 9 the severity of damage to property caused by ground
reasons the Ventnor study attempted to avoid subjective movement;
scales and has concentrated on analysing and present- 9 the causes of damage to property as a result of ground
ing the wealth of available information on movements movement;
within the town in a factual and objective manner. 9 the relationship between past landslide events and
Understanding the geomorphology of the landslide antecedent rainfall.
complex at Ventnor proved to be the key to under-
standing the nature and pattern of contemporary move- The ground behaviour map demonstrated that the
ments. This understanding was used to compile a 1 : 2500 problems resulting from ground movement vary from
scale ground behaviour map. This map summarizes the place to place according to the geomorphological setting.
nature, magnitude and frequency of contemporary pro- This formed the basis for landslide management strate-
cesses and their impact on the local community, being a gies that can be applied within the context of a zoning
synthesis of the following information: framework that reflects the variations in stability rather
192 E. M. LEE & R. MOORE

than a blanket approach to the problem. In support References

of the management strategy a l:2500 scale planning
guidance map was produced which related categories of CHANDLER,M. P. & HUTCHINSON,J. N. 1984. Assessment of
ground behaviour to forward planning and development relative slide hazard within a large, pre-existing coastal
control. The map indicated that different areas of the landslide at Ventnor, Isle of Wight. In: Proceedings of the
landslide complex need to be treated in different ways for IVth International Symposium of Landslides, Toronto, 2,
both policy formulation and the review of planning 517-522.
DEPARTMENT OF THE ENVIRONMENT. 1990. Planning Policy
applications. Areas were recognized which are likely to Guidance PPG14: Development on Unstable Land. HMSO,
be suitable for developments, along with areas which London.
are either subject to significant constraints or mostly DEPARTMENTOFTHEENVIRONMENT.1996. Planning Policy Guid-
unsuitable. ance note 14 (Annex 1): Landslides and Planning. HMSO,
London.
HUTCHINSON, J. N., BROMHEAD,E. N. & CHANDLER,M. P.
1991. Investigations of landslides at St Catherine's Point,
Landslide management Isle of Wight. In: CHANDLER, R. J. (ed.) Slope Stability
Engineering: Developments and Applications. Thomas Tel-
Since the publication of the study the local authority has ford, London, 169-179.
adopted an Undercliff Landslide Management Strategy, JONES, D. K. C. & LEE, E. M. 1994. Landsliding in Great
the objectives of which are: Britain. HMSO, London.
LEE, E. M. & MOORE, R. 1991. Coastal landslip potential
9 to reduce the likelihood of future movement by con-
assessment: Isle of Wight Undercliff, Ventnor. Department
trolling the factors (both natural and man-induced) of the Environment, London.
that cause ground movements; LEE, E. M., DOORNKAMP,J. C., BRUNSDEN,D. & NOTON, N. H.
9 to limit the impact of future movement through 1991a. Ground movement in Ventnor, Isle of Wight. Depart-
the adoption of appropriate planning and building ment of the Environment, London.
controls. LEE E. M., MOORE, R., BRUNSDEN,D. & SIDDLE,H. J. 1991b.
The assessment of ground behaviour at Ventnor, Isle of
The study identified that considerable benefit can be Wight. In: CHANDLER,R. J. (ed.) Slope Stability Engineer-
gained by reducing the frequency and magnitude of ing: Developments and Applications. Thomas Telford, Lon-
ground movement events through engineering works don, 207-212.
designed to improve the stability of the landslide system. LEE, E. M., MOORE, R., BURT, N. & BRUNSDEN,D. 1991c.
This work would best be directed towards coastal protec- Strategies for managing the landslide complex at Ventnor,
tion to prevent toe erosion or unloading of the landslide Isle of Wight. In: CHANDLER, R. J. (ed.) Slope Stability
complex, groundwater management and improving the Engineering: Development and Applications. Thomas Tel-
infrastructure throughout the Undercliff (some struc- ford, London, 219-225.
MOORE, R., LEE, E. M. & NOTON, N. H. 1991. The distribution,
tures have been poorly maintained and neglected to the frequency and magnitude of ground movements at Vent-
extent that they have become dangerous) and thereby nor, Isle of Wight. In: CHANDLER,R. J. (ed.) Slope Stability
improve confidence in the area. Potentially the most Engineering: Development and Applications. Thomas Tel-
serious destabilizing factor associated with development ford, London, 231-236.
in the Undercliff has been the artificial surcharge of RENDELGEOTECHNICS.1995. The Undercliff of the Isle of Wight:
groundwater from septic tanks, leaking water pipes, a review ofgroundbehaviour. South Wight Borough Council.
drains, sewers and swimming pools.
Geological Society, London, Engineering Geology Special Publications
Subsidence map development in an area of abandoned salt mines
E. M. Lee and C. F. Sakalas

Geological Society, London, Engineering Geology Special Publications 2001; v. 18; p. 193-195
doi:10.1144/GSL.ENG.2001.018.01.27

© 2001 Geological Society of London

Subsidence map development in an area of abandoned salt mines

E. M. Lee 1 & C. F. Sakalas 2

~Department of Marine Sciences and Coastal Management, University of Newcastle,
Newcastle-upon-Tyne, U K
2 High-Point Rendel, Birmingham, U K

Purpose of the study was very much a low-cost affair. It was not practicable
to commence with a conventional borehole investiga-
Morphological mapping is not generally associated with tion. The approach to investigating the site was based
mining and subsidence, having traditionally been used to upon a number of pragmatic assumptions:
define the nature and extent of surface features such as 1. at the time of abandonment all the underground
landslides and landforms. However, detailed morpholo- workings were unstable: if they had not already
gical mapping, linked to subsurface investigation, can collapsed, they may be expected to collapse at some
provide a preliminary indication of potential subsidence time in the future;
hazards in some areas of abandoned mine workings. 2. the degree of risk and liability across the area will
This example describes how mapping was carried out in vary with the nature of current and future land uses
an undeveloped part of the Cheshire saltfield, prior to above the workings and the nature of the works
any subsurface investigation, with the aim of establish- themselves;
ing the likely extent to which a number of mines had 3. management responses should reflect the variation
collapsed. in risk across the area, rather than adopting a
blanket approach.

The site
The site is in the Northwich area of Cheshire, England, Techniques used
and has had a long history of rock salt mining and brine
pumping. The site is largely undeveloped, although The project involved a comprehensive desk study review
locally important infrastructure routes cross the area, of the mining and subsidence history of the site, com-
and it has potential for use as a community woodland parison of benchmark levels on different editions of
and open space area. 1:2500 scale Ordnance Survey maps and a survey of
At the site, rock salt (halite) occurs in two discrete structural damage caused by subsidence. Surface mor-
horizons within the Triassic Mercia Mudstone sequence, phological mapping was undertaken at a scale of 1: 1250
separated by around 10m of marls: the 20-30 m thick using a tape, compass and hand-held clinometer. The
Top Bed (at c. 40 m b.g.1.) and the 23-28 m thick Bottom survey was directed towards identifying the subtle (and,
Bed (at c. 75m b.g.1.). This sequence is overlain by in some cases, not so subtle) changes in slope angle and
glacial till. Salt was extensively mined at the site from direction that might indicate the spatial extent of mine
around 1777 to 1933, with the one Top Bed and seven collapse. It quickly became apparent that the technique
Bottom Bed mines opened over this period, varying in was very effective in recording how the natural topog-
size from 0.8 to 13ha. Since abandonment there have raphy changes in form and begins to 'bend' inwards
been both dramatic mine collapses and gradual subsi- around abandoned mine workings.
dence, with some mines believed to have remained open The morphological mapping demonstrated that many
or partially open. Details of the mining and subsidence of the smaller mines had produced no surface signs of
history of the Cheshire area can be found in Calvert collapse. In addition, a number of larger mines that were
(1915) and Wharmby (1987). believed to have completely collapsed were shown to be,
In response to a potential purchaser's need to estab- in places, only partly collapsed by comparing the sur-
lish the risks and liabilities associated with future mine face morphology with the historical mine plans. On this
collapse and subsidence, a preliminary stability assess- basis it was possible to recognize four broad groups of
ment was undertaken. This stage of the investigation mined areas:

From: GRIFFITHS, J. S. (ed.) Land Surface Evaluationfor Engineering Practice. Geological Society, London, Engineering Geology
Special Publications, 18, 193-195. 0267-9914/01/$15.00 9 The Geological Society of London 2001.
c~

o
 SUBSIDENCE MAP DEVELOPMENT IN AN AREA OF ABANDONED SALT 195

1. undermined areas which have no surface evidence of 3. natural slopes: gently sloping land beyond the area
collapse: the ground above these mines is character- of surface subsidence;
istically gently sloping southwards; 4. stream valley slopes: incised side slopes to stream
2. undermined areas which have surface evidence of channels;
partial collapse: these areas are characterized by 5. areas with either steep stream valley slopes or sig-
broad, enclosed depressions which go against the nificant surface subsidence.
'grain' of the natural topography;
Areas which have been affected by mine collapse may
3. undermined areas which have surface evidence of
not necessarily be free of subsidence problems in the
complete collapse, e.g. the major crater which marks
future, as these mines may have only partially collapsed.
the site of the former Top Bed mine;
4. possible collapsed early pits used for extracting
rock-salt from the Top Bed; small enclosed depres-
sions on the northern side of the site. Subsurface investigation
The contemporary subsidence map, based largely on
desk study and morphological mapping, provided a use-
Contemporary subsidence map ful preliminary indication of the likely subsidence risks at
the site. On this basis the client considered it worthwhile
A contemporary subsidence map was prepared (Fig. 1) proceeding further with the plans. It was recognized that
which draws together the varied information about the the potential for future subsidence can only be assessed
subsidence history of the site, i.e.: through subsurface information about present mine con-
ditions and, hence, a further stage of work was carried
9 recorded evidence of subsidence
out. The subsidence map provided a framework for
9 ground movement measurements
planning the subsurface investigation (including bore-
9 building damage holes and ultrasonic surveys). The investigation largely
9 morphological mapping
confirmed the general picture inferred from the surface
9 mine abandonment plans
mapping, suggesting that the morphological mapping
Bearing in mind that the whole of the area is likely to have technique could be a useful preliminary stage in mine
been affected by minor ground movements associated subsidence studies in other settings.
with natural subsidence due to natural salt solution or
brine pumping, five main land classes were recognized on
the basis of the pattern of past mining activity and References
subsequent ground movement:
CALVERT, A. F. 1915. Salt in Cheshire.
1. areas with significant to severe surface subsidence; WHARMBY,P. 1987. A report on the rock-salt mines and brine
2. areas with significant surface subsidence; shafts in the Cheshire saltfield. Cheshire County Council.
Geological Society, London, Engineering Geology Special Publications
The design of remedial works to the Dharan-Dhankuta Road, East Nepal
R. P. Martin

Geological Society, London, Engineering Geology Special Publications 2001; v. 18; p. 197-204
doi:10.1144/GSL.ENG.2001.018.01.28

© 2001 Geological Society of London

The design of remedial works to the Dharan-Dhankuta Road,
East Nepal

R. P. Martin
Geotechnical Engineering Office, Civil Engineering Department, Government of Hong Kong Special
Administration Region, China

Purpose of mapping The road was affected by three main types of earth
quake damage:
Parts of the Dharan-Dhankuta Road in East Nepal (i) widespread instability in roadside cut slopes and
were seriously damaged by the Nepal earthquake of natural rock slopes, especially in the brittle frac-
21 August 1988 and subsequently by a locally intense tured quartzite, ranging from small rockfalls of a
monsoon storm on 12 September 1988. Detailed field few cubic metres to debris slides and rockslides up
mapping was carried out at several of the worst-affected to several thousand cubic metres;
hill sections of the road. The purpose of the mapping (ii) extensive cracking in natural slopes mantled with
was to establish the nature and local extent of earth- colluvial soils overlying fractured rock;
quake, and subsequent landslide, rockfall and erosion (iii) significant displacement and occasional complete
damage as a prerequisite for the design of road remedial failure of retaining structures supporting the road
works. A full account of the work is given by Roughton formation and cut slopes above the formation.
& Partners (1988). This is an example of post-construc-
tion stage land surface evaluation using the five-stage Although not exceptional for the area, the subsequent
classification of mountain road projects proposed by monsoon storm triggered further landsliding on slopes
Fookes et al. (1985). already weakened by the earthquake and caused numer-
ous blockages of culverts and other drainage structures.
Emergency clearance operations re-established four-
The site wheel drive, single-lane access along the road within
four to six days of the earthquake and storm events.
The Dharan-Dhankuta Road is a 50 km long mountain Two-lane access for lorries and buses was re-established
road traversing the first two ranges of the Low Hima- along most of the road by mid-October 1988, following
laya of East Nepal (Fig. 1). It was built between 1977 the removal of c. 65000m 3 of landslide and erosion
and 1982 under the U K Government's overseas aid debris. However, the most heavily damaged section of
programme to the Kingdom of Nepal (Cross 1982). the road, which included two areas of complete failure
The epicentre of the earthquake on 21 August 1988 of the road formation, remained closed to full traffic for
was in the Udayapur District, about 65 km to the west of several months.
Dharan (Fig. 1). The focal depth was c. 60 km and the
surface wave magnitude (Ms) was 6.6 (Dikshit 1991).
Three weeks after the earthquake, heavy rain fell over
part of the site during monsoon thunderstorms. The Techniques used
24 hour rainfall recorded in Dharan was c. 160 mm, most
of which fell within a few hours. This case study describes the mapping and checklist
techniques used for the design of remedial works in the
most heavily damaged section located at Karkichhap,
The problem north of the Sangure Ridge (Fig. 1). The key sources of
information used for the field mapping were copies
Serious earthquake damage to the road was confined of 1 : 500 scale original road alignment plans and 1 : 100
to the first 26 km from Dharan where the route climbs scale cross-section drawings produced by the design engi-
over the Sangure Ridge. This is a range of hills up to neers, together with the relevant design reports. As-built
1500 m in elevation formed in weathered Tertiary Siwalik record drawings (elevations and cross-sections) at similar
sandstones/siltstones and Mesozoic quartzites, meta- scales were also consulted where available, but these were
sandstones and phyllites of the Sangure Series (Fig. 1). not complete. Also valuable were descriptions of the

From: GRIFFITHS,J. S. (ed.) Land Surface Evaluation for Engineering Practice. Geological Society, London, Engineering Geology
Special Publications, 18, 197-204. 0267-9914/01/$15.00 9 The Geological Society of London 2001.
198 R. P. MARTIN

\ ,

. - /i "~'~~ 1 " ~ NKUTA /

/,

..... ~ "/e '~B~Guth;Tar 5Ok9 " ~. "7~

\

'~ ~ I~ ........... 914

V~ f-
\ /
%\
914 \
soe I \
\
7~24

/
Dhara~ani " //
/
L

\
\
? Karkichhap ' 'y~.
~ m ~ ~ 15km
1219
.~,~11 Bhedetar
0
Chhoti Marang tl N ~
914

Shivadhara 1219

9~A

~ Everest Kan~~"inga
L.,,~,,ju _~

DHARAN
N 9 Kathmandu i S.~IKKIMM
NEPAL
Epicentre of 9 Dhankuta
earthquake x 9 Dharan t
-I- Ot 1 2 3km
i __~L~ .............J
INDIA ~/
LEGEND

Road alignmentand 5km chainage marks 9 Village or districtname

914 " ' Contour (metres) ~ Town area

Rivers

Fig. 1. Location of the Dharan-Dhankuta Road in East Nepal.

 THE DESIGN OF REMEDIAL WORKS TO THE DHARAN-DHANKUTA ROAD, EAST NEPAL 199

LEGEND
20+100 Road chainage in km+m -~-~---r--c~ Short steep slope or bluff ~--x--~ In situ rock OLItcmp

Gabion retaining wall or revetment Convex break of slope Local slope angle in degrees

Crest of cut slope or bare scar LS Landslip scar

D o ~~ Masonry retaining wall or revetment

100 300H Crack - numbers refer to approximate width, tote Contour in metres ND No significant alL,tress observed
200D depth (D) or height of bluff (H) in mm Outward displacament
TP Trial Pit 500 of wall crest (ram)
Note: This map extract reduced; original mapping at t : 1 000 scale

Fig. 2. Extract of field map of damaged section of road at Karkichhap.

natural terrain on geomorphological sketch plans, sec- checklist incorporated additional sketch mapping at a
tions and pro formas compiled during a reconnaissance- typical scale of 1:200.
stage land surface evaluation of the provisional road Three hand-excavated trial pits were dug along the
alignment made two years prior to the start of construc- line of the former road edge in the failed section at
tion (Brunsden et al. 1981). Ch 19-+-240-19+280, to provide information on suit-
Field mapping was carried out using 1:1000 scale, or able founding depths for reconstructed retaining walls
reduced 1:500 scale, composite road alignment plans (Fig. 2). These were the only subsurface investigations
with a 2 m contour interval. The mapping included selec- used for the remedial works design. Use of a drilling rig
ted elements of geomorphology, surficial geology and was considered but was ruled out because of difficulties
damage to the road formation and retaining structures. of procurement and access within the works design
Figure 2 is an extract of such a field map in the centre of programme, and the questionable value of point data in
the heavily damaged section of the alignment. Figure 3 such an extensively disturbed area.
shows part of the same area prior to the earthquake and
storm damage (A), during construction of remedial
works (B) and after completion of the works (C). The Conceptual models
mapping was complemented by the use of a standar-
dized checklist (pro forma) to record details of damage Analysis of the results of the field mapping and check-
to the distressed retaining structures (Fig. 4). The lists allowed conceptual models to be developed for two
200 R . P . MARTIN

types of slope failure and a mechanism of progressive

displacement and collapse of masonry retaining walls.
(A) 1. Rockfall and rockslides in quartzite cliffs. The
earthquake dislodged many individual rock blocks and
generated rockslides from two massive quartzite pinna-
cles in between the hairpinning roadline. The mass of the
largest fallen block was estimated at 300 tonnes. Two
sections of roadline supported by gabion retaining walls
up to 7 m high were largely destroyed by the rockfall im-
pact (Figs 2 and 5). This type of failure was not repeated
in the monsoon storm.
2. Cracking and displacement of soil slopes (natural and
cut slopes). Surface cracking of colluvium and intensely
weathered phyllites was visible throughout the area, often
associated with shallow debris slides (Fig. 2). Individ-
ual cracks were typically 50-300 mm wide, 100-300 mm
deep and showed relative vertical displacement of up to
(B) 500 mm. The cracks varied in length from a few metres
up to a maximum of 50m (commonly 5-20m) and in
plan shape from straight to sharply concave downslope.
Most were aligned parallel or subparallel to the slope
contours. The intensity of cracking varied significantly,
with the worst distress concentrated in two areas: above
Ch 19 + 100-19 + 210 and in between the roadlines from
Ch 19 + 320-19 + 430 and 19 + 750-19 + 850 (Fig. 2).
Slopes which suffered the most intense cracking during
the earthquake were noticeably the sites of renewed
displacement or shallow landsliding during the monsoon
storm.
3. Displacement and failure of retaining walls. Both
masonry and gabion gravity retaining walls, up to 7 m
and 10 m high respectively, had been constructed to sup-
(c) port the road formation in this area (Fig. 2). None of the
gabion walls failed in the earthquake. Most experienced
relatively minor movement and settlement (generally
<200 mm), with slight to moderate consequential crack-
ing of the road pavement. In contrast, two sections of
masonry wall failed completely by rotation and most of
the other existing masonry walls between Ch 18 + 800-
19+400 were displaced outwards by 0.5 to 3.0m
(Fig. 6). Based on mapping of these displacements and
cracks within the pavement and backfill, the inferred
mechanism of wall collapse is shown in Figure 7. In gen-
eral, the worst-affected walls were not the highest but
were those with minimum toe embedment ( < l . 5 m ) as
deduced from site observations and as-built wall eleva-
Fig. 3. Upper (Ch 19 + 250-19 + 350) and lower (Ch 19 + 800- tion drawings.
20 + 000) roadlines (A) prior to the earthquake and storm
damage (June 1988), (B) during construction of remedial works
(May 1989) and (C) following completion of the works
(October 1991). In (B) the failed section of retaining wall
(see Fig. 6A) has been rebuilt along the upper line while What the mapping established
landslide debris still covers part of the lower line. In (C) the
slope between the roadlines has been reinstated with The conceptual models were used as a framework for
bioengineering works (terracing and grass/tree planting above a assessing the likelihood of renewed instability or con-
gabion revetment at the toe). tinuing displacement of slopes and walls and for making
engineering judgements for the design of permanent
remedial works at different sections along the alignment.
THE DESIGN OF REMEDIAL WORKS TO THE D H A R A N - D H A N K U T A ROAD, EAST NEPAL 201

CHECKLIST FOR DISTRESSED RETAINING WALLS SKETCH PLAN (NOMINALLY1:200)

I AVERAGEROAD
Ch. Location Wall type Ref. No. Logged by Date WIDTH
17 + 185 - 17 + 235 Masonry RPM 14/10/88
Exposed height over distressed length (m)
;~ Max 4.5 Min 1.5 Critical section 4.2
Plan shape of distressed length
Straight ~ Combination

Culvert p r e s ~ Type ~_. 1x 1.2 gfAnnco

/
Total length of distressed section (m) | 50
No. Max hor disp (ram) Cumulative Max vert disp (mm) Cumulative
CracksCrest5 c. 100 c. 150 100 100
Combined width of soil filled cracks c.250
Face No. Max width (ram) Av Z from horiz (~ Other face damage:
17+200 ~ -CuLvert=____~
~ Cracks 0 -- -- * Y'/N
=~ In plan? Max bulge (mm) In section? Max bulge (mm)
0,3 .~
Displacement ~ }
in planalong " ~
/t,
m Height of max bulge in section above toe (m) construction
joint
Significant? Position of crest with respect to toe
Tilting ~rTrwo main//
Y/N" Less than vert ,/ Near vert Overhanging Cut slope
Wall i (upslopeside)
Monitoring Available? Date started Movement (downslope
side)
Results* y 2/9/88 Stopped? I Continuing? J
Previous seepage signs? * Seepage through wall? * l )~"/N " ]// t, /1
C. 17+213
~ Probable foundation material: Very closely jointed phyllite Hatchedarea
hassignificant//
~) Av. ground slope angle below wall ~ 41 ~ J pavement/
settlement
Signs of distress in ground below toe? ~t'/N Upto 100mrn
vertical
REMARKS * = Items to be emphasised, where relevant displacement
Section up-chainage of Ch 17 + 203 shows I 10mm forward movement at crest and w.r.t,
of top
pavement I
30mm at toe. No cracks observed in ground below toe, but ground displaced of wall
downwards from toe by average 400mm between Ch 17 + 210 - 17 + 230 (fresh
masonry exposed). i

Forward tilting along one construction joint is main cause of apparent bulge along the
crest. Wall not considered to have bulged significantly.

Average face batter of wall 79 ~- 83 ~

Critical X-Section (1:200) Dwg. Nos. Used SK/84A

CH. 1 7 + 2 1 3 ~ ~ ~ q ~ 17+235 / I )avement

/\ CetOe0
nte

Fig. 4. Example of checklist used for distressed retaining walls.

202 R. P. MARTIN

(A)

Fig. 5. Two sections of roadline between Ch 20 + 100 and

20 + 200 destroyed by rockfalls from the quartzite pinnacles.
These sections were rebuilt with new gabion retaining walls.

1. Rockfalls and rockslides. Almost all of the damaging

rockfalls and rockslides were caused by the earthquake.
The potential for renewed rockfalls or slides, or erosion (B)
of existing slide scars, was judged to be low. Hence the
principles adopted for remedial works design were: (i) to
rebuild the damaged sections of road with essentially the
same cross-sections; and (ii) not to attempt to construct
stabilization or improvement works to the existing rock
slopes. (Full stabilization of the quartzite pinnacles was
beyond the scope of reasonable improvement works and
it was recommended that future seismically generated
rockfalls should be an accepted risk.)
2. Cracked and displaced soil slopes. The crucial
observation on these slopes was that cracking and the
occurrence of debris slides appeared to be confined to the
top 2-3 m of colluvial soil. Underlying fractured phyl-
lites, where exposed, showed some evidence of loosening
in the earthquake but there were no very large tension
scars on the slopes or other signs of incipient deep-seated
movement. The probability of a major hillside failure,
deep enough to destroy both sections of road, was judged
to be very low. Hence remedial works designs were based
on: (i) excavating and removing cracked material to re-
instate the cut slopes, or providing structural support in
the form of gabion revetments wherever the angles of
reinstated cut slopes would be greater than 50~ and
(ii) grading over and sealing with compacted soil the
larger cracks not removed by reinstated earthworks, to Fig. 6. The failed section of roadline (Ch 19 + 240-19 + 280)
minimize infiltration in future storms. (A) shortly after the earthquake (October 1988) and
(B) during construction of the remedial works (May 1989).
3. Displaced and failed retaining walls. The response to In the foreground of (B) a new gabion retaining wall
the earthquake clearly demonstrated that the flexible has been constructed to replace the failed section of
gabion walls were superior to the masonry walls in their masonry; behind this the displaced section of masonry
ability to accommodate differential foundation settle- wall has been repaired by foundation underpinning and
ment and moderate seismic loads without serious distress buttressing.
 THE DESIGN OF REMEDIAL WORKS TO THE DHARAN-DHANKUTA ROAD, EAST NEPAL 203

I. PRE-EARTHQUAKE / \ / / /
/ ~ // / Edgeof pavement
Static condition
Pa assumed to be active 4~u~ e / -/J :~ P"/ / ~ Probableextent
Note shallow depth of embedment at toe
I ~V Foundationmostlyin
coUuvium2-3mthick
perpendicularto slope
II. ONE CYCLE OF GROUND MOTION

Wall displaced outwards

Initial cracking and settlement of backfill
Outward movement of wall > backfill .'. Pa reduced ///
Slope at toe displaced downwards and outwards
(prior to wall movement? due to wall movement?.)
Bearing capacity reduced at toe, outweighs effect of reduced bearing pressure
p
~.-t~- .- - .- - .~ .
Ill. SEVERAL CYCLES OF GROUND MOTION /~~1//
Further downwards and outwards displacement of ground at toe
Bearing capacity exceeded at toe causing local shear failure and rotation of wall / ! \~r /
Major tension zones developed in backfill: cracks widened and deepened, [ ,~ ~ \\ / /
settlement increased, signficant relative vertical displacement of pavement / ~ I//

IV. MANY CYCLES OF GROUND MOTION ~" / "\ / / /

A /
Wall fails by rotation about toe, may slide further downslope Jk / \~,--"/~,/
Backfill collapses into void f \ / //~\\ /~
Rapidloadingmayinitiatemoreextensiveshallowfailurein slopebelowtoe / /~~__ __~/ collapseofbackii,i

Fig. 7. Inferred mechanism of collapse of masonry retaining walls.

(Martin 1982). Remedial works designs were based on slopes, but this has been tolerated within the normal
the following guidelines: (i) the failed sections of masonry maintenance programme, as is the case for minor insta-
wall should be rebuilt in gabion with a minimum toe bility on other hill sections of the road.
embedment of 2 m; (ii) distressed (rotated) masonry walls
judged to be unserviceable should be removed and rebuilt
in gabion; (iii) distressed masonry walls judged to be
serviceable require substantial improvement to toe sta- Similar applications
bility by foundation underpinning and/or buttressing;
and (iv) distressed gabion walls generally require only Large-scale field mapping and use of checklists proved
partial backfill and pavement reconstruction. invaluable for appreciating the scale and nature of the
The mapping and checklist results were of particular damage in this study. The techniques were applied
value in guiding the judgement on serviceability for items rapidly and cheaply, allowing completion of remedial
3(ii) and 3(iii) above. Removal of displaced (but not works design drawings within eight weeks. This type of
completely rotated) pieces of masonry weighing up to rapid post-construction assessment has widespread appli-
several hundred tonnes was a major logistical constraint cation to low-cost mountain roads subjected to earth-
for remedial works design and construction. In the event, quake or rainstorm damage.
only three sections of severely distressed wall, of com- The use of post-construction field mapping on the
bined length 72 m, were judged completely unserviceable Dharan-Dhankuta Road follows a long tradition of
and required demolition, removal and replacement. engineering geomorphological and geological mapping
Construction of remedial works commenced in early applied at the reconnaissance, investigation and con-
1989 and was substantially completed by mid-1990. struction stages of the project (Brunsden et al. 1975;
The rebuilt and repaired walls have performed satisfac- Fookes et al. 1985). This case study reinforces the view
torily in the interim, with no undue further movements of Fookes et al. that field mapping in various forms is
experienced. Further minor instability and erosion has the fundamental tool for investigations of low-cost
occurred from time to time within the reinstated cut mountain roads.
204 R . P . MARTIN

Acknowledgements. The author was seconded to Messrs CROSS, W. K. 1982. Location and design of the Dharan-
Roughton International to undertake the work described. The Dhankuta low cost road in eastern Nepal. Proceedings of
support given by Mr A. H. Cutler, Mr A. Murphy and Mr M. the Institution of Civil Engineers, Part 1 72, 27-46.
Pearce is gratefully acknowledged. This paper is published with DIKSmT, A. M. 1991. Geological effects and intensity distribu-
the permission of the Department for International Develop- tion of the Udayapur (Nepal) earthquake of August 20
ment (U.K. Government), the Department of Roads of the 1988. Journal of the Nepal Geological Society, 7 (special
Kingdom of Nepal and Messrs Roughton International. issue), 1-17.
FOOKES, P. G., SWEENEY,M., MANBY,C. N. D. & MARTIN, R. P.
1985. Geological and geotechnical engineering aspects of
low-cost roads in mountainous terrain. Engineering Geol-
References ogy, 21, 1-152.
MARTIN, R. P. 1982. Discussion on 'Retaining Walls and
BRUNSDEN, D., DOORNKAMP, J. C., FOOKES, P. G., JONES, Basements'. Proceedings of the Seventh South-east Asian
D. K. C. & KELLY, J. M. H. 1975. Large scale geomor- Geoteehnical Conference, Hong Kong, 2, 279-280.
phological mapping and highway engineering design. ROUGHTON & PARTNERS. 1988. Nepal Roads Remedial Works
Quarterly Journal of Engineering Geology, 8, 227-253. Project: Dharan-Dhankuta Road 1988 Damage Report.
BRUNSDEN, D., JONES, D. K. C., MARTIN, R. P. & DOORN- Roughton & Partners International, Southampton, 2 vols
KAMP, J. C. 1981. The geomorphological character of part (unpublished).
of the Low Himalaya of Eastern Nepal. Zeitschrift fur
Geomorphologie, supplementband 37, 25-72.
Geological Society, London, Engineering Geology Special Publications
Terrain evaluation for military purposes: examples from the Balkans
C. P. Nathanail

Geological Society, London, Engineering Geology Special Publications 2001; v. 18; p. 205-208
doi:10.1144/GSL.ENG.2001.018.01.29

© 2001 Geological Society of London

Terrain evaluation for military purposes: examples from the Balkans

C. P. Nathanail
School of Chemical Environmental Engineering, Nottingham University, Nottingham, U K

Aim/purpose were applied to advise on the ease of installation and,

malicious, extraction of border marker posts.
Terrain has influenced military commanders and the
outcome of military operations since ancient times (Rose
& Nathanail 2000). The aim of terrain evaluation in Terrain evaluation in Bosnia Herzegovina
support of military operations revolves around gaining
U N and NATO forces have operated in Bosnia Herze-
maximum operational advantage from the ground.
govina since the early 1990s in a variety of humani-
In war this includes maximizing the mobility of your
tarian relief and peace keeping roles. The terrain, cli-
own forces and ensuring the survival of your troops and
mate and their interaction were major obstacles that
in turn denying both of these to the enemy's forces.
required considerable engineering and logistics efforts
Armed forces are also increasingly being used in peace
to overcome.
keeping or peace enforcing roles. In such cases the aim
Geological information was available for much of the
of terrain evaluation is to support the humanitarian
country. However a lot of it was old, at a small scale or
efforts of the military.
written in Serbian. Some of the geological maps had
The United Nations Protection Force (UNPROFOR)
been produced by Yugoslav geologists for the Waffen SS
was tasked with delivering humanitarian relief in Bosnia
during World War II. Anecdotal information indi-
Herzegovina (Fig. 1). British troops were based in four
cated that the local geologists had deliberately intro-
main compounds at Vitez (School and Garage), Tuzla,
duced errors into those maps. This was subsequently
Gornji Vakuf and Tomislavgrad. Terrain evaluation
confirmed during the reconnaissance of sites near Vitez
principles were used to advise on the feasibility of con-
in central Bosnia (Nathanail in prep).
structing water supply boreholes within the perimeter
wire of each compound to ensure a secure supply of
water independent of any of the warring factions. Techniques used
Later, the NATO Implementation Force (IFOR)
was tasked with ensuring that the provisions of the The principal techniques involved in providing advice
Dayton Peace Agreement were fully implemented were: literature review, vehicle based reconnaissance,
(Haynes et al. 1997). One of the concerns was to en- very limited (due to the security situation) walkover
sure that the boundary between warring factions, which surveys, topographic and geological map appreciation,
had been agreed on maps, was securely and accurately interviews with local experts and limited interpretation
delineated on the ground. Terrain evaluation principles of satellite imagery.

Fig.1. Position of Bosnia Herzegovina (after http://www, graphicmaps.com/graphic_maps.htm).

From: GRIFFITHS,J. S. (ed.) Land Surface Evaluation for Engineering Practice. Geological Society, London, Engineering Geology
Special Publications, 18, 205-208. 0267-9914/01/$15.00 9 The Geological Society of London 2001.
206 C. P. NATHANAIL

The literature review focused on obtaining geologi- Zadar and north of Dubrovnik (Fig. 2). The zone is
cal maps of Bosnia and relevant papers on specific dominated lithologically by limestone, other carbonates,
locations. Although publications in English were pre- and some Triassic evaporites.
ferred, Serbian and German references were also found The High Karst zone comprises white limestone
to be relevant. Although limited translation support was mountains which have been eroded into deep valleys
available, a good dictionary used by a terrain evaluation and caves. Much of the surface water has disappeared
specialist was found to be more efficient in interpreting underground and given rise to the terrain known as
the technical terminology used in say borehole logs or karst. The limestones are in places intercollated with
illustrations in papers. bauxites.
Flysch constitutes the main part of the narrow Bos-
nian zone between two major thrusts and represents a
Site description: trough that persisted at least from the beginning of
geology of Bosnia Herzegovina Jurassic until the late-Cretaceous (Ager 1980). Sarajevo
is built on the Bosnian zone flysch. The Jurassic succes-
Bosnia Herzegovina forms part of the Dinarids, in the sion generally comprises pelagic limestone with chert
Alpine mountain chain that marks the collision between passing up to radiolarites then into the thick Bosnian
the African and European continental plates. The flysch. Much of this zone and the Serbian zone discussed
Dinaric Alps comprise mainly carbonate rocks which below are concealed beneath post-orogenic Oligocene-
have been folded parallel to the coastline of former Miocene molasse.
Yugoslavia and subsequently uplifted and eroded. The The Serbian zone largely comprises ophiolites whose
former Yugoslavia contains many of the type localities dark colour give Montenegro its name. In places the
of karst features. Many geomorphological terms have pre-ophiolite continental crust comprising Palaeozoic
been derived from the Slovenian and Serbo-Croatian basement is exposed overlain by early Triassic red
languages such as doline; p o l j e - a linear depression; sandstone. Within Bosnia the ophiolite sequence is over-
p o n o r - a sink-hole; u v a l a - a large depression due to lain conformably by flysch which spans the Jurassic-
the coalescence of dolines; and jama. Cretaceous boundary and corresponds with the flysch
of the Bosnian zone to the SW. Ager (1980) reports that
this is one of the rare places in Europe where the direct
Tectonic and Palaeogeographic Zones relationship of ophiolites and flysch can be seen.
The Dinarids, named after Mount Dinara, comprise Dearman et aL (1989) recognized that Bosnia lies
eight tectonic and palaeogeographic zones (Fig. 2) within one of the tectonically active regions of the world
(Ager 1980). where tectonism would affect the distribution of the
The Dalmatian zone extends from a peninsula of Istria, principal types of weathering. The country is character-
on the Italian border, all the way down to Albania. The ized by complex mountain relief and extensive geody-
general northwesterly strike of the rocks is evidenced by namic processes. The lower and middle levels suffer
the extremely elongated islands off the Croatian coast at chemical weathering of the rocks resulting in the deposi-
tion of a thick clayey cover over bedrock. Abundant
rainfall and groundwater enhances the development of
landslides.
In the Dinarids, the carbonate series of the Dinarid
zone and the inner Dinarian syncline region are subject
to extensive karstification. Seismic activity is high with
the southeastern end of the region being the most active.
As the intra-montaine basin and some poljes lagged
behind during the general tectonic uplift of the Dinarids
in the Neogene, sediment that once formed part of a
larger mantel are preserved in them (e.g. the basin of
Sarajevo). The role of surface runoff in the evolution
and dissection of the relief is evident; the flysch and
molasse hills in eastern Serbia are less dissected than are
the western parts of Slovenia which receive more rain.

Sources of information
A desk study to examine geological maps, papers
Fig. 2. Main tectonic zones in the Dinarides (after Ager 1980). and reports was carried out. This allowed the general
 TERRAIN EVALUATION FOR MILITARY PURPOSES: EXAMPLES FROM THE BALKANS 207

terrain and geological framework of the country and fractions, was not going to be affected by the ease or
specific locations of interest to be determined - at least otherwise of marking it.
in outline. The Dayton Peace Agreement was initialled at Day-
Limited fieldwork was carried out at each site ton, Ohio on 21 November 1995, following some six
(Nathanail 1996, 1998). This amounted to a walkover weeks of negotiations. To support the talks, the US
survey within the perimeter fence of the U N compound. military mapping agency - the Defence Mapping Agency
In one location (Vitex School) a hand auger was used to (or D M A ) - deployed a team of mapping specialists
examine soils down to 2 m. In another, Tomislavgrad, a and equipment to Dayton. This team of approximately
small borrow pit some 2-3 km away from the compound 40 technicians literally worked around the clock for
was examined to provide some indication of the nature those six weeks, producing 'the map' and accompany-
of the soils in that particular valley. ing overlays that were to define the partition of Bosnia
Attempts to access locally held information proved Herzegovina. A digital terrain elevation model of the
variably successful. The offices of a quarry near the Vitez region onto which aerial photography of the whole of
School compound were visited however the site appeared Bosnia was superimposed was created.
to be in use as a H Q for one of the warring factions and The operator then superimposed vector line data over
no information was retrieved. A former employee of the this image and in particular the Inter-Entity Boundary
Local Authority in one area had logs of two public water Line between the former warring factions, or the IEBL
supply boreholes some hundreds of metres from the as it became known. The computer image appeared
Vitez Garage compound. H a n d coloured maps of near on one screen and a co-ordinated map display on the
surface materials, compiled principally for groundwater other. Wherever the operator flew and drew a line on
and aggregate resource assessment purposes, were made the aerial photography, it would automatically appear
available, but only for inspection, at the Tuzla offices of a on the map.
former public body. In the end the two sides even disagreed on the type of
marker they wanted. The C r o a t / M u s l i m Federation
wanted a ground mark for administrative purposes only,
as they saw the IEBL somewhat akin to a British county
The conceptual model - groundwater boundary. The Bosnian Serbs, however, seemed to be
resources using the former Berlin Wall as their model boundary.
Finally they ended up with a marker similar to those
At each of the five U N bases visited, a simple conceptual used in U K to mark gas and water p i p e l i n e s - a pole
model of the near surface geology and hydrogeology with a brightly coloured apex hat. Not surprisingly, this
was developed from the desk study information, ter- proved to be a collectors item for those who did not
rain appreciation and walkover survey (Nathanail 1998). want any form of boundary.
In each case the model focused on where groundwater
resources were likely to be present in sufficient quantities
and on potential threats to the quality of such water.
The results were used to support the decision to deploy What did the techniques reveal?
the British Army's well drilling team for a number of
Despite the limited time and information available and
successful tours of duty in the Balkans (Wye 1994).
restricted access to the ground, the use of terrain evalua-
tion principles enabled a recommendation that a well
drilling team be deployed from the U K to construct
The conceptual m o d e l - border abstractions wells at each of the U N bases where British
Troops were stationed. The team successfully con-
marker posts structed wells at each of the bases and then went on to
A map of surface soils was available. The requirement construct wells in many other bases for other armed
was to contribute to the selection of boundary marker forces of other nations.
posts to be installed along the agreed line between the Terrain advice was also provided to assist in the
various factions. The requirements were that posts be planning for the eventual marking of the negotiated
easy to install but difficult for belligerents to remove. boundaries. However, in this instance terrain issues were
The map legend was interpreted in terms of ease of secondary in the overall decision making process.
installation (e.g. loose sand, soft clay versus bedrock or
stiff clay) and ease of extraction (essentially the inverse
of installation). The optimum ground being one in Similar situations and logical applications
which it was moderately difficult to install but diffi-
cult to extract marker posts. Of course, the course of The techniques of rapid terrain evaluation on the
the boundary, being negotiated between the warring basis of insufficient information can be used in other
208 C. P. NATHANAIL

military situations, in support of humanitarian relief NATHANAIL, C. P. 1996. Environmental constraints on United
operations, post disaster rehabilitation (Nathanail & Nations operations in Bosnia Hercegovina. In: COULSON,
Nathanail 1998) and in preliminary conceptual design of M. G. & BALDWIN, H. (eds) Proceedings of the Interna-
most major projects. Despite the incomplete database, tional Symposium on the Environment and Defence, NATO
CCMS Report Number 211, North Atlantic Treaty Orga-
the benefits of early appreciation of the terrain outweigh nisation, Brussels, 257-263.
the relatively minor costs of staff time, information NATHANAIL, C. P. 1998. Hydrogeological Assessments Of
acquisition and travel related expenses to visit the site(s). United Nations Bases in Bosnia Hercegovina. In: UNDER-
WOOD, J. & GUTH, P. (eds). Military Geology in War and
Peace, Geological Society of America Reviews in Engi-
References neering GEOLOGY,XIII, 211-215.
NATHANAIL, C. P. & NATHANAIL,J. F. 1998. Mitigating geo-
AGER, D. V. 1980. The Geology of Europe. McGraw Hill, hazards affecting mountain roads in northeast Somaliland.
London. In: MAUND, J. G. & EDDLESTON, M. (eds) Geohazards in
DEARMAN, W. R., SERGEV, E. M. & SHIBAKOVA,V. S. 1989. Engineering Geology. Geological Society, London, Engi-
Engineering Geology of the Earth. Nauka Publishers, neering Geology Special Publications, 15, 231-237.
Moscow. ROSE, E. P. F. & NATHANAIL,C. P. 2000. Geology and War.
HAYNES, S., FAGG, A. & RIGBY,N. 1997. Mapping For Peace - Geological Society, London.
The Challenges of 250 Years Of Crisis Support, presenta- WVE, T. 1994. Well drilling in Bosnia: The Royal Engineers
tion at Nottingham Trent University. Journal, 108, 149-153.
Geological Society, London, Engineering Geology Special Publications
Hazard assessment in Eastern Taiwan
D. N. Petley

Geological Society, London, Engineering Geology Special Publications 2001; v. 18; p. 209-213
doi:10.1144/GSL.ENG.2001.018.01.30

© 2001 Geological Society of London

H a z a r d assessment in Eastern T a i w a n

D. N. Petley
Department of Geography, University of Durham, Durham, U K

Introduction as a result of the ongoing collision of the Philippine Sea

(oceanic) and Eurasian (continental) plates. Damaging,
The topography of Taiwan is dominated by the Central shallow seismic activity is frequent. The most seismically
Mountain Range, which has steep, densely vegetated active region is in the central portion of the east coast of
slopes, deeply incised valleys and fractured, unstable the island. However, the recent 'Chi-Chi' earthquake of
rock masses. Although most of the population is located 21 September 1999 occurred in the centre of Taiwan and
on the low-lying coastal plains, there is a need in this caused the loss of over 2100 lives and induced damage
rapidly developing island to construct transportation island-wide.
links through the mountains. Unfortunately, the steep
topography makes development difficult and there have
been many instances in which poor terrain evaluation Terrain evaluation in Taroko
has led to inappropriate engineering solutions. National Park
This paper outlines how landslide hazard assessment
is being used as part of the planning process for the Taroko National Park is located on the eastern margin
development of infrastructure along one such road of Taiwan 121~ ~ E and 24~ ~ N (Fig. 1). It is centred
through the mountains. upon the drainage basin of the Li-Wu River, with terrain
that varies from lowland plains to steep, rugged, densely
vegetated mountainsides. Morphologically, the drainage
The physical setting of Taiwan basin is dominated by Taroko Gorge, a spectacular
canyon up to 1000 m deep and 17 km long. The geology,
Taiwan is located on the eastern margin of the Eurasian which is highly complex, comprises marble, green schist,
continental plate (Fig. 1). The island, which is approxi- black schist, pelite, sandstone, shale, and isolated bodies
mately 385 km in length and 143 km in width, has formed of gneiss. The general structure is that of a complex fold
and thrust belt formed as a result of east-west com-
pression, but with a number of large faults and thrusts
120 E 125 E
9 displacing and off-setting the units (Petley et al. 1997).
Running east-west through the park is the only sub-
stantial road across the Central Mountain Range, the
EURASIAN
Central Cross Island Highway (CCIH). The road runs
PLATE
along the base of Taroko Gorge, positioned between
15 and 50m above the river-bed on a series of inter-
connecting ledges. Thereafter it is threaded along the
25"N valley sides of the Li-Wu River, gaining altitude towards
""~.~. "......................... Ryuku 0 .. ........
the west in order to climb over the high mountain passes.
Improvements to the highway, which have included road
Ta,chun~f..~-/.~..~-"-':". ~ ualien". widening and the construction of new sections, and the
] /': ~ :/./li "-. RyukuTrench
development of tourist infrastructure, have caused con-
~ ~adung ~ , siderable disturbance to the geomorphic system (Petley
~.':~ [ " ~) 7cma, 1998a, b).
"V :/r \ N55 W
Kentingl~ i ~'\ The study reported here was undertaken as part of the
planning process for Taroko National Park. In recent
ManUla ~ ~Luzon" years; considerable investment has been made in up-
Trench ~ ~rc grading the infrastructure. Unfortunately, the success of
t ~....
this construction has been quite variable, with several
Fig. 1. The physical and tectonic setting of Taiwan. areas being adversely affected by gullies, landslides and

From: GRIFFITHS,J. S. (ed.) Land Surface Evaluation for Engineering Practice. Geological Society, London, Engineering Geology
Special Publications, 18, 209-213. 0267-9914/01/$15.00 9 The Geological Society of London 2001.
210 D . N . PETLEY

river channels. As a consequence, the National Park to the west of Tien Hsiang and 1370m above sea level.
Headquarters has recently produced a comprehensive Structures at this site currently comprise:
planning zonation to ensure that all new construction
was undertaken in a manner that minimizes the hazard 9 a large stall constructed without permission on the
posed by natural processes. inside of the bend in the road, with a parking area in
The study reported here was undertaken for six front;
centres of population along the CCIH. A draft devel- 9 a number of other buildings on both sides of the road,
opment plan for each site had already been developed. some of which are used for trade and some of which
The aim of this study was to utilize geomorphological are dwellings: m a n y of the buildings have been built
terrain evaluation to assess the validity of the proposed illegally;
scheme. Thus, the aims of this analysis were: 9 a terraced agricultural area in the northwest of the
site, used for the cultivation of fruit trees.
9 to independently evaluate the geomorphology of the
study areas; The site is located on a narrow, gently sloping piece of
9 to establish the location of any potential geomorpho- land about 100m above the Hualu Hsi, a tributary of
logical hazards within the planning area and to assess the Wahel Hsi (Fig. 2). The river, situated to the east
the likely impact on the site; of the site, is deeply incised such that the slopes are
9 to re-examine and interpret hazards that had been steep, and they are densely forested. The slopes above
previously identified by National Park staff; the site to the west are also steep and densely vegetated,
9 to provide advice on the redesignation of areas in line reaching a maximum height of 2253 m at Cheuh Shan,
with the hazard assessment where applicable. about 1.5 km away. This slope is traversed by the CCIH,
and about 150 m above th~ site there is a hairpin bend in
the highway. The slopes above Hualushi are character-
Methodology ized by a number of gullies, many of which flow onto the
highway with no realistic means of controlling the water.
Each site was mapped using aerial photograph inter- In Hualushi itself, two large gullies cross the road, one
pretation based on two epochs of 1 : 50 000 scale, vertical, on the apex of the bend in the centre of the settlement
colour, stereoscopic air photographs. Detailed geomor- and one about 75 m uphill to the south.
phological mapping was undertaken at each location The development plan for the site is as follows
during site visits. Where possible, sketches were made (Fig. 3):
from a variety of oblique perspectives. Additional data
were obtained on the nature of the local climatic system, 9 a narrow corridor has been reserved for further devel-
and the frequency of occurrence of both typhoons, with opment of the highway;
their associated extreme precipitation events, and earth- 9 the stall and car park will be removed;
quakes. These data were supplemented by the acquisi- 9 small areas will be reserved for the development
tion of local information via an interpreter. 9 of administrative structures, tourist facilities and
Thus, the approach adopted was similar to that dwellings;
used by Brunsden et al. (1975): namely, to construct a 9 the remainder of the area is to be protected and kept
morphological map based on observed landforms, and for conservation
then to use this as a base for the interpretation of
morphogenesis and contemporary processes. In general the plan appears to be sympathetic to the
No attempt was made to undertake a risk assessment environment. There is no evidence at the site of sub-
as the vulnerability of different planning elements was stantive mass movements or that seismic activity would
not considered (e.g. Dearman 1991; Brabb 1984). How- trigger major problems. However, the failure to ade-
ever, the study did give an indication of areas in which quately consider the two large gully systems that cross
natural hazards have previously occurred and/or are the road in the vicinity of the township provides con-
likely to occur in the future, the nature of that hazard, its siderable concern. The gully that crosses the highway on
magnitude and frequency of occurrence. It provides the the apex of the bend in the centre of the town provides
basic information that, combined with a vulnerability the greatest cause for worry. This is a large feature (up to
map, would allow a risk assessment map to be produced. 5 m wide) draining a significant area of steep land above.
In addition, it allows the National Park H Q to identify Parts of this area have been developed for agriculture,
areas to which further resources should be directed. whilst the remainder is forested. In 1997, flow in this
gully resulting from the high precipitation intensities
associated with the passage of Typhoon Amber over-
Case study: Hualushi topped the drainage culvert. Flow was diverted through
the town, causing considerable damage to buildings.
Hualushi is a small farming community situated on a Closer examination of the drainage basin for this gully
sharp bend in the highway, about 21 km along the road casts light on the nature of the problem at this site.
 H A Z A R D ASSESSMENT IN EASTERN TAIWAN 211

v Major concave break of slope

9 Major convex break of slope N
# Ridge
" ' - - ~ ' ~ Gully
A Scale
0 metres 50
Scarp or cutting
Direction of slope
40 Indicative slope angle (degrees)
(~ Forested area
u Vegetated area, including cropland

74o
40 ~ ~f ~ ~' ~40 "e
~P
~f
25
~f
~f
33 21 "r
40

r 27k~
33k~

,'~ ~'40

d~ 41,~
" ~ "~ ~'48
(~ 55~~ (~ (~ 391~
4..~ ~ '~22

Fig. 2. Basic geomorphological map of Hualushi. Note the two major gully systems adjacent to the settlement.
212 D . N . PETLEY

Fig. 3. Proposed planning scheme for Hualushi prior to this study.

 HAZARD ASSESSMENT IN EASTERN TAIWAN 213

(i) The drainage basin is relatively large, steep and opment. Great care is needed to ensure that the slopes at
densely vegetated, with about 700 m of relief. This this site are not disturbed by, for example, road building,
provides the potential for short lag times, provid- which could trigger potentially serious problems.
ing high discharges in typhoon events.
(ii) Some agricultural development of the drainage
basin has occurred. The removal of the forest Discussion and conclusions
reduces the environment's ability to retain water,
further reducing the lag time. The case study presented above demonstrates how terrain
(iii) Above the township, the highway traverses the evaluation in the form of geomorphological mapping
slope via hairpin bends. Examination of the road may contribute to the processes of creating development
at this point suggests that inadequate provision plans. The terrain evaluation demonstrated that the plan
has been made for allowing water in other gullies had limitations in that it failed to consider properly the
to cross the road, such that there is a tendency for hazards associated with the gully systems. As a result, a
water to utilize the highway as a fluvial channel. new plan has been put in place to allow for these features
This has the potential to feed water from adjacent and thus to protect the infrastructure and inhabitants.
slopes into the gully in question, especially as one Use of similar methods to assess the hazards asso-
of the hairpin bends lies within the drainage basin. ciated with cliffs above the road (Petley 1998b) has
It is difficult to predict the path of water as the demonstrated that the accuracy of the method is high.
differential temporal activation of the gullies will Variations in the nature of the processes operating in the
lead to different fluvial pathways being created, geomorphological hazards and in their frequency and
but in some storms it is likely that extra water magnitude are likely to occur, and these problems are
enters this gully via this route. likely to be exacerbated by variations in the techniques
(iv) Upon examination of the culvert under the road, it used to identify, examine and classify from one practi-
was observed that it was clogged with rubbish. tioner to another. Fortunately, the technique does not
If the pipe is choked then all of the flow must be need to suffer from the problems associated with quali-
routed across the road and through the town, tative assessments of hazard rating. However, overall
creating a significant hazard. the use of geomorphological mapping for the evaluation
of terrain during the planning process has considerable
The other gully that provides some concern is the large merit. As it is relatively low in cost and accurate, the
feature 75m or so up the road. Here another very system should be adopted widely.
large gully crosses the highway via a culvert. At this site
the culvert is large and appears to be well constructed.
However, there are signs that the road has suffered References
damage at this site, perhaps suggesting that water flows
across the road at this location during high flow events. BRABB, E. E. 1984. Innovative approaches to landslide hazard
This is a source of concern as this water has a good and risk mapping. International Symposium on Landslides
(Toronto, Canada), 1, 307-323.
chance of flowing down the road and through the BRUNSDEN, D., DOORNKAMP, J. C., FOOKES, P. G., JONES,
settlement, possibly combining with the flow in the other D. K. C. & KELLY, J. M. H. 1975. Large scale geomorpho-
gully. Similar concerns exist as at the previous gully in logical mapping and highway engineering design. Quarterly
relation to the road above, although there is potential in Journal of Engineering Geology, 8, 227-253.
this case for the road to remove water from the system. DEARMAN, W. R. 1991. Engineering Geology Mapping. Butter-
At present, this is difficult to predict and must not be worth Heinemann, Oxford.
relied upon. Fortunately, the catchment is not used for PETLEY, D. N. 1998a. Engineering hazards in the Taroko Gorge,
agriculture. However, this does not detract from the Eastern Taiwan. In: MAUND, J. G. • EDDLESTON, M. (eds)
concerns about these features. Geohazards and Engineering Geology. Geological Society,
London, Engineering Geology Special Publications, 15,
The gullies on the eastern side of the highway are 125-132.
deeply incised and active during storms. Care is also PETLEY, D. N. 1998b. Geomorphological mapping for natural
needed to ensure that these gullies are kept clear of hazard assessment in neotectonic terrains. Geographical
development and garbage. Journal, 164, 183-201.
No major slope problems were identified at the site, PETLEY, O. N., LIU, C-N. & LIOU, Y-S. 1997. Geohazards in a
although care is needed to ensure that the slopes adjacent Neotectonic Terrain, Taroko Gorge, eastern Taiwan.
to the incised gully systems are stable if used for devel- Memoir of the Geological Society of China, 40, 135-154.
Geological Society, London, Engineering Geology Special Publications
Slope instability within a residential area in Cleveland, UK
P. J. Phipps

Geological Society, London, Engineering Geology Special Publications 2001; v. 18; p. 215-219
doi:10.1144/GSL.ENG.2001.018.01.31

© 2001 Geological Society of London

Slope instability within a residential area in Cleveland, U K

P. J. Phipps
Mott MacDonald, Croydon, Surrey, U K

Purpose of survey Techniques used

An integrated geomorphological survey involving de- Detailed geomorphological field mapping was under-
tailed geomorphological field mapping and associated taken within the Valley Drive/Denevale area of Yarm at
desk studies is a powerful tool in identifying the locations an original scale of 1 : 500. The base maps were obtained
and nature of slope instability within built-up areas. from enlargements of existing 1:1250 Ordnance Survey
When supplemented with existing or specifically designed mapping. A total area of 400 m x 150 m was mapped in
ground investigation, initial instability processes and a two-day period. The mapping methods were based on
mechanism models can be re-evaluated and refined. This those described in Cooke & Doornkamp (1990) and
paper provides an example of a geomorphological survey Savigear (1965).
at Yarm, County Cleveland, which aimed to delineate The field mapping was supplemented by a desk study
unstable areas and address the potential causes of ground which included a review of historical Ordnance Survey
movements within a residential area. maps and available historical stereoscopic aerial photo-
graphs. These assisted in the identification of slope form,
water and drainage features, and areas of instability to
be set against the residential development and land-
The site scaping of the area. Some monitoring information was
also available from existing inclinometers, piezometers
Valley Drive and Denevale are situated at the far eastern and groundwater wells.
end of Yarm, on the south side of the River Tees, near
its confluence with the River Leven. The natural valley
side slopes are about 10-15 ~ in this area and rise up
towards a plateau surface at 30 m A.O.D., some 20-25 m The conceptual model
above the Tees Valley. In order to facilitate development
The natural slopes on this section of the Tees Valley
on the valley side slopes, extensive cut and fill operations
were extensively modified by cut and fill operations
have been carried out. Artificial slopes steepened up to
for housing development. At the outset of the geomor-
37 ~ and vertical walls to 3 m high were also evident.
phological field mapping the importance of under-
The published 1:10 560 scale British Geological Survey
standing the relationships between artificial slopes and
map (1976) and the 1:50000 scale British Geological
the natural geomorphology to aid stability assessments
Survey map (1987) indicate Triassic Sherwood Stand-
were recognized. The geomorphological maps produced
stone formation (formerly 'Bunter' Sandstone) as out-
delineated the different slope forms and their associ-
cropping within the river channel. The valley side slopes
ated instability features (Fig. 1). Three main types of
have been developed in thick superficial deposits which
slope instability were identified within the mapped area
comprise Upper and Lower Glacial Till units, separated
consisting of the following.
by Glacial Sands and Gravels overlying Laminated Clay.
These materials are believed to have been deposited 1. Compound landslides had progressed up-slope from
during the last Cold Stage (Devensian 80000-10000 the river edges. Failure occurred through in situ
years BC) (Catt 1991). Glacial Till and had typically incorporated variable
The published geological maps show extensive areas amounts of overlying fill. The landslides consisted
of landsliding within both the Tees and Leven Valleys. of a series of failures that appeared to move along
Prior to the start of operation of the Tees Barrage in non-rotational failure surfaces at depth. Existing
December 1994, a confidential survey of the landslides morphology was represented by a number of failed
within the valley was carried out by an independent blocks and associated scarp slopes. Surface expres-
consultant in 1989 that provided a record of instability sion of the failed areas appeared to be controlled by
features. the distribution of overlying weaker fill.

From: GR/FFITHS, J. S. (ed.) Land Surface Evaluationfor Engineering Practice. Geological Society, London, Engineering Geology
Special Publications, 18, 215-219. 0267-9914/01/$15.00 9 The Geological Society of London 2001.
216 P.J. PHIPPS

Fig. 1. Geomorphological map of Valley Drive and Denevale, Yarm, Co. Cleveland.

Translational landslides were typically limited to 3. Rotational slumps developed in fill slopes, which
locally oversteepened cut and/or fill slopes. The consisted of single rotational failures.
failure surface was dominantly planar and sub-
parallel to the ground surface. The depth of failure Each of these failure types had different implications
typically did not exceed 2 m. Surface form was often regarding their risk to various structures and potential
poorly defined on the ground in the study area. mitigation measures. The distribution of slope instability
 SLOPE INSTABILITY WITHIN A RESIDENTIAL AREA IN CLEVELAND, UK 217

el it

|- ~i
.~. -~i
ill

~L~-~"~ /
I~
o

6 c~

i 0~

ol %/
9 .~
g
6 .
o

,.o

I ~ _ l !" --,-,
= i io

.,.~

i i

i! o

._~

.~
~.~ ~- -a

i i! i ~ I i
218 P.J. PHIPPS

was found to be fundamentally controlled by the arti- geotechnical characteristics suggested that most instabil-
ficial modifications to the natural slopes. Five zones ity, whether shallow or more deeply seated, occurred
could be delineated within the side slopes with specific on slopes in excess of 25 ~ (i.e. Chandler 1984). Failure
stability characteristics. These are presented in Figure 2 tended to be independent of height although this factor
and described below. usually controls the failure mechanism. Slopes that were
less than 4 m in height and affected by instability tended
Z o n e A. Slopes adjacent to the river that had not been to be subject to translational landslides or rotational
altered by excavation activities. Vegetation slumps, with major compound landslides confined to
was that existing prior to development. Fill slope sections that exceed 4 m in height.
slopes directly upslope within Zone B had the The overall natural slope angles at the site did not
effect of loading the rear of these Zone A exceed 15~ and the majority of existing natural land-
slopes. slides in the Tees Valley were associated with slopes
Z o n e B. Extensive landscaping of fill slopes that had exceeding 20 ~. Slope instability in the study area tended
been built up on the existing lower valley side to be confined to failures involving artificial slopes,
slopes. These activities generally steepened although some natural instability related to bank ero-
the overall slope angles and increased shear sion by the River Tees was also evident.
stresses in the underlying slopes, including The occurrences of slope instability identified during
those in Zone A. the survey were not attributable to a single causal factor.
Z o n e C. Slope angles have been decreased by excava- Different forms of instability were present in various
tions in the existing natural slopes, mainly for slope zones which had their own stability characteristics.
house benches. Some of the causal factors that had an impact on the
Z o n e D. Slopes steepened by excavation. Small walls slope stability included the natural slope development
that may not have been designed as retaining with associated artificial modification to slope geometry
structures were often at the base of such cut and groundwater conditions. These were dominated by
slopes. the cut and fill operations and usage of the Tees Barrage.
Z o n e E. Original, more gently inclined upper valley Vegetation changes on the valley slopes related to devel-
side slopes and plateau surfaces that have been opment activity were also identified as a contributory
developed without the need for extensive cut factor to local shallow instability.
and fill operations.

Similar applications
What the survey established
The type of geomorphological survey adopted for the
The geomorphological survey identified the distribution residentially developed valley side slopes at Yarm can be
of three major landslide types. Both the translational applied to a variety of geomorphological situations for
landslides and rotational slumps were confined to indi- different project requirements. However, the basic tenet,
vidual cut/fill slopes and did not tend to exceed 3 m depth that the geomorphological field mapping records the
below ground level. Compound landslides were mainly existing landforms and processes and the desk study
present in the natural side slopes of Zone A directly adja- element provides general information on the site and its
cent to the River Tees, although they did locally impinge history, is a very powerful tool to assist land surface
into the lower fill slopes of Zone B. Existing ground evaluation for engineering purposes.
investigations indicated that the depth of the compound For the Yarm site it was a fundamental requirement
landslides was controlled by the presence of an eroded to delineate and classify the various instability types.
Sherwood Sandstone surface roughly at river bed level. Subsequent information enabled a factual description of
Derivative ground behaviour maps were developed the ground behaviour at the site to be produced which
for the site which incorporated both the form of slope then formed the basis for addressing the risks to indi-
instability and the rates/magnitudes of recorded move- vidual properties and the potential for major valley side
ments. Such maps provide a factual description of differ- slope failures.
ent recorded impacts associated with the various ground
movement types in each slope category and identify the
susceptibility of the slope categories to instability.
The potential for the site to be subject to major valley References
side failures was also addressed as events of this mag-
nitude could not be mitigated against without recourse BRITISHGEOLOGICALSURVEY.1976. Sheet N Z 41 S W, 1:10 560.
to major ground investigations and subsequent remedial BRITISH GEOLOGICALSURVEY. 1987. Sheet 33, Stockton Solid
works. Studies of other glacial till slopes with similar and Drift, 1:50 000.
 SLOPE INSTABILITY WITHIN A RESIDENTIAL AREA IN CLEVELAND, UK 219

CATT, J. A. 1991. The Quaternary History and Glacial Deposits COOKE, R. U. & DOORNKAMP, J. C. 1990. Geomorphology in
of East Yorkshire. In: EHLERS,J., GIBBARD,P. L. & ROSE, J. Environmental Management, second edition. Clarendon
(eds) Glacial Deposits in Great Britain and Ireland. Balkema, Press, Oxford.
Rotterdam, 185-193. SAVIGEAR, R. A. G. 1965. A technique of morphological
CHANDLER, R. C. 1984. Recent European experience of land- mapping. Annals of the Association of American Geogra-
slides in over-consolidate clay and soft rocks. Proceedings phers, 53, 514-538.
of the 4th International Symposium on Landslides, Toronto,
September 1984.
Geological Society, London, Engineering Geology Special Publications
Ground models for the design and construction of a high-speed rail link
P. J. Phipps

Geological Society, London, Engineering Geology Special Publications 2001; v. 18; p. 221-225
doi:10.1144/GSL.ENG.2001.018.01.32

© 2001 Geological Society of London

Ground models for the design and construction of a high-speed rail link

P. J. Phipps
Mott MacDonald, Croydon, Surrey, U K

Purpose of survey Pleistocene cold phases and to a lesser extent by more

quiescent inter-glacial phases (Jones 1980). The most
Engineering geomorphological, geological and geotech- recent effects on the landscape have been from the
nical information, pertinent to a long linear engineering geomorphological processes acting during the present
project, was required to provide support to a Parliamen- Holocene, and from the impacts of human activities.
tary Bill and assist the future design and construction of
the Channel Tunnel Rail Link. A wide variety of reports,
maps and borehole logs had been collated for locations Techniques used
in the vicinity of the route alignment. These were to be
supplemented with Phase I ground investigations for It was recognized early on at the Phase I stage that to
limited locations along the route during the winter of provide a relevant structure for the ground investiga-
1993/94. Phase I information was to be incorporated tions, associated studies and presentation of the infor-
with the desk study information to provide an initial mation, then a robust and appropriate methodology had
classification of the route geomorphology and ground to be adopted. Terrain systems mapping was identified
conditions, and provide a sensible basis from which to as providing a suitable hierarchical framework for pro-
design subsequent Phase II ground investigations. ducing initial ground models, designing the Phase II
ground investigations, and incorporating engineering
geological and geotechnical information back into the
The site ground models. The actual techniques used were based
on those described in Cooke & Doornkamp (1990) and
The Channel Tunnel Rail Link (CTRL) is to be con- for engineering projects in Brink et aL (1968). Satellite
structed between St Pancras and the Channel Tunnel imagery interpretation at 1 : 50 000 scale was undertaken
Railway Terminal at Cheriton, near Folkestone. The to assist with the delineation of the initial terrain systems.
CTRL is a high-speed railway, 108 km in length, and its Detailed geomorphological field mapping at 1"1250
construction will incorporate a variety of engineered was carried out for a number of type site areas to pro-
structures. The route traverses a wide range of topog- vide detailed information on specific landscape char-
raphy, and solid and drift geology, which produce a acteristics including morphology, near-surface materials,
wide variety of ground conditions to be considered for slope instability, drainage and areas of soft ground
design and construction purposes. (i.e. Brunsden et al. 1975). The detailed field mapping
At a macroscale the CTRL is located in three distinct was also supplemented by aerial photograph interpre-
geomorphological areas: the London Basin, the North tation for the entire route using stereoscopic colour
Downs and the northern limb of the Weald. A wide photographs at 1 : 2500 scale.
variety of solid and drift geologies were evident from
available British Geological Survey mapping: these
included Lower London Tertiary deposits, Pleistocene The conceptual model
Terrace deposits and recent alluvium in the London
Basin, the Upper Cretaceous Chalk-dominated land- The wide variety of ground conditions expected over the
scape from the Thames Valley to the North Downs, and route alignment and the various engineering structures
the Cretaceous strata of the Weald. that had to be designed and constructed for those
The geomorphological and structural geological fea- ground conditions required a robust and concise way of
tures of the route alignment had been developed mainly recording, assessing and generating the relevant infor-
as a function of post-Cretaceous landscape evolution mation (Waller & Phipps 1996). The terrain evaluation
during the Tertiary controlled by tectonic and iso- had to provide a basis for designing relevant and cost-
eustatic fluctuations. The landscape was further signifi- effective Phase II ground investigation and be flexible
cantly modified by periglacial conditions during the enough to allow revision by it.

From: GRIFFITHS,J. S. (ed.) Land Surface Evaluation for Engineering Practice. Geological Society, London, Engineering Geology
Special Publications, 18, 221-225. 0267-9914/01/$15.00 9 The Geological Society of London 2001.
222 P . J . PHIPPS

~d

e~

"~0
o
"-6

o
 GROUND MODELS FOR THE DESIGN AND CONSTRUCTION OF A HIGH-SPEED RAIL LINK 223

Initial interpretations o f satellite imagery were refer- m a p areas up to 1 k m 2 that translated to over 20 k m o f
enced against available published geological m a p p i n g to the C T R L route length.
provide small-scale division of the route and its environs Following the detailed geomorphological mapping,
into terrain systems. These divisions reflected character- the type site areas were divided into typically less than
istic topography, drainage and vegetation correlated to eight terrain facets. The facets summarized the detailed
g e o m o r p h o l o g y and geology. field data as each one comprised characteristic slopes,
The detailed geomorphological field m a p p i n g was drainage, weathering profiles, associated relict and con-
u n d e r t a k e n for type site areas within those terrain sys- t e m p o r a r y processes, and engineering concerns. G r o u n d
tems where it was expected to have benefits for the ter- profile information was also summarized for each facet
rain evaluation and g r o u n d model d e v e l o p m e n t (Fig.l). from assessment o f Phase I and historical borehole data,
N o m a p p i n g was therefore u n d e r t a k e n in the floodplain field observations o f drift and solid geotechnical expo-
areas or the L o n d o n Basin where the C T R L would be in sures, and h a n d augering. These data were presented as a
a deep tunnel. Two to three days were typically taken to series of s u m m a r y tables accompanying the facet maps

Table 1. Example extract o f a facet summary table: Terrain system XI: ttythe Beds surface

Facet Facet Slopes Materials Drainage Contemporary Relict processes Engineering

code name processes concerns

Hythe 0-3 ~ Typically 0-2 m of sandy No evidence of Very slow Solution Variability of
Beds clays and silty sands with ephemeral movement of weathering depth to
plateeau fragments of flint chert drainage in superficials from Tertiary rockhead.
surface and ragstone (head) the study downslope by of upper Variability in
overlying 18 m or less of area. creep and layers, extent and
interbedded sands, hillwash Periglacial type of
sandstone and limestone. processes. processes superficial
The basal members consist of Plateau surface Slow solution during the last deposits.
silty or clayey sands and well drained. weathering of cold phase Presence of
thin to medium bedded Slopes are underlying have led to palaeoshear
weak to moderately strong oversteepened Hythe Beds reworked surfaces in
sandstones (Hythe Beds). to 3~ around possible. materials to superficial
Sands at the base grade into the heads of 3 m depth deposits.
silts and clays in the spring sapped or more, Presence of
transition zone with the valleys and fracturing infilled
underlying Atherfield (see III). of the quarries with
Clay. underlying little or no
Hythe Beds. surface
expression.
Cambered 3-6 ~ See I. No perennial Slow processes Dislocation and See I.
Hythe Nature of the cambering drainage of hillwash cambering Lateral
Beds leads to a greater thickness features. and creep, of Hythe variability in
zone of superficials masking the Slopes are Cambering of Beds blocks Hythe Beds
underlying Hythe Beds. oversteepened blocks of especially sandstone
Up to 4 m of head to 8~ around Hythe Beds during the bands.
recorded, typically the heads of downslope Anglian cold Possible
consisting of stiff orange spring sapped may be phase, presence of
brown and green clays valleys. occurring at Evidence of large-scale
with a little angular fine a very slow former palaeoshears
to coarse flint or quartz rate. perennial developed
gravel. Spring sapping drainage in during the
Disturbed Hythe Beds at Atherfield dry valley Quaternary.
caused by the cambering Clay/Hythe forms.
of the plateau surface Beds
near the scarp front leads interface.
to the formation of gulls
and open joints, which
tend to fill with
superficials.
224 P . J . PHIPPS

e~

e4
.a
 GROUND MODELS FOR THE DESIGN AND CONSTRUCTION OF A HIGH-SPEED RAIL LINK 225

(Table 1). Facets typically included dry valley bottoms, Similar applications
steep valley side slopes and plateau surfaces (Fig. 1).
Many of the boundaries between the terrain facets Terrain evaluation for civil engineering purposes is
were morphological and the aerial photograph inter- rarely utilized, mainly owing to the lack of specific
pretation was used to extend the facet mapping away detailed information required by designers and contrac-
from the type site area in each terrain system. The tors. However, such studies can provide appropriate
process proved to be very rapid and accurate to within summaries of engineering geomorphological features
5 m for areas of open fields. Ground truthing was and typical ground conditions. They may also provide
undertaken to verify the facet boundaries. the basis for designing more detailed ground investiga-
tions to address specific issues or enable more subsurface
information to be obtained for terrain systems or indi-
What the survey established vidual facets where appropriate information is lacking.
The technique of terrain evaluation can be adopted
A total of 14 terrain systems were identified through for any long linear engineering project including roads,
which the CTRL route was to pass. Within each of the railways and pipelines, and additionally for resource
terrain systems outside the London Basin tunnels area, surveys. A large amount of data can be synthesized into
facet mapping was produced at a scale of 1 : 2500 from a a simpler format, or conversely limited data in certain
combination of detailed geomorphological field map- locations can often be extended into similar terrain
ping and aerial photograph interpretation. systems that are lacking in information.
Relevant engineering geomorphological and geologi-
cal information to assist in the design and construction
of the CTRL was synthesized for each facet. Sum-
mary tables (Table 1) and three-dimensional conceptual
ground models that showed the relationship between References
surface features, contemporary processes and the under-
lying geotechnical materials and structure were pro- BRINK,A. B. A., PARTRIDGE,T. C., WEBSTER,R. & WILLIAMS,
A. A. B. 1968. Land classification and data storage for the
duced to aid in visualization of the issues (Fig. 2).
engineering usage of natural materials. Proceedings of the
The majority of the Phase II site investigations were Symposium on Terrain Evaluation for Engineering. Aus-
designed on the basis of the preliminary facet mapping. tralian Road Research Board, 1624-1647.
Locations of boreholes or trial pits were often decided by BRUNSDEN, D., DOORNKAMP, J. C., FOOKES, P. G., JONES,
a requirement to improve the subsurface information for O. K. C. & KELLY, J. M. H. 1975. Large scale geomor-
certain terrain facets. Both facet boundaries and terrain phological mapping for highway engineering. Quarterly
system boundaries were often subject to specific Phase II Journal of Engineering Geology, 8, 227-253.
investigation, especially when the boundaries were un- COOKE, R. U. & DOORNKAMP,J. C. 1990. Geomorphology in
clear or contentious. In this way the terrain evaluation Environmental Management. Clarendon Press, Oxford.
could be refined in response to the improved data JONES, D. K. C. 1980. The Shaping of Southern England. Insti-
tute of British Geographers, Special Publication, No. 11,
and the knowledge of geomorphological, geological and Academic Press.
engineering characteristics and issues within the facets. WALLER, A. M. • PHIPPS, P. J. 1996. Terrain systems mapping
Anomalous features identified in the field and from the and geomorphological studies for the Channel Tunnel Rail
aerial photograph interpretation were also investigated Link. In: CRAIG, C. (ed.) Advances in Site Investigation
in a further attempt to understand their origin. Practice. Thomas Telford, London, 25-38.
Geological Society, London, Engineering Geology Special Publications
Slope stability hazard assessment: Coalport Railway Bridge, Shropshire
D. T. Shilston

Geological Society, London, Engineering Geology Special Publications 2001; v. 18; p. 227-231
doi:10.1144/GSL.ENG.2001.018.01.33

© 2001 Geological Society of London

Slope stability hazard assessment: Coalport Railway Bridge, Shropshire

D. T. Shilston
W. S. Atkins Consultants Ltd, Epsom, Surrey, U K

Objectives dismantled in 1960, but the roads, bridges and trunk

sewer are in everyday use.
Escalating costs, disputes and delays can arise during The solid (bedrock) strata in this part of the Iron-
construction as a result of inadequate research at the early bridge Gorge comprise mudstones and siltstones with
stages of project development. The assessment of slope persistent sandstone layers of Carboniferous age (Fig. 3).
stability hazard described in this case history illustrates They are mapped by the British Geological Survey (BGS)
the relative ease, and therefore low cost, of rapidly access- as dipping at a shallow angle to the east. Two north-south
ing and evaluating a wide range of desk study and walk- trending faults are mapped by BGS in the area, one of
over survey data pertinent to a construction project. which (if extrapolated) would pass very close to the rail-
way bridge. These strata are overlain by thin impersistant
drift deposits and made ground. The BGS map shows
The project large landslides adjacent to the Railway Bridge, but none
at the bridge itself. However, man-made topographical
Coalport is a village on the north bank of the River changes have obscured much of the subtle geomorpho-
Severn in the Ironbridge Gorge, near the new town of logical evidence for the original form of the hillside (and
Telford, Shropshire. CoMport Railway Bridge (BC 17) is hence landslips) in the vicinity of the Railway Bridge.
at the site of the former Coalport Railway Station. It is
immediately adjacent to the much older CoMport (river)
Bridge and carries the road from the river bridge across
the former railway line (see Figs 1 and 2). Techniques
As part of a programme of bridge inspection and
One key aspect of the studies for Railway Bridge BC17
strengthening, the Commission for New Towns engaged
was the integration of conventional engineering geolo
WS Atkins to assess the condition of Railway Bridge BC
gical and geomorphological data (including mapping)
17 and make recommendations for remedial measures.
with information about the site's history and human
An initial inspection of the Railway Bridge concluded
influence on the landscape. Only when this was done was
that it was in a poor state of repair and that it may be at
it possible to arrive at a broadly based understanding of
risk from ground movements caused by landslipping,
the landslide hazard at the site.
being in an area that is well known for the presence of
The classes of information used in the work are sum-
landslips (Carson & Fisher 1991). A more detailed desk
marized on Table 1. They may appear to be unusually
study, including a walk-over survey, was therefore car-
extensive but, in the author's experience, this is not the
lied out specifically to assess the landslide hazard at the
case. For many parts of Britain, particularly old indus-
railway bridge and to help design the ground investiga-
trial areas and areas of potentially adverse geotechnical
tion that was planned as part of the design of remedial
conditions, investigators will find a large, varied and use-
works to the bridge's abutments.
ful corpus of background and site-specific information
The natural topography of the site comprises the bank
(Shilston et al. 1998).
of the River Severn leading up to a narrow river terrace
Three sources of information were found to be
and then the steep sides of the gorge whose crest is about
particularly useful:
55 m above the river terrace. In the vicinity of the Rail-
way Bridge, the natural topography has been extensively 9 history of the CoMport branch Railway (Smith 1987);
modified by the construction of Coalport Bridge (built in 9 historical study of CoMport (river) Bridge (Trinder
1818 on the foundations of an earlier timber bridge of 1979);
1780; see Fig. 2); the CoMport Canal (1792); the Coalport 9 series of vertical stereographic aerial photographs of
Branch Railway with Railway Bridge BC 17 and railway the site area: the earliest was from 1946 and shows the
station (1861); roads; various private houses; and recently topography and geomorphology of parts now ob-
a trunk sewer from Telford new town. The railway was scured by modern forestry.

From: GRIFFITHS,J. S. (ed.) Land Surface Evaluation for Engineering Practice. Geological Society, London, Engineering Geology
Special Publications, 18, 227-231. 0267-9914/01/$15.00 9 The Geological Society of London 2001.
228 D. T. SHILSTON

Fig. 1. Ordnance Survey map extract, 1883.

Fig. 2. View of Coalport River Bridge (from Trinder 1979).

SLOPE STABILITY HAZARD ASSESSMENT: COALPORT RAILWAY BRIDGE, SHROPSHIRE 229

Fig. 3. Engineering geological sketch map.

230 D.T. SHILSTON

Table 1. Summary of information used in the desk study Building upon this body of background information,
the desk study and associated walk-over survey enabled
Class of information Number specific conclusions to be drawn about the character-
of items istics of slope instability that could effect the Railway
Topographic maps (historic and modem) 10 Bridge:
Aerial photographs (various dates) 6 9 very shallow, near-surface landslipping is likely wher-
Geological maps (including maps of boreholes, 6 ever clay or mudstone strata are present on ground
mines, wells and shafts
steeper than say 8~ that is, throughout the valley side;
Technical papers and reports (site history) 9
Technical papers and reports (geology 14 9 groundwater egress at springs is causing localized very
and geotechnical engineering) shallow landslipping at the present day;
9 apparently shallow landslipping is present to the west
of the Railway Bridge on the broad expanse of mud-
stone outcrop which forms the lower zone of the hill-
side below a prominent sandstone unit;
The walk-over survey was carried out on two occasions. 9 apparently complex landslipping may be present to
The first visit, which was at the start of the desk study, the east of the Railway Bridge where the outcrops of
provided a general understanding of the site and its two sandstone units drop in elevation down the valley
complex geomorphology and helped focus the desk side towards the River Severn;
study work on areas and technical aspects that seemed 9 there is no firm geological evidence for landslipping
most likely to be significant. The second visit was made in the area immediately uphill of the Railway Bridge;
when the desk study was nearly complete. Its primary 9 there is no geological evidence for instability affecting
objectives were to provide 'ground truth' for the desk- the River Terrace Deposits that occur at the base of the
based work and to investigate in the field specific areas slopes adjacent to the River Severn;
of interest that had been identified. Both visits were 9 the satisfactory long-term (175 years) performance of
made more cost-effective by being co-ordinated with the cast-iron Coalport (river) Bridge provides evidence
visits to local sources of information. of the stability of the lower part to the hillside as a
The methodology for the engineering geological whole; long-term squeezing or creep, as experienced for
mapping and presentation of the sketch map (Fig. 3) example some 2 km upstream at the Iron Bridge, does
follows that illustrated by Dearman & Fookes (1974). not appear to be taking place.
Figure 3 brings together information obtained by the
It was concluded that the known deeper-seated landslips
desk study and walk-over survey: it illustrates the engi-
in the vicinity of the Railway Bridge do not appear to be
neering geology, geomorphology and history of the site
active at the present day, although there are signs of very
on a single figure. One or more synoptic maps such as this
shallow movements on most steep slopes and where
can be a very effective way of conveying complex infor-
there is groundwater egress. The known landslips can
mation about the conditions and development of a site.
therefore be classified as dormant. However, the general
experience of slope instability in the Ironbridge Gorge
warns that movement could be initiated in the future by
man-made or natural changes in the condition of the
Results existing landslips, such as the ingress of water, placing of
fill at their crests or excavation at their toes.
The desk study work readily identified published Evidence for landsliding was sought in the ground
information about landslips, elsewhere in the Ironbridge investigation for the remedial works at the Coalport Rail-
Gorge, that provided general guidance on slope in- way Bridge. The costs of doing so were much reduced by
stability in the area. Principal conclusions were: the findings of the desk study and walk-over survey
which provided a rational basis for both the design and
9 major deep-seated landslipping in the Ironbridge interpretation of the ground investigation.
Gorge was initiated by rapid erosion of the gorge by
the River Severn; Acknowledgements. The author would like to thank the Com-
9 shallower landslipping is present, both as a veneer to mission for New Towns for permission to publish this paper.
the deeper-seated landslips and in areas apparently
lacking such landslips;
9 some landslips are active today or have been stabilized
by modern civil engineering works;
References
9 recent reactivation of dormant shallow and deeper- CARSON,A. M. & FISHER,J. 1991. Management of landslides
seated landslips or accelerated movement of active within Shropshire. In: CHANDLER,R. J. (ed.) Slope Stabil-
landslips is generally the result of interference by man. ity Engineering. Thomas Telford, London, 95-99.
 SLOPE STABILITY HAZARD ASSESSMENT: COALPORT RAILWAY BRIDGE, SHROPSHIRE 231

DEARMAN, R. W. & FOOKES, P. G. 1974. Engineering geo- The Value of Geotechnics in Construction. Construction
logical mapping for civil engineering practice in the United Research Communications Ltd, London, 25-36.
Kingdom. Quarterly Journal of Engineering Geology, 7, SMITH, W. H. 1987. The Coalport Branch. British Railway
223-256. Journal, 19, 399-415, 436-440.
SHILSTON, D. T., HARRISON, E. N., PARSONS, A. S. & LEE, K. TRINDER, B. S. 1979. Coalport Bridge: A study in historical
1998. Giants' shoulders: the cost-effective use of geotech- interpretation. Industrial Archaeology Review, 3, 153-157.
nical desk studies in civil and structural engineering. In:
Geological Society, London, Engineering Geology Special Publications
Evaluation in urban and industrial environments: the Docklands Light
Railway, Lewisham Extension, London
D. T. Shilston, N. E. Harrison and D. J. French

Geological Society, London, Engineering Geology Special Publications 2001; v. 18; p. 233-237
doi:10.1144/GSL.ENG.2001.018.01.34

© 2001 Geological Society of London

Evaluation in urban and industrial environments: the Docklands Light
Railway, Lewisham Extension, London

D. T. Shilston, N. E. Harrison & D. J. French

W. S. Atkins Consultants Ltd, Epsom, Surrey, U K

Objectives Table 1. Aide-memoire to classes of information for engineering

geological, geomorphological and geotechnical desk studies
This case history describes the use of information
Classes of information Examples
obtained from the initial desk-based engineering geolo-
gical and geotechnical studies carried out as part of Topography Maps
the design of civil engineering works for the Lewisham Aerial photographs
Extension of the Docklands Light Railway (DLR) which Satellite imagery
was constructed in a heavily developed urban environ- Geology and Maps
ment. The desk study aimed to make best use of available hydrogeology Aerial photographs
information in order to evaluate the terrain, limit the Satellite imagery
need for further ground investigation and provide geo- Memoirs, reports, papers and books
logical and geotechnical information to the design teams Mine and quarry records
Thematic databases (e.g. in UK for
as quickly as possible. landslides, natural cavities, etc.)
A desk study is an integral part of terrain and site Previous ground investigations
evaluation and, wherever possible, should be carried
Environment and Planning maps
out in conjunction with a reconnaissance survey (or planning Aerial photographs
walk-over) of the site. Together they are likely to be of Satellite imagery
greatest value and most cost-effective when carried out Archaeological site and historic
early in the investigations for a proposed development. building records
They can provide the project team with an early indica- Mine and quarry records
tion of the conditions at the site, such information being Contaminated land records
Landfill and waste disposal records
essential for master planning and concept design. As the
Environmental statements
project is advanced the desk study findings become the Meteorological records
basis for the planning and interpretation of the sub- River and coastal information
sequent more expensive and lengthy stage of physical Site conditions, land Historical maps
investigation by (for example) boreholes and trial pits. use and history Historical documents
Conceptually, it should be self-evident that it is cheaper Aerial photographs
to obtain and evaluate information that already exists Satellite imagery
than to obtain new investigation data. But this is not the Land use and planning maps
only reason for carrying out desk studies. Their ability to Site walk-over/ 'Skilled eye' inspection of site and
recover information that cannot be obtained by other reconnaissance survey locality
means is also very valuable. For example, in an urban or Local knowledge and Local history societies
industrial environment, the study of historical maps, experience Library local studies departments
archive aerial photographs and other documentary infor- Local press
mation can provide vital insights into the past condi- Neighbours
tion or uses of a site and the hazards that such use uses Previous site users
Construction records
could pose to a new development (e.g. the presence of an Building control office
ancient landslip, of old coal mines or of a potentially
contaminating former industrial process). Thus, desk Precedent Case histories
Construction records
studies are part of the process of understanding and
managing ground hazards and risks. Codes, standards and Professional bodies and institutes
guidance Government departments
When planning desk studies it is helpful to use an aide- Research organizations and
memoire or checklist of major classes of information universities
that could be examined, such as that given in Table 1

From: GRIFFITHS,J. S. (ed.) Land Surface Evaluation for Engineering Practice. Geological Society, London, Engineering Geology
Special Publications, 18, 233-237. 0267-9914/01/$15.00 9 The Geological Society of London 2001.
234 D. T. SHILSTON E T AL .

MOUSE
LINK R O A D T U N

_- NORTH
-GREENWICH

CROSS~UR --
=--'-WATER

C ' E r ~ S~IK

ELVERSON R D ;

LEWISHAM-

I I I

SOLID GEOLOGY

RAILWAY
STAIIONS
FAULT, TICK O N
1 l L~176 cLAY ~ W SIDE
ms. m m Ft'm~JmM
I~I~NSION

IXX:IO..ANDSl.lOIrr AJ~OF~
RAILWAY HX]8"HIqO
-

m m T O
ItA]LWAY
M ~ ~ • ~ I S OF SYI~LiI~

_-- ~_ GRBBNWICH P o o r ~
STRUCllJlU~ I101~8
CHALK

Fig. 1. Geological sketch map.

 EVALUATION IN URBAN AND INDUSTRIAL ENVIRONMENTS 235

(Shilston et al. 1998). The table is based on the authors' was essential that a senior member of staff comprehen-
own experience and can be augmented by guidance sively briefed team members on all relevant aspects of
for individual countries or regions. As an example, for the project. In addition, experienced staff from the vari-
England, more detailed information on many of the ous disciplines involved in the project regularly reviewed
sources mentioned is given in Transport Research Labo- the work as it progressed.
ratory Report 192 (Perry & West 1996), The Geologist's The desk study comprised a review of existing geologi-
Directory (Geological Society 2000) and the English cal and hydrogeological data, site history records and
Local Studies Handbook (Guy 1992). construction precedent as summarized in Table 2. The
table also summarizes the technical benefits of this work.
Presentation of information and results in an appro-
The project priate format for issue as formal reports can be time-
consuming. An informal desk study file was maintained
The Lewisham Extension to the Docklands Railway in the early stages of the project. It could be easily up-
(DLR) runs approximately 4.2 km from Mudchute in dated and held the key reference information along with
London's Docklands, beneath the River Thames to summary tables similar to that reproduced in Table 2.
Greenwich, and thence to Lewisham (Fig. 1). From north In addition, figures that summarized information, such
to south, the new railway comprises a cut and cover as overlays of historical maps, were produced so that the
tunnel section from Mudchute to Island Gardens, twin desk study findings could be properly assessed and
bored tunnels under the River Thames to just west of presented to all parties.
Greenwich mainline station, a short length of cut and Further details of the desk study are given in Shilston
cover tunnel and then a surface section (partly on a et al. (1998). Here we comment on one specific aspect
viaduct) from Greenwich Station to Lewisham. As well of the work: the value of information about site history
as the railway line itself, two existing stations were and construction precedent to the overall evaluation of
replaced and five new stations constructed. the site. Such information can be particularly important
LRG Contractors, a joint venture between John where previous urban and industrial land uses are a sig-
Mowlem Civil Engineering plc and Mitsui-Nishimatsu, nificant component of the terrain through which a pro-
were awarded a contract to design and build the exten- ject is to be constructed.
sion. WS Atkins were appointed by the Joint Venture
partners to undertake the detailed design of the works.
This was a fast-track multidisciplinary project where
information for preliminary design of the various struc- Results
tures was required almost immediately, to be updated
with more accurate information as the design process In common with many urban and industrial areas, Lon-
progressed. The initial geotechnical tasks were to review don's Docklands and the London Boroughs of Green-
the geotechnical parameters used in the tender design wich and Lewisham have undergone numerous phases
and to determine what additional site investigation of development and redevelopment. Information on site
would be necessary for detailed design to be carried out history and construction precedent in the area was
with confidence. Due to the very tight construction pro- important in assessing the ground hazards faced by the
gramme, the scope of any additional ground investiga- project, determining the values of engineering properties
tion had to be kept to a minimum and was restricted to a for the design of the works and in the selection of con-
limited number of high quality boreholes and in situ struction methods. Examples of the use of, and conclu-
tests. Collation and evaluation of existing information sions drawn from, such information are given below.
by means of a desk study were, therefore, central to the
planning of the additional investigations and to the geo-
technical design.
Identification of construction hazards and risks
Groundwater Regime. The groundwater regime in this
Techniques area of east London is complicated by the proximity of
the Thames and rising groundwater levels in the under-
Appropriate management, resourcing and presentation lying Chalk aquifer. This was of particular concern for
of the desk study work were critical to achieving the the construction of the 23 m deep station box at Cutty
project requirements in the very short timescale avail- Sark, which required the use of a very sophisticated
able. A flexible team approach was adopted which could dewatering system.
be tailored to the changing project priorities. Junior staff Unexploded bombs. The area was heavily bombed
carried out most of the collection and preliminary review during the Second World War. There was therefore a
of reference information. For this to be successful, and possibility that unexploded bombs might be encountered
in order that all relevant information was obtained, it during construction work.
236 D. T. SHILSTON ET AL.

Table 2. Summary of availability and benefits of some of the information assessed during the desk study for the Lewisham Extension to
the Docklands Light Railway

Classes of information Source Description of information Benefits

Geology and hydrogeology British Geological Survey BGS maps show the Better understanding of local
(BGS) maps and memoirs Greenwich Fault and geological structures,
associated folding and including faults, folds and
disturbances. unconformities and of
potential hazards such as
swallow holes.

Borehole records from BGS have numerous boreholes Enhanced appreciation of

previous site investigations in the vicinity of the DLR ground conditions, thus
extension, limiting the amount of
additional site investigation
required.

Published geological papers Howland (1991) provides a As above

and literature useful overview. Recent
review by BGS has led
to a reclassification
of the Tertiary strata in
the London area (Ellison
et al. 1994).

Published hydrogeological CIRIA (1989) and Lucas & Better understanding of

papers and literature Robinson (1995) indicate groundwater regime
that the DLR extension is including influence of rising
in an area which may be groundwater table.
affected by the rising
groundwater table in the
Deep Aquifer.

Mining records Department of the Identified potential hazard of

Environment (1991) encountering shafts/
recorded mining activities deneholes.
in the Greenwich area.

Site conditions, tand use Historical maps Miscellaneous historical Identified potential hazards
and history maps from 1703 onwards associated with previous
and Ordnance Survey land use, including
maps from 1869 to the contaminated land, old
present day. foundations (piles, footings,
etc.) and other construction
works (river walls,
basements, tunnels, etc.)

Precedent Published case histories of Limehouse Link project Obtain details of previous
construction projects in the experienced problems with construction experience,
area, particularly those dewatering part of the particularly identification of
involving deep excavations Tertiary strata which potential construction
and tunnels. contained perched water problems and appropriate
and had high permeabilities. solutions.
Records of the construction
of the Greenwich Foot
Tunnel and shafts
(Copperthwaite 1901/02).
 EVALUATION IN URBAN AND INDUSTRIAL ENVIRONMENTS 237

Mining instability. Chalk and sand were mined in the in complex urban and industrial terrain such as that in
nineteenth century in the Greenwich area. the area traversed by the Lewisham Extension of the
Contaminated land. Ground and groundwater con- D L R . The desk study work described in this paper
tamination were likely to be present from a wide range played a fundamental role in identifying hazards and
of former industrial activities at a number of localities risks, developing ground models, and facilitating the
along the route. development of effective mitigation measures through
Building foundations. Piles and shallow foundations the selection of construction methods and the design of
were likely to be present beneath some of the existing temporary and permanent works. These contributed to
and demolished buildings in the vicinity of the surface the successful construction of the project, the D L R
sites and bored tunnels. Extension to Lewisham being opened in November 1999
some two months ahead of programme.
Identification of risks to adjacent structures
Acknowledgement. The authors would like to thank LRG
Ground movements caused by the works may affect Contractors for permission to publish this case study.
some adjacent buildings. Where possible, construction
records were obtained for prestigious or particularly sen-
sitive structures. These included the existing Greenwich
mainline railway station which was built in the nine- References
teenth century and is a Grade II listed historic building.
Drawings of the station's superstructure and founda- CIRIA. 1989. The Engineering Implications of Rising Ground-
tions were acquired which proved invaluable in assessing water in the Deep Aquifer Beneath London. Special Pub-
the impact on the station of the proposed tunnelling and lication 69. CIRIA, London.
surface works. COPPERTHWAITE,W. C. 1901/02. The Greenwich Footway Tun-
nel. Proceedings of the Institution of Civil Engineers, 150.
CRUIKSHANK,J. 1993. Limehouse Link Supplement. New Civil
Precedent experience Engineer. May.
Greenwich Foot Tunnel. Opened in 1902, this tunnel DEPARTMENT OF THE ENVIRONMENT. 1991. Review of Mining
Instability in Great Britain. Regional Report, South East
connects Island Gardens to Greenwich and runs beneath England including Greater London, Vol. I/ft.
the Thames close to the proposed route of the D L R ' s ELLISON,R. A., KNOX, R. W. O'B., JOLLEY,D. W. & KING, C.
twin bored tunnels (see, Fig. 1). There is good documen- 1994. A revision of the lithostratigraphical classification
tation for the Foot Tunnel, including a contemporary of Early Palaeogene strata of the London Basin and East
paper published in the Proceedings of the Institution o f Anglia. Proceedings of the Geologists' Association, 105,
Civil Engineers which describes the method of construc- 187-197.
tion and the ground conditions encountered (Copper- GEOLOGICAL SOCIETY. 2000. The Geologists" Directory. Geo-
thwaite 1901/02). logical Society of London.
Recent construction projects. The D L R Lewisham GuY, S. 1992. English Local Studies Handbook. University of
Exeter Press.
Extension is close to several major engineering pro- HOWLAND,m. 1991. The Engineering Geology of the London
jects such as the Limehouse Link Tunnel, Jubilee Line Docklands. Proceedings of the Institution of Civil Engi-
Extension and Canary W h a r f (see Fig. 1). Information neers. Part 1, 90, 1153-1178.
about these projects highlighted potential construc- LUCAS, H. C. & ROBINSON, V. K. 1995. Modelling of rising
tion difficulties and particular ground hazards. These groundwater levels in the Chalk aquifer of the London
included, for example, difficulties encountered with de- Basin. Quarterly Journal of Engineering Geology, 28, Sup-
watering the excavations for the Limehouse Link project plement 1.
(Cruikshank 1993). PERRY, J. & WEST, G. 1996. Sources of Information for Site
Investigations in Britain (Revision of TRL Report LR 403).
Report 192, Transport Research Laboratory, Crowthorne.
SHILSTON,D. T., HARRISON,N. E., PARSONS,A. S. & LEE, K.
Overview 1998. Giants' shoulders: the cost-effective use of geotech-
nical desk studies in civil and structural engineering. In:
Management of ground hazards and their consequent The Value of Geotechnics in Construction. Construction
risks is an essential feature of civil engineering projects Research Communications Ltd, London.
Geological Society, London, Engineering Geology Special Publications
Land surface evaluation: conclusions and recommendations
D. Brunsden and J. S. Griffiths

Geological Society, London, Engineering Geology Special Publications 2001; v. 18; p. 241-243
doi:10.1144/GSL.ENG.2001.018.01.35

© 2001 Geological Society of London

Land surface evaluation: conclusions and recommendations

D. Brunsden & J. S. Griffiths

Emeritus Professor of Physical Geography, Kings' College, London, UK

Results from the Second Working based on a sound scientific rationale. In this way the
local characteristics can be set within a general synoptic
Party Report appreciation of the wider situation and factors that
As discussed in the introduction to this report (Grittiths might influence a possible development site from outside
& Edwards), the Second Working Party on Land Surface the immediate area can be taken into account.
Evaluation in Engineering Practice recognized that the The First Working Party envisaged land surface
subject area was so extensive that an 'instruction man- evaluation being primarily the combination of geomor-
ual' or handbook, similar to previous working party phological mapping and aerial photograph interpretation
reports, would not be appropriate. The end product, as (i.e. remote sensing). This would be carried out based on a
presented, therefore, is a compilation of examples of framework of land classification and supported by com-
techniques and case studies written by practitioners in the prehensive desk studies. Comparing the examples pre-
field. The papers on techniques only provide an intro- sented in the two working party reports, the importance
duction to the subject and an indication where details on of these techniques has continued. These techniques are
methodology can be found. The case studies are taken exemplified in the second report by a series of case
from the files of working professional engineering geol- studies:
ogists, or consultants in the subject who are now in 9 Shilston, Harrison & French (desk studies for the
academia, although many have a background of indus- Docklands light railway);
trial experience. Most of the case studies have not been 9 Hearn (aerial photograph interpretation for road
published before, those that have either been expanded to construction in mountainous terrain);
highlight the importance of land surface evaluation (e.g. 9 Birch (geomorphological mapping for gas pipelines in
Hearn, Blong & Humphreys; Edmonds) or they represent a South Wales);
key development in the subject (e.g. Jones). 9 Phipps (land classification for the Channel Tunnel
The material presented in this report confirms that high speed rail link);
Land Surface Evaluation is established as part of essen- 9 Fookes, Lee & Sweeney (terrain evaluation for a
tial good practice in land care, management and the pipeline in Algeria).
development of the earth's surface. Since the first report
(Anon 1982), it has become clear that the most important However, since the first report there has been a vast
achievement is the realization that land surface data can increase in the techniques associated with remote sens-
be easily integrated with all other ground information. ing which are now beginning to impinge on engineering
The culmination of this is the production of a geological projects. Whilst satellite image interpretation has only
ground model that embraces the work of all earth science really found a place on very large regional projects, aerial
disciplines (Fookes 1997). In the case studies it is appar- photography and spatial data handling systems have
ent that the over-tiding objective, and the main purpose undergone significant changes. Of particular interest are
of employing land surface evaluation techniques, is the the developments in digital photogrammetry (Chandler),
development of an accurate ground model that minimizes and Geographical Information Systems (GIS) (Nathanail
the unknowns, identifies potential and actual problems, & Symonds; Hearn, Hodgson & Woddy). However, over
avoids hazards and maximizes the opportunities pro- the next decade we expect to see the full application of
vided by a site. Carrying out an effective land surface stereo-satellite imagery, pre-defined multi-spectral wave-
evaluation for engineering projects prior to going on to band sensing on aerial surveys, synthetic aperture radar,
site allows planning decisions to be made, ground investi- and airborne laser scanning. In addition, with the advent
gations to be designed in the most cost-effective way, and of more work in the nearshore and offshore zones, the
preliminary designs to be drafted. The ground model present standard techniques of seabed remote sensing
developed provides a framework within which all subse- such as seismic profiling, bathymetric and side-scan sonar
quent data can be checked and analysed, as well as survey will be increasingly utilized to establish geological
allowing extrapolations between data points to be made ground models.

From: GR1FFITHS,J. S. (ed.) Land Surface Evaluation for Engineering Practice. Geological Society, London, Engineering Geology
Special Publications, 18, 241-243. 0267-9914/01/$15.00 9 The Geological Society of London 2001.
242 D. BRUNSDEN & J. S. GRIFFITHS

The other area where the case studies in the second studies and, increasingly, risk management will continue
report clearly show development since the first report, is to grow in importance for practitioners in land surface
in the growing recognition that 'geomorphology', as one evaluation at all stages of future and existing engineering
of the critical disciplines within land surface evaluation development projects.
studies, is not merely the science of 'morphology'. It is Another new technique that has been developed is
now established that geomorphology provides a basis ground behaviour analysis. This involves mapping land-
for understanding the processes that create and maintain form types that are classified by genesis or mechanism,
'landforms.' Unfortunately, for many engineers the only and plotting the known vectors or rates of displacement.
useful product of a geomorphological survey was a map, In the report, the case study presented by Lee & Moore
and in order to understand the dynamics of the land- demonstrates the application of this powerful new tech-
scape they would turn to other disciplines such as soil nique that clearly has many potential applications in
mechanics, hydraulic engineering and hydrology. How- planning and engineering.
ever, the case studies by Lee & Brunsden (coastal sediment In addition to these broad themes, there is a wide range
dynamics) and Edmonds (subsidence in Chalk) demon- of applications demonstrated by the case studies includ-
strate that the broader view that geomorphologists have ing resource survey (Allison), defining the borders of
about landscape development has a significant role to countries in military conflict (Nathanail), and establishing
play in engineering projects. This broader geomorpholo- geological ground models for tunnels (Fookes & Shil-
gical view takes into account space and time, inheritance ston). Whilst the case studies presented in the second
of long-term trends, evolution of landscape, extreme working party report do not cover a full range of cli-
events, and environmental change. These are factors that matic regions, examples in the literature demonstrate the
are only now beginning to be recognized as critical in the applicability of the land surface evaluation techniques to
proper investigation of terrain in an engineering context. all types of environmental conditions. For example, glac-
It's for this reason that the second working party envis- ial hazards are not dealt with by the case studies in this
ages a much closer integration of engineering geology report but Reynolds (1998) examines aspects of these
and geomorphology in the future, as demonstrated by problems. Clearly the range and type of applications will
the case studies of the Channel Tunnel (Griltiths), the continue to expand, although it should be recognized that
Cardiff Barrage (Edwards) and the 'Ave' railway study many of the best examples of good practice will remain in
in Spain (Birch). consultants files. Finding time and getting permission to
The major change that has taken place between the publish such work will continue to be difficult.
first and second working party reports, and one that was
expected, has been in the development of new techniques
in land surface evaluation. These now play a role in
engineering at all levels from planning, through ground The future
investigation, design, and construction to remediation.
The first identifiable change is that land surface or In view of the proven success of land surface evaluation
geomorphological evaluations are now being used in over thirty years, it remains a disappointment that many
actual construction work. Martin exemplifies this on the engineering failures still occur because the nature of the
design of remedial works for the Dharan-Dhankuta road natural terrain has not been understood (Hutchinson
in Nepal, and Hearn in the general construction and 2001). Land surface evaluation is an integral compo-
rehabilitation of low-costs roads in mountainous areas. nent in the development of the Fookes (1997) geologi-
The development of techniques to evaluate the perfor- cal ground model and as such is a key aspect of site
mance of structures in the natural environment (Char- investigation for all engineering works. The omission of
man, Carey & Fookes) is a further refinement of this any reference to land surface evaluation, and indeed the
process. These examples demonstrate how the civil works lack of detailed reference to any form of ground mapping,
can only be effective if carried out with a full under- in the latest issue of the Code of Practice for Site Invest-
standing of the natural earth surface processes that are igation BS 5930:1999 (BSI 1999) is worrying and should
active in the contemporary landscape. be rectified.
The next major development has been in the field of Ground investigations are expensive and systematic
hazard studies. Anticipating future geomorphological land surface evaluations remain one of the best ways of
events, particularly extreme events with a very significant ensuring clients get value for their money. Professor P G
environmental impact, has come to be a major part of Fookes in his 1997 Glossop lecture provided the format
land surface evaluation practice. These hazard studies for carrying out site investigations (see Fookes 1997,
have dealt with a range of problems including mass move- figure 42). This format contained one task entitled 'Site
ment (Hearn & Griffiths; Charman; Petley; Shilston), Reconnaissance/and Survey' that, along with local
flood flows and debris flow runout (Hearn, Blong & Hum- knowledge and experience, fed into the preparation of
phreys), ground subsidence (Edmonds; Lee & Sakalas), preliminary geological/geomorphological maps/sections.
and sand and dust (Jones). It is anticipated that hazard Land surface evaluation has to be regarded as critical to
 LAND SURFACE EVALUATION: CONCLUSIONS AND RECOMMENDATIONS 243

this task, but it should also be emphasized that a land References

surface evaluation perspective needs to be included in all
phases of a development. In order to achieve this prac- ANON. 1982. Land surface evaluation for engineering practice.
titioners are advised to review the techniques and case Quarterly Journal of Engineering Geology, 15, 265-316.
studies included in this report as they provide exam- BRITISH STANDARDSINSTITUTION. 1999. Code of Practice for
ples both of the tools that are available and the way they Site Investigations. BS 5930:1999, British Standards Insti-
have been successfully applied to engineering projects tution, London.
FOOKES, P. G. 1997. Geology for engineers: the geological
over the past few decades. model, prediction, and performance. Quarterly Journal of
As a final statement, however, the Working Party came Engineering Geology, 30, 293-424.
to one very clear conclusion. The best way to understand HUTCHINSON,J. N. 2001. Reading the ground: morphology and
the problems of the ground is still to walk over it, to learn geology in site appraisal. Fourth Glossop Lecture. Quar-
to observe, to record the observations and measurements, terly Journal of Engineering Geology and Hydrogeology, 34.
to formulate questions and hypotheses, to design investi- REYNOLDS,J. M. 1998. High-altitude glacial lake hazard assess-
gations, and to discuss the results in an informed way ment and mitigation: a Himalayan perspective. In: MAUND,
using a good scientific rationale. This is the message of J. G. 8~ EDDLESTON, M. (eds) Geohazards in Engineering
Geology. Geological Society, Engineering Geology Special
this book.
Publications, 15, 25-34.
Index

aeolian transport 117, 120-21, 171 9 Caversham, Berkshire 97 105

aerial photographic information sources 19-20 Chalk 29-37, 97-106, 189, 190, 221
aerial photographic interpretation 3, 13, 15, 59, 60, 241 Chalk Marl 129 33
geological and geomorphological mapping 4, 5, 143,160, 210, 241 Channel Tunnel
ground behaviour 107 portal 129-33
hazard mapping 74, 79, 85, 105, 210 rail link 59, 61,221-5
landslide hazard mapping 43, 46, 49, 91, 144, 145, 154, 227, 230 terminal 53
railway design and construction 85, 221,225 checklists s e e proformas
road construction and rehabilitation 135, 137, 138, 140, 210 Cheshire 193-5
sediment transport 182 Chesil Beach, Dorset 181, 182, 187
subsidence hazard 32, 105 Chilterns 30
aerial photography s e e a l s o photogrammetry; remote sensing classifications 3-4
aggregate resource maps 55 building damage by salt weathering 167
aggressive ground 162 environmental performance of structures 23 8
Algeria 115-21 landslide hazard and risk 144
alluvium 107, 221 mountain land 92
anthropogenic fill 107 subsidence hazard 105 6
Applied Earth Science Mapping (DOE) 4 Clay with Flints 30, 31
Arabian peninsula 70; s e e a l s o Saudi Arabia Cleveland, UK 215 18
Arc-Info 68 climate/weather 67-8, 94-5, 138, 197, 210; s e e a l s o tropical
ArcView 151 environments
arid zone coalfield hazards 73-82
ground characterization 115-21 Coalport, Shropshire 227-30
pipeline routing 115-21 coastal environments 23-8, 159 69, 181 7, 189 92
water resources survey 65-71 Code of Practice for Civil Engineering No. 1
s e e a l s o deserts; drylands (1950/1957) 3
Australia 59-61 Code of Practice for Site Investigation s e e BS 5930
conceptual models 109, 110, 171, 199 200, 207, 22l 5
Bahrain 53, 70 construction materials 23
Balkans 205 8
contaminated land 236
Barbados 151-6
contamination potential of groundwater 106
basalt lava flows 66-7 Coombe Rock 129-33
Berkshire 97-105
CP2001 3
Bhutan 91-5
Crag 31
black box method 46
Black Ven, Dorset 14, 15, 16
data collection 40-41
Blackgang, Isle of Wight 53
data handling 5
Blackheath Beds 31
data types 57
block diagrams 125, 126, 224
databases
border marker posts 205, 207 ground and groundwater behaviour 113
borehole data
landslides 49, 144, 154
information sources 20, 126 natural cavities in chalk 29-37, 105
investigations 79, 121, 130, 144, 160, 225, 235, 237 s e e a l s o GIS
boreholes s e e a l s o wells
Dawlish, Devon 28
Bosnia-Herzegovina 205-8
dayas 118
boundary delineation 205, 207
deserts
brickearth 31
blown sand and dust hazard 171-9
bridges 28, 227-30 engineering geological/geomorphological mapping 159-69
British Geological Survey 20, 81, 126
terrain system mapping 61
BS 5930 (1981) 4, 39, 53 s e e a l s o arid zone; drylands
BS 5930 (1999) 19, 39, 242
desk studies 5-7, 19-20, 233, 235, 241
buildings abandoned salt mine subsidence 193
blowing sand and dust 171-9
assessment of slope instability types 215 18
salt weathering 162, 167
assessment of unstable ground 189 90
and slope instability 215-18 and the geological model 123, 127
s e e a l s o structures; urban areas
GIS-based landslide hazard mapping 151-6
Cambrian rocks 83 ground models in an urban area 107-10
canals 81, 91 military purposes 2 0 6 ~
Cardiff Bay barrage 107, 112 pipeline routing 74, 76, 115
Cardiff City, ground model construction 107-13 railway bridge slope stability 227-30
case studies index (cross-tabulation) 6 railway construction 85, 233-7
Castle Hill Landslide, Kent 129-33 road construction and rehabilitation 135
246 INDEX

desk studies ( c o n t i n u e d ) geochemistry 69

shingle transport 18, 184 7 geohazard mapping 37, 55; s e e also hazard mapping
water resources 67-8 Geological Society 20
determinability of geology 123, 124, 127 geology
Devon 26--8, 53 geological domains 125, 126
Devonian 126-7 geological models 107-10, 123-7, 241; s e e also modelling
Dharan-Dhankuta Road, Nepal 197-203 information sources 20, 227, 230, 233, 235, 236
digital elevation models (DEMs) 13--17 and landslide hazard assessment 92-4
Dinarid mountain chain 206 maps and mapping 59, 60, 143, 205, 221
direct mapping 46 see also engineering geology
Docklands Light Railway, London 233-7 geomorphological mapping 3, 37, 50, 53-5, 241
Dorset 14, 15, 16, 30, 35, 181-7 assessment of unstable land 189-90, 209-13
drainage 55, 68, 69 copper mine landslide hazards 143--9
former paths 30, 31, 32, 33, 97 for high pressure gas pipeline 73-82
drylands 53; s e e also arid zone; deserts modelling sediment transport 182, 183---7
dune mobility 117, 120-21; s e e also sands railway construction audit 81, 85, 87
durability of coastal structures 26 redidential/urban areas 107, 215-18
duricrusts 115, 117-19 subsidence hazard 97-106
dust hazard 171-9 geomorphology 4, 227, 230, 242; s e e also engineering
dust drift potential map 177-8 geomorphology; geomorphological mapping
dust rose diagrams 174-5, 177, 178 geophysical techniques 79; see also s p e c i f i c t e c h n i q u e s
geostatistical techniques 59
earthquakes 49, 138, 141, 143, 197, 209, 210, 241 geotechnical mapping 4
East Anglia 30, 35 GIS (Geographical Information Systems) 5, 20, 57- 8, 241
Egypt 53, 159-69 and landslide susceptibility mapping 43, 46, 151--6
electrical conductivity tests 68, 69, 162, 166-9 and terrain system mapping 59
embankment settling 89 water resources survey 68
engineering geology 117, 227, 230, 242 glacial deposits 30, 31, 107, 193, 215; see also periglacial features
engineering geological mapping 4, 39- 41, 138- 40, 143-5, glacial hazards 242
159 69, 199, 227-30 GPS (Global Positioning Systems) 17, 54
and the geological model 123-6, 129 33, 221 gravels 15, 30, 31, 83, 87, 105-6, 221
engineering geological plans 39, 40, 110--11, 113, 137 Greece 81
s e e also geology
Greensand 189, 190
engineering geomorphology 4, 55, 115 Greenwich Foot Tunnel 237
engineering geomorphological mapping 4, 135-40, 159-69, 221 ground behaviour mapping 190-92, 242
s e e also geomorphology
ground characterization 115 --21
Environment, Department of the (DOE) 4, 33 ground models s e e modelling
environmental information sources 233, 235, 236 ground truthing 79, 85-7, 115, 121, 143, 225, 230; see also field
environmental performance classification 23-8 reconnaissance; walkover surveys
erosion 53, 91, 171 groundwater
failures, engineering 89, 200, 202, 242; s e e also landslides; slope behaviour modelling 107-13
instability contamination potential 106
faulting 125, 126, 127, 138, 151 hazard 235
feasibility studies 61, 91, 143, 189, 209 hazard intensity map 166--9
field mapping 4 landslide hazard assessment 95, 215
abandoned salt mine subsidence 193-5 resources 65-71,207; see also wells
assessment of slope instability types 215-18 gypsum heave 118
assessment of unstable ground 189-90
rail link design and construction 221,223 hazard assessment
for remedial roadworks 199, 203 hazards to construction work 235 7
sediment transport 182 landslides 91-5, 135-41, 190-92, 209-13
field reconnaissance 79, 85--7, 115, 121,233 railway bridge slope 227-30
and the geological model 123, 127 using ground models 112-13
GIS-based landslide hazard mapping 154 see also risk assessment
railway construction audit 87 hazard mapping 81-2
road construction and rehabilitation 135, 136-7, 138 landslides 43-52, 55, 83-9, 138, 143-9, 151- 6, 190-92, 210-13
vehicular 121, 173, 205 pipeline routing 73- 82, 115, 121
s e e also ground truthing; walkover surveys S Wales coalfield valley 75-81
flood hazard potential 161, 165, 166 solution features 29-37, 97-106
fluvio-glacial sand and gravel 107 hazard potential, near-coastal deserts 159-69
flysch 206 hazard studies 242
folding 151 helicopter reconnaissance 121
Folkestone, Kent 53, 129-33 hill canals 91
France 53 Hill Gravel 97-105
Himalaya 61, 91-5, 135-8, 140-41
gas pipeline routing 73-82, 115-21 historical data 19-20, 160, 233-7
Gault Clay 129-33, 189, 190 and ground models 107, 129, 133
 INDEX 247

and hazard mapping 47, 129 33, 144, 215 Modflow 110
and shingle transport 184-7 morphological mapping 53-4, 193-5; s e e also geomorphological
Hong Kong 49, 50, 51 mapping
Hualushi, Taiwan 210-13 mountainous areas
hydrogeological information sources 233, 235, 236 copper mine hazard mapping 143-9
hydrological survey 65-71 hill canals 91-5
Hythe Beds 223 military activities 205-8
road construction and rehabilitation 135-41,
indirect mapping 46
197-203, 209 13
infiltration tests 68, 69
multispectral imaging 5, 105, 241
information sources 19-20, 233
military mapping, Balkans 205-6, 207 national park development plan, Taiwan 210-13
pipeline routing 74 National Remote Sensing Centre 20
railway bridge slope stability 227-30 Negev Desert, Israel 70
railway construction audit 85 Nepal 46, 47, 50-51, 53, 91-5, 197-203
rock tunnel geological model 123-7 New Jersey, USA 37
water resources, Jordan 67-8 Northwich, Cheshire 193-5
see also databases
irrigation canals 91 Oddicombe, Devon 26-8
Isle of Wight 4, 53, 55, 189-92 Ok Tedi Copper Mine, PNG 53, 143-9
isotope analysis 69 Oman 70, 71
Israel 70 ophiolites 206
Jordan 65-71 Ordnance Survey 19, 20
orthophotos 15, 16
karst 34-7, 97-106, 206
Kent 53, 81, 129-33 Papua New Guinea 47, 48, 53, 143-9
land classification 3, 241 Pennsylvania, USA 37
land instability modelling 37 performance modelling 23-8
land surface evaluation, definitions 3 periglacial features 30, 31, 75, 97, 100, 221; see also glacial deposits
land use, and landslide hazard assessment 95, 189-92 Peru 81
landscape development 242 Philippines 135, 138-40
landscape hazard mapping 154 6 photogrammetry 5, 13-17, 59, 241
landslides 37, 151 photography 5, 54; s e e also aerial photography
conceptual models 199-200, 215-18 photomosaic maps 87, 88, 172, 176
development of ground model 129-33 pipeline routing 73-82, 115-21
GIS-based hazard mapping 151-6 planning guidance 4, 106, 151
ground behaviour mapping for land use planning 189-92 planning information sources 19, 233, 235, 236
hazard assessment 91-5, 190-92 Plateau Drift 30, 31
hazard mapping 43-52, 55, 83-9, 138, 143-9, 151-6, 190-92, Poland 53
210-13 probability 47
management strategy 192 process units 181
photogrammetry 14, 15, 16 proformas
remedial roadwork design 135, 138 40, 202-3 coastal structure performance rating 23-6
risk assessment 4, 43-52, 143-9 landslide hazard assessment 91-5, 199, 203
laser scanning 13, 241 near surface materials 54
Libya 61 s e e also record sheets
limestone subsidence 37 proto-Solent 30, 33
Lincolnshire 30 proto-Thames 30, 33, 97
literature review 205-6; see also desk studies; information sources PUCE terrain classification 3
London 221,233-7 Purbeck Downs 30
longshore transport model 181-7
Lyme Bay, Dorset 181-7 quantitative risk assessment (QRA) 43, 46-7, 49
quartzite 200
mapping, information sources 19-20; s e e also field mapping; a n d
specific t e c h n i q u e s radar remote sensing 13, 241
marls 83, 87, 129 33 railway bridge slope stability 227-30
MEXE system of classification 3 railway design and construction 61, 81, 83 9, 221-5
military activities 3, 5, 159, 160, 166, 171,205 8 railway maintenance strategy 82
mining areas 37, 79, 143-9, 193-5, 236 railways s e e also Channel Tunnel
modelling 5, 37 raster data 57
landslides in remote areas 91-5, 135-41, 199-200 Reading Beds 30, 31, 97-105
for military purposes 205-8 record sheets 85, 86; s e e also pro formas
pipeline routing 115 remedial works 28, 81, 135, 138 40, 197-203
rail link 221-5 remote sensing 5, 13-17, 241
rock tunnel 123-7 information sources 20
sediment transport 182, 183-7 of karstic spatial patterns 34-7
urban areas 107-13 and landslide hazard mapping 43, 46
see also conceptual models sand drift 171, 172, 175-6, 179
248 INDEX

remote sensing (continued) stereoplotting 17; see also aerial photographic interpretation;
and terrain system mapping 59, 60 photogrammetry
see also aerial photography; satellite imaging structures 23-8, 61, 89, 200, 202
residual soils 61 subsidence 33, 37, 97-106, 193-5
resource maps 55 subsurface investigations 54, 79, 129-33, 138, 140, 160, 163-4; see
retaining wall failure 200, 202 also specific techniques
Ripon, Yorkshire 53 Suez City, Egypt 53, 159 69
risk assessment 112-13 surface materials map 160-61
coastal environments 28 surface processes, recording 55
landslides 4, 43 52, 143-4 Syria 70, 71
risk management 242
rivers Taiwan 209-13
former courses 30, 31, 32, 33, 87-9, 97 tectonism 206
management 53 Tees Valley 215-18
riverbank erosion 15 terrain systems 3, 59-61, 91-5, 107, 115 21, 135, 205-8, 221 5;
roads see also modelling
geomorphological mapping 53, 135-40, 209 13 Thames Valley 97-100
landslide hazard mapping 46, 47, 50 51,210-13 Thanet Beds 31
mountainous areas 135-41, 197-203 thematic mapping 20
remedial works on unstable land 81, 135, 138-40, 197 203 topographic maps and information 20, 32, 53, 205, 227, 230, 233
routing and construction 135 8, 140-41 Torbay, Devon 53
terrain system mapping 61 Transport & Road Research Laboratory 3, 135, 137
rock falls/avalanches 140-41, 143, 144, 145, 200, 202 trench backfill availability 118
Royal Society, R i s k 51 trench excavatability 117, 119-20
trial pitting 79, 130, 133, 138, 160, 163, 199, 225
salt mines, subsidence mapping 193-5 triggering mechanisms 49
salt weathering potential map 162 tropical environments 61, 143, 151,209
sands tunnels 81, 123-7, 129-33,223, 235, 237
blowing sand and dust hazard 171-9 typhoons 138, 210
drift potential map 177-8 UNESCO 39
geological/geomorphological mapping 159 69, 172-3, 176 urban areas 107 13, 233-7
sand rose diagrams 174, 177 urban developments 159 69
satellite imaging 59, 60, 241 USA 37, 70
military purposes 205
pipeline routing 115 Valley Loam 97 105
rail link design and construction 221-3 vector data 57
see also remote sensing Ventnor, Isle of Wight 4, 55, 189-92
Saudi Arabia 71, 171-9 vulnerability 43
seabed remote sensing techniques 241
sediment budget analysis 181 7 wadi channels 68 70, l l8, 161, 165, 166, 171
sediment cells 181, 183-6 Wales 53, 73 82, 107-13
seismic profiling 241 walkover surveys 32, 54, 233, 243
Severn Levels, ground models 107 10 landslide hazard in remote areas 91 2
shear surface types 130-33 miltary purposes 205
Sherwood Sandstones 215 railway bridge slope stability 227, 230
shingle transport 181 7 rock tunnel geological model 125, 126
Shropshire 227-30 sand and dust drift potential 173
site investigation 19 20, 107 13; see also specific techniques urban and industrial areas 233
see also field reconnaissance; ground truthing
sketch mapping 39, 203, 230
slope instability 81, 118, 209-13, 215-18, 230; see also water resources survey 65-71
water supply boreholes 205, 207; see also wells
landslides
slope profiles 68 weather see climate/weather
software, photogrammetric 15, 17; see also GIS weathering 61, 126, 127, 162, 167, 171-9
soils wells 65-6, 68, 69-70, 205, 207
erosion 53 West Bay, Dorset 181-7
white box method 46
residual 61
saline 161 2 wind erosivity 171
slope failure 200, 202 wind speed data 173-4, 176
Woolwich-Reading Beds 31
surveys and mapping 20, 53, 68
Working P a r t y Report on Land Surface Evaluation
solifluction deposits 31, 55, 129
solution features 29 37, 97 105 (lst, 1982) 3-4, 59, 241,242
Working P a r t y Report on M a p s and Plans (lst, 1972) 39, 41, 53
sonar surveys 241
Spain 81, 83 9 Yarm, Cleveland, U K 215-18
spatial analysis 57, 241 Yorkshire 30, 53
landscape hazard mapping 154-6 Yugoslavia (former) 205-8
natural cavity occurrence 29-37
stereo-satellite imagery 15, 241 zoning maps 39 40
 Land Surface Evaluation for Engineering Practice
Geological Society Engineering Geology Special Publication No. 18

edited by

J. S. Griffiths
(University of Plymouth, UK)

This volume presents a collection of papers on techniques and case studies in land surface
evaluation for engineering practice written by specialist practitioners in the field. The volume
arose out of deliberations by the Second Working Party on Land Surface Evaluation set up by
the Engineering Group of the Geological Society in January 1997 and chaired by Dr J. S.
Griffiths. The book provides examples of cost-effective methods for collecting land surface
and near surface data prior to carrying further detailed ground investigations of engineering
sites. The book should be of interest to practising engineering geologists, geotechnical
engineers, geomorphologists and planners who have the responsibility for planning and
designing investigations of potential sites of development.

248 pages
over 126 illustrations; several in colour
35 papers
index

Visit our on-line bookshop: ISBN 1-86239-084-3

http://bookshop.geolsoc.org.uk

Cover illustration: Reconnaissance of the Ok Ma landslide near

Tabubil, Papua New Guinea. Photograph provided by Professor D. Brunsden.
!JI!