GS+ Users Guide

Version 7

GS

GeoStatistics for the

Environmental Sciences
G A M M A

D E S I G N

S O F T W A R E

Accessible Geostatistics for Everyday Science

GS

GeoStatistics for
the Environmental Sciences
Version 7

Gamma Design Software, LLC

Plainwell, Michigan 49080
800/873-0690 phone
269/685-0910 fax
www.gammadesign.com

Copyright
Copyright 1989-2005 Gamma Design Software, LLC. All Rights Reserved
ISBN number 0-9707410-0-6
Information in this document is subject to change without notice and does not represent a commitment on the part of Gamma Design Software. The software described
is provided under a license agreement and may be used or copied only as specified
in the agreement. No part of this document may be reproduced in any manner
whatsoever without the express written permission of Gamma Design Software.
Gamma Design Software
P.O. Box 201
Plainwell, Michigan 49080
U.S.A.
Citation
The appropriate citation for this documnent is
Gamma Design Software. 2004. GS+: Geostatistics for the Environmental Sciences.
Gamma Design Software, Plainwell, Michigan USA.

Trademarks
Microsoft and Windows are trademarks or registered trademarks of Microsoft Corporation. Surfer is a registered trademark of Golden Software, Inc. ArcView and
Arc/Info are registered trademarks of ESRI, Inc. Other brands and their products
are trademarks or registered trademarks of their respective holders and should be
noted as such. GS+ is a trademark of Gamma Design Software.
February 2004
October 2004 rev

Table of Contents

Table of Contents
Chapter 1
Introduction
Overview ..............................................................................................1
System Requirements ..........................................................................2
Installation ............................................................................................2
Updates ................................................................................................2
Licensing and Copy Protection ............................................................3
Single-User License Agreement ..........................................................3
Chapter 2
Getting Started
From Data to Maps: How to Proceed...................................................5
General Screen Layout ........................................................................5
Menus...................................................................................................6
User Preferences ...............................................................................13
Graph Settings ...................................................................................17
Printing Graphs ..................................................................................23
Printing Data or Text ..........................................................................24
Using Older-Version GS+ Files...........................................................24
Chapter 3
Working with Data
The Data Worksheet Window ............................................................25
Column Assignments .........................................................................28
Covariate Values Warning..................................................................29
Missing Values ...................................................................................29
Data Filter Dialog................................................................................30
Duplicate Values ................................................................................31
Chapter 4
Importing Data from External Files
File Import Dialog ...............................................................................33
Importing Text Files............................................................................34
Text Input File Formats
GS+ .............................................................................................36
GeoEas.......................................................................................37
ArcView XYZ.............................................................................38
Surfer XYZ ................................................................................39
Spreadsheet and Database Input Files..............................................40
Input File Formats (File Import Properties) ........................................41
i

Table of Contents
Viewing Files (File View Window) ..................................................... 42
Appending Data to an Existing Worksheet (Data Append Dialog).... 42
Chapter 5
Summary Statistics
Z Variate Summary............................................................................ 43
Frequency Distributions
Frequency Distribution Histograms ........................................... 45
Cumulative Frequency Distribution............................................ 47
Normal Probability Distribution .................................................. 49
X,Y Coordinates Summary ................................................................ 51
Coordinate Postings (Quantile Plots)
2-Dimensional Data ................................................................... 52
1-Dimensional Data ................................................................... 54
Defining Posting Intervals .......................................................... 55
Regression Analysis Tab................................................................... 56
Cross-Variate Regression Window............................................ 57
Regression Values..................................................................... 58
Chapter 6
Semivariance Analysis
Overview ................................................................................................... 59
The Autocorrelation Window ..................................................................... 60
Define Nonuniform Lag Class Intervals............................................. 64
Variograms Window .......................................................................... 66
Semivariance Values Window........................................................... 67
Isotropic Variogram Models............................................................... 68
The Spherical Isotropic Model ................................................... 71
The Exponential Isotropic Model ............................................... 72
The Linear Isotropic Model ........................................................ 73
The Gaussian Isotropic Model................................................... 74
Anisotropic Variogram Models .......................................................... 75
The Spherical Anisotropic Model............................................... 77
The Exponential Anisotropic Model ........................................... 78
The Linear Anisotropic Model.................................................... 78
The Gaussian Anisotropic Model............................................... 79
Anisotropic Semivariance Surface (2D Variogram Map)................... 80
Cross-Autocorrelation or Cross-Semivariance Analysis ................... 83
Chapter 7
Other Autocorrelation Measures
Standardized Variograms .................................................................. 85
Madograms........................................................................................ 86
ii

Table of Contents
Rodograms.........................................................................................87
Drift ....................................................................................................88
Correlograms......................................................................................89
Covariance Analysis...........................................................................90
General Relative Variograms .............................................................91
Pairwise Relative Variograms ............................................................92
Morans I Analysis ..............................................................................93
Fractal Analysis ..................................................................................94
Chapter 8
Variance Clouds and h-Scattergrams
Overview ............................................................................................95
Variance Clouds .................................................................................95
Variance Cloud Pairs .........................................................................99
h-Scattergrams ................................................................................100
h-Scattergram Pairs ........................................................................104
Chapter 9
Interpolation Basics
Overview ..........................................................................................105
The Interpolation Window ................................................................107
Z Estimate Boundaries .....................................................................109
Defining a Uniform Interpolation Grid ..............................................110
Defining a NonUniform Interpolation Grid ........................................112
Polygon Outlines (Interpolation Masks) ...........................................114
Polygon Outline Map .......................................................................116
Cross-Validation Analysis ................................................................117
Cross-Validation Values ..................................................................118
Chapter 10
Conditional Simulation
Overview ..........................................................................................119
Secondary Data for Simulations.......................................................121
Number of Simulations .....................................................................122
Chapter 11
Kriging and Cokriging
Kriging ..............................................................................................123
Cokriging ..........................................................................................125
Chapter 12
Inverse Distance Weighting
Inverse (IDW) and Normal (NDW) Distance Weighting ...................127
iii

Table of Contents
Chapter 13
Interpolation Output File Formats
GS+ (.krg) Format ............................................................................ 129
ArcInfo or ArcView (.asc) Format............................................... 130
Surfer (.grd) Format ...................................................................... 131
Chapter 14
Mapping
The Mapping Window ..................................................................... 133
Map Contour Intervals .................................................................... 137
3D Maps ......................................................................................... 139
Standard Deviations ............................................................... 140
Rotation ................................................................................... 141
2D Maps ......................................................................................... 142
Standard Deviations ............................................................... 142
Sample Posting ....................................................................... 143
Transects (1d Maps) ....................................................................... 144
Standard Deviation ................................................................. 144
Chapter 15
Bibliography ......................................................................................... 145
Chapter 16
Glossary ................................................................................................. 147
Index

iv

................................................................................................. 155

Chapter 1 Introduction

Chapter 1
Introduction
GS+ is a GeoStatistical Analysis program that allows you to readily measure and
illustrate spatial relationships in geo-referenced data.
GS+ analyzes spatial data for autocorrelation and then uses this information to make
optimal, statistically rigorous maps of the area sampled. The maps can be created in
GS+ or in other mapping programs or geographic information systems.

When do I need geostatistics?

You need geostatistics when you need to make accurate, statistically rigorous maps
created from incomplete data which means whenever you make a map for a property that cannot be exhaustively sampled. Whether you are mapping oil deposits or
plankton distributions, geostatistics allows you greater confidence in the interpolated
values for the locations not actually sampled.

Statistics Provided by GS+

GS+ provides spatial autocorrelation analysis:
Semivariance analysis produces variograms and different types of variogram
models, including isotropic and anisotropic variograms;
Anisotropic variogram maps make it easy to recognize anisotropy;
h-Scattergrams and Variance clouds provide an easy way to recognize data
outliers; and
Many other types of autocorrelation measures, including Moran's I, fractals,
correlograms, covariograms, madograms, rodograms, drift, standardized
variograms, and general and pairwise relative variograms.
GS+ provides fast interpolation:
Block kriging and punctual kriging provide optimal interpolation of a discrete
point or an area around a sample point location;
Conditional simulation provides probability based interpolations and estimation error;
Cokriging provides optimal interpolation when you have only a few samples
for the primary variate but lots of samples for an easier-to-measure, related
variate; and
Inverse distance weighting provides simple nearest neighbor interpolation
based only on distance to nearby samples.
1

Chapter 1 Introduction
GS+ provides basic parametric statistics:
Sample means and variance;
Frequency distributions, probability distributions, and skewness and kurtosis
measures for determining departures from normality; and
Quantile plots or coordinate maps show the distribution of sample values
across the spatial domain;
Transformations for returning the data to normality; and
Regression analysis for covariates vs. primary variates.

System Requirements
PC Compatible with Pentium or faster processor
Windows NT4 / 2000 / XP or higher Operating System
A minimum of 40 MB of free hard disk space
A minimum of 64 MB RAM

Installation
To install GS+:
1) Insert the CD-ROM or download the installation file to a temporary folder
2) From the Windows Start Button, click Run
3) Type g:\Setup [use your CD-ROM drive letter for g:]
4) The Setup program will prompt you through the installation process. Follow
the instructions on the screen. The serial number for your copy of GS+ can
be found on the GS+ CD-ROM package.

Updates
Maintenance updates are available free of charge to registered users. Update files
are available by download only from http://www.gammadesign.com. The current
version of GS+ can be checked from the Help menu as described in Chapter 2.

Chapter 1 Introduction

Licensing and Copy Protection

Copy Protection
The standard version of GS+ is not copy-protected. We rely on national and international copyright law and the integrity of our users to abide by the license agreement
printed on the CD or diskette envelope and below. This agreement limits you to
installing your serialized copy of GS+ on only one computer unless you have a
site license.
Site Licensing
If you have reason to install GS+ on more than one computer in the same laboratory,
or on a network that allows more than one user at a time to access the program,
please contact Gamma Design for information about converting your single-user
license to a site or classroom license. It is a violation of your single-user license
agreement if the program resides on more than one computer.

Single-User License Agreement

Gamma Design Software License Agreement
Please read carefully; this is a legal agreement between you (the end user) and
Gamma Design Software, LLC (Gamma Design). When you break the seal on the
software media you signal your agreement to be bound by the terms of this agreement, including the Software License and the Limited Warranty. If you do not agree
to be bound by the terms of this agreement, do not open the package and return the
package together with accompanying written material to Gamma Design at the address below for a full refund.
SINGLE USER Software License
+
1. Gamma Design retains ownership of the GS program enclosed. Gamma Design
gives you (the end user) the right to use a single copy of GS+ on a single computer.
+
You may use GS on a network or file server ONLY if access is limited to one user at
a time and you have the original copy of the documentation and program disks. You
do not have the right to install or use GS+ on more than one computer, hard disk
drive, or file server at a time.
2. GS+ is owned by Gamma Design and is protected by United States copyright laws and
international treaty provisions. GS+ must be treated like any other copyrighted mate+
rial although you may either 1) transfer GS to a single hard disk drive so long as you
+
keep the original copy for the purpose of backup, or 2) make one copy of GS for
+
backup purposes. The written material accompanying GS may not be copied.
3. GS+ may not be rented or leased, but may be permanently transferred if you keep no
copies of any version of GS+ and the recipient agrees to the terms of this agreement.
4. You may not decompile, disassemble, or reverse-engineer GS+.
5. Gamma Design retains all rights not granted expressly herein. Nothing in this Agree-

Chapter 1 Introduction
ment constitutes a waiver of Gamma Designs rights under any federal or state law.
Limited Warranty
1. Gamma Design warrants that GS+ will conform substantially to the accompanying writ+
ten materials for a period of 1 year from the date of purchase, provided that GS is
used on computer hardware and with the operating system for which it was designed.
2. Gamma Design disclaims all other warranties, either express or implied, including implied warranties of merchantability and fitness for a particular purpose. This applies
to both the software itself and accompanying written materials. This limited warranty
gives you specific legal rights; you may have others that vary from state to state.
3. Under no circumstances shall Gamma Design be liable for any damages whatsoever
arising out of the use of or inability to use GS+, even if Gamma Design has been advised of the possibility of such damages. Such damages include but are not limited
to damages for loss of profits or revenue, loss of use of the software, loss of data, the
cost of recovering such software or data, the cost of substitute software, or claims by
third parties. In no case shall Gamma Design be liable for more than the amount of
the license fee, as set forth below. Some states do not allow the exclusion or limitation of liability for consequential or incidental damages, so this limitation may not apply to you.
User Remedies
1. Gamma Designs entire liability and your exclusive remedy shall be, at Gamma Designs discretion, either (1) refund of the purchase price or (b) replacement of the
software that does not meet Gamma Designs limited warranty. In either case software must be returned to Gamma Design with a copy of the sales receipt. This warranty is void if failure has resulted from accident, abuse, or misapplication. Any replacement will be warranted for one year.
2. The warranties and remedies set forth above are exclusive and in lieu of all others,
oral or written, express or implied. No Gamma Design Software distributor or employee is authorized to make any modification or addition to this warranty.
U.S. Government Restricted Rights
+
GS software and documentation are provided with RESTRICTED AND LIMITED RIGHTS.
Use, duplication, or disclosure by the Government is subject to restrictions as noted in subparagraph (c)(1)(ii) of The Rights in Technical Data and Computer Software clause at 52.2277013. The manufacturer is Gamma Design Software, P.O. Box 201, Plainwell, MI 49080.
General
You must fill out and return the Warranty Registration Card to be eligible for customer support
and service. If you have questions about this agreement, write to Gamma Design Software,
P.O. Box 201, Plainwell, MI 49080, U.S.A.

Chapter 2 Getting Started

Chapter 2
Getting Started
From Data to Maps: How to Proceed
To make a map using GS+:

First, collect samples from known locations. The sample locations do not
need to be evenly spaced or even to lie on a grid, you simply need to know
their location in a Cartesian (x,y) coordinate system;

Second, bring the data into the GS+ Data Worksheet; you can enter the data
directly into the worksheet or import the data from a text file, spreadsheet, or
another source; often the easiest way to import data is to cut-and-paste from
the source spreadsheet or text file;

Third, perform Semivariance Analysis to produce a variogram model of the

autocorrelation present in the data;

Fourth, use Kriging or Conditional Simulation to produce an interpolation file

that will contain optimal estimates of values at evenly-spaced intervals over
the sample area; and

Finally, draw a 3D or 2D Map of the property. This map will be an optimal,

unbiased representation of the property over the area of interest. You can
also produce a confidence map for the estimates, which will allow you to assess the statistical error associated with each estimated contour interval.

General Screen Layout

The main GS+ window has a command menu at the top and holds each of the individual analysis windows that are currently open. The main windows are:

The Data Worksheet Window

The Data Summary Window

The Autocorrelation Window

The Interpolation Window

The Map Window

Chapter 2 Getting Started

Main Menu
The Main Menu presents access to the windows that provide GS+ analyses. Below
the command menus are icons that represent short-cuts to many of these functions.

Commands and toolbar buttons that are shaded out are not selectable. For example, the print icon is not available when viewing a window without a graphic or worksheet to print. The six covariate icons for covariate functions are not available
unless a covariate has been defined in the Data Worksheet window.
The tool bar is customizable via the Edit menu or by right-clicking on the main menu.
You may add or remove shortcut icons for most analyses. It is also movable and
reconfigurable; it can be dragged to any spot on the screen with the mouse.

The File Menu

The File Menu provides commands for saving and retrieving GS+ parameter files,
used to store and retrieve analysis settings, and also commands for printing and
setting user preferences.

New File Clears existing analysis parameters.

Open File Allows the user to load an existing parameter file. To open a text
data file, use the Import Command in the Worksheet window.

Save File Save the existing parameter file.

Save File As. Save analysis parameters in a file to be named.

Chapter 2 Getting Started

Print Prints a spreadsheet or graph.

Printer Setup Make changes to the print format

Preferences Set user preferences to be saved from session to session

Exit Exit GS+

The Edit Menu

The Edit menu provides access to the cut-copy-paste-delete editing commands.
These commands are available whenever the cursor is in an editable field within a
particular window.

Cut remove selected material to the clipboard

Copy copy selected material to the clipboard

Paste paste the clipboard into the selected area

Delete delete selected material

Select Select a worksheet row or column. If the cursor is not in a worksheet

this command will be dimmed.

Toolbar customize the toolbar to contain a variety of different tool shortcuts

Chapter 2 Getting Started

The Data Menu
The Data menu provides access to the data worksheet and to summary statistics
windows, and provides commands for importing and exporting data files and for manipulating data within the worksheet.

WorkSheet display the data worksheet window

Assign Column assign a variate to a data sheet column

Sort sort the highlighted column; the sort alternates between ascending
and descending order. If the cursor is not in a worksheet cell the sort command will be dimmed, as it is here.

Change Decimals increase or decrease the number of decimal places to

show in a particular column; this has no effect on calculations. The cursor
must be in an active worksheet window.

Insert insert a row or column into the active worksheet

Delete delete a row or column in the active worksheet

Import Data import a data file into the active worksheet

Export Data export the contents of either the active worksheet or the active
data arrays to a text file. The active data arrays are the non-missing data in
the ID, X-Coordinate, Y-Coordinate, and Z-data columns of the Data Worksheet window. If the Z data have been transformed (using a transformation
command), the transformed z values will be printed in addition to the original
z values in the Data Worksheet.

Summary Statistics display basic statistics for the primary variate (Z) and
covariate (Z2) if defined.

Chapter 2 Getting Started

The Autocorrelation Menu
The Autocorrelation menu provides access to one of several types of autocorrelation analyses in GS+:

Variogram

General Relative Variogram

Correlogram

Madogram

Covariance Analysis

Morans I Analysis

Drift

Pairwise Relative Variogram

Fractal Analysis

Standardized Variogram

For each analysis you may choose to analyze the primary variate Z, the covariate
Z2, or the cross-variate Z x Z2, although cross-variate analysis is not available for
every autocorrelation measure. If Z or Z2 are undefined the menu command will be
dimmed.

Chapter 2 Getting Started

Interpolate Menu
The Interpolate menu provides access to one of the two types of Interpolation provided by GS+.

Kriging Analysis display the Interpolation Window with access to Cokriging,

Block kriging, and Point Kriging
Inverse Distance Weighting (IDW) display the Interpolation Window with
access to Inverse Distance Weighting and Normal Distance Weighting interpolation methods
Simulation display the Interpolation Window with access to Conditional
Simulation
The Map Menu
The Map menu provides access to GS+ mapping functions.

3D Map display the Mapping window with 3D mapping selected

2D Map display the Mapping window with 2D mapping selected
1d Transect display the Mapping window with 1D mapping selected; if a 2dimensional map file (one having both an x and a y coordinate) is currently
selected, the 1D option cannot be selected

10

Chapter 2 Getting Started

Window Menu
The Window menu allows you to quickly move to open windows in GS+.

Cascade rearrange all open child windows

Data Worksheet display data worksheet window
Data Summary display data summary window for the primary or secondary
variate
Autocorrelation Analysis display autocorrelation windows for the:
Primary Variate (Z)
Secondary Variate (Z2)
Cross Variate (Z x Z2)
Interpolation display kriging, conditional simulation, and IDW analysis window
Map display mapping window
The Help Menu
The Help menu provides access to GS+ help functions.

GS+ Help display context-sensitive help topic.

11

Chapter 2 Getting Started

Check for GS+ Update (via Internet) send an automatic query to Gamma
Design's web address to see if the version of GS+ currently running has been
updated. If a newer version is available, you will be asked if you would like to
connect to the proper web page for an update.
For this feature to work, your computer must have access to the internet.
Communication is conducted using your browser. You can also check the
GS+ update status manually by checking your programs version (available
from the About GS+ screen) against the version displayed at
www.gammadesign.com.
Go to www.gammadesign.com connect to Gamma Design's home page
through your normal internet provider using your default browser.
Email Gamma Design send an email message to Support@GammaDesign.com using your default email program. For this feature
to work, you must have access to the internet and a default email program
installed.
How to Contact Gamma Design show how to contact Gamma Design Software.
About GS+ display title and copyright screen, and also display the current
GS+ version number.

12

Chapter 2 Getting Started

User Preferences General Tab

The Preferences dialog allows you to set user-default values for some GS+ behaviors. There are three categories available General settings (as described here),
settings for Data File Import, and settings for Analyses.

Reset
The Reset command returns all user-default values on this tab to original default
values. To reset all values on all tabs, use the Global Reset command.
Missing Value Indicator
Specify the value used by GS+ on output files and when importing data files to indicate that a value is missing. Missing values are ignored during data builds and subsequent analyses. This value can be overridden by values specified in individual
dialog windows.
Show tips
Display short explanations of commands and input boxes when mouse hovers over
buttons and boxes.
Places Past Decimal
For different types of variates, allow GS+ to format values automatically or specify
directly the number of places past the decimal to report in windows and printouts.
All calculations are performed on double-precision values regardless of the values
requested here. These values can be overridden by values on specific dialog windows such as the Field Assignment Dialog of the Data Worksheet Window.
Global Reset
Sets all user-default values on all tabs to original (GS+ - defined) default values.
13

Chapter 2 Getting Started

Save/Exit
Close window and keep preference changes.

User Preferences - Data File Import Tab

The Preferences Data File Import dialog window allows you to set user-default
values for importing data files into the GS+ Data Worksheet.

Reset
The Reset command returns all user-default values on this tab to original (GS+ - defined) default values. To reset all values on all tabs, use the Global Reset command.
Filename Extension
Default extension for the data file name specified when importing data files to the
Data Worksheet.
Default Fields
When importing data files, these values indicate which fields to assign initially to different variates. Field (column) assignments can be changed in the Data Worksheet
window.
Default Import File Type
When importing data files, this file type will be the default type.
Places Past Decimal to Show
When importing data files, report this many places past the decimal in the Data
Worksheet cells. This value is overridden when specifying places past the decimal
in the Field (Column) Assignment Dialog of the Data Worksheet window.
Automatically Rebuild Data Arrays
When new data is imported and when columns are reassigned (e.g. when the X14

Chapter 2 Getting Started

Coordinate column is reassigned to column 3 from column 2), the internal data array
used by GS+ must be rebuilt. This can take some time if the array is big because
duplicates must be checked, missing values tested, and autocorrelation boundaries
calculated. The alternative to automatic recalculation is to press the Rebuild button
on the Data Worksheet Window. Check or uncheck these boxes to turn off automatic array rebuilds.
Global Reset
Sets all user-default values on all tabs to original default values.

User Preferences Analyses Tab

The Preferences Analyses dialog window allows you to set user-default values for
some analyses.

Reset
The Reset command returns all user-default values on this tab only to original (GS+
- defined) default values. To reset all values on all tabs, use the Global Reset
command.
Variogram Defaults
Offset start of first lag by interval with this check box GS+ will (when
automatically defining uniform interval lag classes) shift the first lag class to
of its normal separation distance. This can provide better resolved
variograms when there are sufficient pairs of points in shorter separation distance classes. The disadvantage is that if there are few pairs of points for
the shortest distance classes when the first lag is not offset (a common prob15

Chapter 2 Getting Started

lem), there will be even fewer pairs available with the first lag offset. If this is
the case, there will be little if any improvement to the variogram.

Automatically make anisotropic surface map the Surface Map command on the Semivariance Window will be automatically enacted after
variograms are calculated. This will make the exploration of anisotropic relationships more automated but there is a performance penalty as creating a
surface map requires further computation.

IDW Interpolation Defaults

Weighting Power (0-100) this value provides the initial default weighting
power for inverse distance weighting (IDW) interpolation. This value may be
overridden from the Interpolate Window; see the Interpolate Window for further information on this parameter.

Smoothing Factor (>0) this value provides the initial default smoothing
factor for inverse distance weighting (IDW) interpolation. This value may be
overridden from the Interpolate Window; see the Interpolate Window for further information on this parameter.

Default to Normal Distance Weighting (NDW) sets the default IDW interpolation to Normal Distance Weighting (NDW). This value may be overridden from the Interpolate Window; see the Interpolate Window for further information on this parameter.

Interpolation Defaults
Draw map after interpolating check this box to automatically draw the
map of interpolated values after interpolation.
Global Reset
Sets all user-default values on all tabs to original (GS+ defined) default values.

16

Chapter 2 Getting Started

Graph Settings General Tab

The Graph Settings Dialog appears when you edit graphs. The dialog will look
slightly different depending on whether you are editing a bar graph, an x-y scatter
graph (e.g. a variogram), or a 2D or 3D map. In the case of a 3D map, for example,
there will be a place for scaling and renaming the Z axis in addition to the X and Y
axes. In addition to the General tab, there is also a tab for Axis Scaling, Axis Titles
and Labels, and Contour Details (only available for maps).

Graph Colors Tab

You may set background colors for three different parts of the graph. Click on the
color bar to the right of the component name to bring up a Color Dialog Window that
will allow you to change the color of that component; the color of the bar indicates
the current color.
Graph Borders Tab
You may set border styles for the graph and for the legend (if shown). Width refers
to the width of the border and Print allows you to choose to print or not print the border when printing:

17

Chapter 2 Getting Started

Lines Tab
On graphs with lines (for example variogram models) you may specify line pattern,
width, and color:

Graph Title
A title is text that appears at the top center of the graph area. To change the font of
the title press Change, which will bring up a Font Dialog Window.
Graph Legend
The legend appears on maps and quantile postings to indicate the value ranges for
different symbols or colors. Legend text appears at the top of the legend table. To
change the font of the legend title and values press Change, which will bring up a
Font Dialog Window.
Graph Footnote
A footnote is text that appears at the bottom left of the graph area. To change the
font of the footnote press Change, which will bring up a Font Dialog Window. To
have variogram model parameters printed as a footnote to variogram graphs, place
in this field the exact phrase (model results) [do not include quotes]. This notation
is placed in the field by default whenever a new variogram and model are calculated;
you will need to remove it if you do not want variogram model parameters to appear
as a graph footnote.
Apply Now
Press Apply Now to apply any changes made to the graph and keep the Graph Settings dialog window open.
Cancel
Press Cancel to exit the Graph Settings Dialog without applying any changes since
the last Apply Now command.
Exit
Press Exit to close the Graph Settings Dialog window. Any changes made since the
last Apply Now command will be applied to the graph.

18

Chapter 2 Getting Started

Graph Settings - Axis Scaling Tab

The Axis Scaling tab of the Graph Settings Dialog Window allows you to specify how
to scale the graph axes, e.g. how long or short an axis should be, and how the bars
and symbols should look.

X Axis Range and Tick Marks

The X Axis range can be set to automatic or user-defined (manual). If the range is
automatic and the lowest value in the graphed data set is greater than zero, then the
axis range is set to a minimum value of zero and a maximum value of 10% greater
than the highest value in the data set. If the range is automatic and the lowest value
is less than zero, then the axis range minimum is set to 10% less than the lowest
value.
The Number of Labels (bar graphs only; not shown above) refers to the number of
values placed along the x-axis.
The Number of Ticks (x-y graphs and maps only) refers to the number of ticks along
the x-axis. Major ticks are accompanied by labels (values); minor ticks are not labeled and appear between major ticks.
To set the appearance of the labels (font, precision, etc.) see the Axis Titles, Labels
tab.
Y Axis Range and Tick Marks
The y-axis range is set identically to the x-axis range. Note that for maps of 1dimensional transects there is no y-axis; the z-axis represents the vertical dimension.
19

Chapter 2 Getting Started

Z Axis Range and Tick Marks
The z-axis range (on maps) is set identically to the X and Y Axes. Note that a zaxis is only present in maps and transects (1-d maps).
Bars
For bar graphs you may specify the number of bars to be plotted, their color, and
pattern. The Bars section is not shown above.
Symbols
For x-y graphs (e.g. variograms) you may specify the type of symbol (open box,
closed circle, etc.) as well as the size and color of the symbol. The Symbols section
is not shown in the example screen above.

Graph Settings - Axis Titles, Labels Tab

The Axis Titles and Labels tab of the Graph Settings Dialog Window allows you to
specify the text that accompanies each axis and to format the values that accompany the major tick marks.

Axis Format
Use these boxes to set the axis titles and how axis values are formatted. Decimals
refers to the number of places past the decimal to format axis values (e.g. 3.1415
has 4 places past the decimal); check exponential to format the axis value in scientific notation (e.g. 3.14E0). For 3D maps, boxes for the Z-Axis are also provided as
shown above; for x-y and bar graphs, there will be no boxes for the Z-Axis. The font
for axis titles and values can be reset with one of the Change commands.

20

Chapter 2 Getting Started

Graph Settings - Contour Details Tab

The Contour Details Settings tab of the Graph Settings Dialog Window allows you to
edit map contour parameters. Some of these settings are identical to those that appear in the Map Window and changing values here will also change values there; for
other maps (e.g. Anisotropic Surface Maps) this window is the only place where contour details can be set.

Projection
3D displays a 3-dimensional map of the data. The height of the map can
be adjusted by changing the 3D Proportion; the perspective can be adjusted
with the Rotate (Mouse Action) command of the calling window
2D displays a flat, 2-dimensional map of the data
3D Proportion
for 3D projections, sets the height-to-width ratio of the map. For tall, skinny
maps set this value to 2 or higher. For flatter maps, set to less than 0.5.
This setting is not available for 2D projections. This command will not affect
the aspect ratio of the map, i.e. the x and y axes will remain proportional to
one another
Map Surface
Contour lines draws lines between contour intervals
Color bands fills the space between contour lines with different colors; colors can be specified from the Define Contour dialog window by pressing Define Contour Levels, below
Smoothing apply slight smoothing to the data to improve visualization
21

Chapter 2 Getting Started

Solid pedestal for a 3D map fill in the area beneath the surface with a solid
color.
Wireframe for 3D maps, drapes an x-y grid over the surface
Wireframe Weave specifies the density of the wireframe grid. A weave of 0
puts a grid line at every data row and column; a weave of 1 skips one
row/column, a weave of 2 skips 2, etc.
Contour Levels
Number the number of contour levels to put on the map
Define brings up the Map Contour Intervals dialog window that allows you
to set break points for individual intervals and colors for contour bands (see
Chapter 12 for a description of the Contour Intervals dialog)
Map Legend
Show legend displays the contour legend next to the map
Continuous vs. Stepped display legend as continuous scale with values
next to break points (e.g. between color bands) or display legend as stepped
format, with values next to boxed colors
Ceiling
Contour lines project contour lines above the surface onto the ceiling of the
plot (3D maps only)
Color bands project color bands onto the ceiling of the plot (3D maps only)
Floor
Contour lines project contour lines beneath the surface onto the floor of the
plot (3D maps only)
Color bands project color bands onto the floor of the plot (3D maps only)
Grid Lines
X axis place a vertical grid line along the back walls of the 3D plot x axis
Y axis place a vertical grid line along the back wall of the 3D plot y axis
Z axis place horizontal grid lines along the back walls of 3D plots

22

Chapter 2 Getting Started

Printing Graphs
The Graph Print dialog window allows you to specify how you would like graphs
printed to what device or file, where on the page, and how big the image should
be:

Destination
Choose where to send the image. If Printer is checked, the graph will be sent to the
printer specified in the Printer box. If File is checked, the graph will be sent to a file
that is specified in a subsequent Select File dialog window. If Clipboard is checked,
the graph will be sent to the Windows clipboard, from which it can be transferred to
another program via the Edit-Paste command.
Printer and Page Setup
Provides access to printer settings as provided by the operating system, including
printer selection, orientation, and margins. All printers available to your Windows
system are available to GS+. This command is not available unless Printer is the
destination specified elsewhere in this dialog window.
Scale
Set the scale of the finished graph image as % of full page for the printer specified.
The Scale command is not available unless Printer is the selected destination.
Center on page
Check this box to have the graphic centered on the output device. Otherwise the
graph will be printed to the upper left corner of the margin area specified in the Settings command. This box is not available unless Printer is the selected destination.
File Format
Use the File Format list box to choose a graphics file format. File format options
include a standard Windows Metafile (.wmf) format, an Enhanced Windows Metafile
(.emf) format, a standard Bitmap (.bmp) format, a web-ready JPEG format (.jpg),
and a web-ready PNG format (.png) . These commands are not available unless
File is specified as the destination for the image.
23

Chapter 2 Getting Started

Printing Data or Text
The Print dialog window allows you to specify how you would like text data printed.

Destination
Choose where to send the image. If Printer is checked, the graph will be sent to the
printer specified by Settings. If File is checked, the graph will be sent to a file that is
specified in a subsequent Select File dialog window. If Clipboard is checked, the
graph will be sent to the Windows clipboard, from which it can be transferred to another program via the Edit-Paste command.
Range
Choose how much of the data to print. Selection prints only the highlighted portion
of spreadsheets, First Page prints only page 1.
Printer and Page Setup
Provides access to printer settings as provided by the operating system, including
printer selection, orientation, and margins. All printers available to your Windows
system are available to GS+. This command is not available unless Printer is the
destination specified elsewhere in this dialog window.

Using Older-Version GS+ Files: The Convert FileDialog

When loading a parameter (.par) file created with a DOS version of GS+, you will be
queried to convert the DOS file to a GS+ Windows file.

Make backup
Check this box to make a backup copy of the file before converting. The backup
copy will have the same name as the original file except will have the extension .old.

24

Chapter 3 Working with Data

Chapter 3
Working with Data
The Data Worksheet Window
The Data Worksheet contains the data for GS+ analyses. Data can be entered
manually or can be imported via the Import file command. Entered data can be
edited, filtered (or bounded), and can be temporarily or permanently deleted from
subsequent analyses. Field assignments (assigning fields or columns to xcoordinate values, y-coordinate values, etc.) are made in the worksheet window by
clicking on the second row.

Base Input File

The Base Input File is the external file from which worksheet data first were loaded
if none of the data were loaded from an external file then this field will be blank. To
import data to the worksheet from an external file, press Import to bring up a File
Import Dialog window.
Import File
Press Import File to bring up a File Import dialog from which text and binary (e.g.
worksheet) files can be imported into the worksheet (see Chapter 4).
Filter
Press Filter to bring up a Filter Dialog that allows the data to be constrained to a
particular range; data outside of the specified range are treated as temporary missing data, i.e. excluded from subsequent analyses. The filter check box turns filter25

Chapter 3 Working with Data

ing on and off.
Print
Press Print to print the worksheet.
Copy
Press Copy to place the contents of the worksheet onto the Windows clipboard.
From the clipboard values can be pasted into other applications.
Rebuild
The Rebuild command builds the data arrays on which all statistical analyses are
based. This command is enabled whenever data records have been edited. The
data arrays must be rebuilt prior to semivariance or other analyses whenever data
have been changed. Sometimes this occurs automatically, such as when columns
are reassigned. Other times, such as when individual cells are edited, you must rebuild the data arrays yourself. When rebuilding is needed, the Rebuild command
will become enabled and the color of the font will change to red. During rebuilding
the data are checked for duplicate coordinate locations and for a sufficient number
of valid records.
Note that rebuilding is only needed when column assignments change or when data
within a column assigned to a coordinate or z-variate changes. In the screen above,
changing a value within the Al column will not require arrays to be rebuilt because
no variates (X, Y, Z, Z2, or ID) are assigned to the Al column.
Clear
Press Clear to empty the data worksheet and reset all analysis windows. Has the
same effect as the File New menu command.
Exit
Exit closes the Data Worksheet window and brings up the Data Summary window.
If the Rebuild command is enabled the data arrays will first be rebuilt.
Data Title / Description
Any text information desired can be entered in this field. When importing text files,
the header records in the file the records that appear prior to the data records
are automatically read into this box. If specified in the text file import window, these
records can also contain variate names that appear as data column titles. The first
line of the Data Description becomes the default graph and worksheet title.
Data Records
The second row of the data worksheet specifies the Field or Variate Assignments,
i.e. which field or column contains the X-Coordinate Data, Y-Coordinate Data, ZVariate Data, or Sample ID Data. These assignments can be changed by clicking
within the top row, which will bring up a Field Assignment Dialog window.
The third row of the data worksheet specifies the user-supplied Field or Variate
Names for the various data fields or columns. To enter or edit names, click on the
26

Chapter 3 Working with Data

cell to be edited. When data is imported from a text or external worksheet file, variate names can be read from the header records.
The data area of the worksheet contain data for each variate. To enter or edit data
in any given cell, double-click on that cell. To enter a formula rather than a value,
begin the cell with an = sign.
The data worksheet may contain 2 billion records and up to 64 columns.
To sort the datasheet, click on the topmost row of the column. Sorts will alternate
between ascending and descending order. Or use the Sort command in the Dat
Menu.
To change the number of decimal places to show for any given column, use the
Decrease Decimals or Increase Decimals command in the Data Menu, or click on
the respecitive tool icon in the menu bar.
To change column, move the cursor to the top of the column and use the mouse to
stretch or contract the column margins.
To temporarily delete a cell from analyses, change it to a Temporary Missing
Value with a click of the right mouse button (its color turns red and the font becomes
italicized; see the value 0.32 in the diagram above). Another click restores it to the
worksheet (its color will return to black and font to normal).
To change a value to a Permanent Missing Value, delete its contents by highlighting it and pressing the Delete key (note the empty cells in the figure above).
To insert or delete a row or column in the worksheet use the Edit | Insert command
or Delete on the menu bar.

27

Chapter 3 Working with Data

Column Assignments
In the Data Worksheet window, you may specify which field or column to associate
with the Sample ID, X-Coordinate, Y-Coordinate, the Z variate, or the Z2 variate by
clicking on the second cell from the top in any column. This will bring up the Column
Assignment dialog:

Column name
The name of the column or variate name.
Assign column as
None make column unassigned.
Sample ID the specified column contains Sample Number or Sample ID information. This can be text or numeric data.
X Coordinate the specified column contains values for the X-Coordinate location. If you choose a column that is already assigned, the other variates
column will switch with the original X Coordinate column.
Y Coordinate -- the specified column contains values for the Y-Coordinate location. If you choose a column that is already assigned, the other variates
column will switch with the original Y Coordinate column.
Z (Primary Variate) the specified column contains values for the Z variate.
If you choose a column that is already assigned, the other variates column
will switch with the original Z column.
Z2 (Covariate) the specified column contains values for the Covariate.
The Covariate is used in cokriging. If you choose a column that is already
assigned to another variate, the other variates column will switch with the
original Z2 column.
Rebuild automatically when reassign
Rebuild the internal worksheet when window closes. This is equivalent to pressing
the Rebuild button on the Data Worksheet window.

28

Chapter 3 Working with Data

Covariate Values Warning

In cokriging there are usually more Z2 covariate values than primary (Z) values. If
your column (field) assignments for Z and Z2 result in data arrays with more Z values than Z2 values then you will get the warning notice below.
You may continue the analysis with more Z than Z2 values, but there is no advantage to doing so in cokriging Z values without corresponding Z2 values are treated
as missing. Likewise, there is no advantage to cokriging when you have just as
many Z2 covariate values as you have Z values.

Missing Values
Data in the Worksheet that are marked as Missing are ignored during analyses.
Permanent missing values appear as blank cells and temporary missing values appear in red. You can use the right mouse button to make cells temporarily missing
and vice versa.
When importing or exporting files special placeholders (values or symbols) can be
used to indicate missing values These placeholders are specified by the user in the
Preferences window, or during file imports by a value or symbol specified in the File
Import Properties window. In GS+, the default missing value indicator is the numeric
value -99.0 but this can be changed in the User Preferences window (Chapter 2).
Missing values appear in output files when a value cannot be interpolated because
the location appears in an exclusive polygon or because numerical limitations disallowed its computation (such as when a variogram model is inappropriately used during kriging).

29

Chapter 3 Working with Data

Data Filter Dialog

The data within a worksheet can be collectively filtered or bounded using the Filter
command on the Data Worksheet window. With this command all records are
scanned and if a record falls outside of the specified Range to Use, the variate outside of its range will be treated as missing values during the data build.

Range to Use
In these fields specify how to constrain the data in the Worksheet to a particular
range. In the example shown, only those records for which the X direction field is
within the range of 0.60 to 40.00 will be included in analyses even though there are
records across the entire range of 0.60 79.90 in the worksheet. Although Worksheet values are treated as missing during data builds, they are not marked as missing in the Worksheet.
Reset
Reset Range to Use to the actual data range for a given coordinate or variate. Data
ranges can be set individually or (with the Reset All command) all at once.
Cancel
Cancel filter dialog.
OK
Accept the ranges specified. Note that for the filter to take effect you must check the
filter box on the Worksheet window.

30

Chapter 3 Working with Data

Duplicate Values
When worksheet data are rebuilt into data arrays, GS+ checks each record against
every other record to check that no duplicate coordinates are present. If one or
more records are found to be duplicates, a warning screen will appear:

Use first value (make other values missing)

This option includes the first record encountered (record 1 in the case described) in
the analysis and marks other duplicates in the worksheet (records 8, 25, and 26) as
temporarily missing values. So long as these values remain marked as missing, on
subsequent worksheet rebuilds this Duplicate Query warning will not occur.
Use last value (make other values missing)
This option is the same as the first option above except that the last record encountered is used in the analysis (record 26 in the case described) rather than the first.
Use average value (make none missing)
This option directs GS+ to use the average of the duplicate records. It does not
mark any values in the worksheet missing, which means that the next time the worksheet is rebuilt this Duplicate Query warning will occur again.
Stop data build and examine worksheet
Specifying this option will cause the data build to halt and will display the first duplicate record of the data worksheet. Any other duplicates already found during the
data build will be reset to non-missing will be returned to their pre-build state.
Also take this action for other duplicates of this build
Checking this box causes other duplicates encountered to be treated the same as
this set of duplicates. For data sets with many duplicate values, this option saves
time by keeping this dialog window from displaying again during the present build.
Cancel
Cancel the data build. Any duplicates set to missing values during this build will be
reset to nonmissing.
OK
Accept the option specified and continue the data build. The data build restarts
every time a duplicate is encountered to provide a comprehensive scan.
31

Chapter 3 Working with Data

32

Chapter 4 Importing Data from External Files

Chapter 4
Importing Data from External Files
File Import Dialog
A number of different file types can be imported into GS+. These include spreadsheets (e.g. Excel formats), databases (e.g. Access, dBase, Paradox, and FoxPro
formats), HTML (web) pages, and text files. Files are imported via the Import command within the Data Worksheet Window, which brings up the dialog box below.
From this dialog you choose the file and the format (as illustrated in the pull-down
list box at the bottom of the menu).

Open
Press Open to open the identified file. The window that next appears will depend on
the type of file to be opened. A text file will bring up the Import Text File dialog
whereas spreadsheet, database, and HTML formats will bring up the Import Spreadsheet dialog.

33

Chapter 4 Importing Data from External Files

Importing Text Files

Choosing to import text files from the File Import Dialog brings up the window below. The contents of the file are displayed in the large preview area. When the OK
button is pressed the contents of the file are read into the GS+ Worksheet according
to rules specified by the chosen File Format. These rules define how many rows of
header data precede the data rows (2, in the example below), how missing values
are identified (with the number 99 in the example below), how column names are
separated (with commas, below), and how data fields (columns) are separated from
one another within data records (with spaces in this example).
A different dialog window is available for importing spreadsheet, database, and
HTML files (described later).

File Format
A variety of text file types can be imported into GS+, including formats defined by the
user. Each type has its own manner for separating fields within data records, for
handling missing values, for allowing header records, and for specifying names of
fields (column titles) within the file. Predefined input format types include the following; examples of files appear in the next section:
GS+ format , in which fields are separated by spaces (free format) or tabs,
missing values are indicated with the placeholder -99, the number of header
records is automatically detected, and column titles (variate names) appear
on the 2nd record separated by commas.
GeoEas format, in which fields are separated with commas, there are no
missing value indicators, the number of header records is specified on the
2nd record of the file, and column titles appear as individual records following
this second record.
34

Chapter 4 Importing Data from External Files

Surfer XYZ format, in which fields are separated by spaces (free format),
missing values are indicated by blank fields, and the first record in the file is
a header record in which column titles (variate names) appear as fields
separated by spaces. Note that the Surfer XYZ format also allows fields to
be separated by commas, which should be specified separately as described below. Note that this format is not the same as the Surfer Grid file
format that can be used for Krig output files or Map input files.
ArcView XYZ format, in which fields are separated with commas, missing
values are indicated by placeholder commas, there is a single header record, and within the header record variate names are separated by quotes
and commas.
Custom, in which any of these format specifications can be changed or customized as specified in File Import Properties.
Properties
The Properties Command brings up a File Import Properties dialog window (described below) within which you can specify how GS+ should adjust its input parameters.
Change Column Assignments
The Change command brings up a Field or Column Assignment window from which
you can assign variates (e.g. X coordinate) to columns or fields in the data file. Columns can be reassigned later from the Data Worksheet window.
OK
Press OK to import the specified files data into the GS+ Worksheet and close the
dialog window. If data already exists in the GS+ Worksheet, you will be asked what
to do with it.
Cancel
Press Cancel to exit the Import Text File dialog without importing data.

35

Chapter 4 Importing Data from External Files

Text Input File Formats: GS+ Format
The standard GS+ input file format is comprised of header records and data records:
Data records are space-delimited or tab-delimited XYZ type data. This
means that each data record contains at least 3 fields: an x-coordinate location, a y-coordinate location, and the value for at least one z variate measured at that x-y location (single-dimension transects will have only xcoordinate and z variate data values). Additional fields can hold a sample
ID value and multiple z-variates for a particular x-y location.
Header records precede the data records and contain whatever text information about the file that the user feels is useful. There can be any number
of header records; for this format GS+ determines the number of header records automatically, which means that data records start with the first allnumeric record. The last header record can contain column titles (variate
names), separated by commas.
Missing values are denoted by the number -99.
Any of these parameters (field delimiters, number of header records, missing value indicators, etc.) can be changed to a custom format from the Import Data File window.
The following listing is the first 11 records of a standard GS+ input file that
has fields for sample ID, x-coordinate, y-coordinate, and two z variates.
Note the variate names in the second record, and two missing values I nthe
last column:
File Demo2d.dat;1994
sample, m east, m north,
1
4.5
11.9
2
2.7
29.4
3
1.6
32.6
4
4.1
44.5
5
0.6
64
6
2.4
71.8
7
7.8
3.5
8
6.7
10.2
9
6.7
16.3

36

Pb,
0.42
0.6
0.6
0.43
0.51
0.34
0.37
0.61
0.46

Al
0.42
0.45
0.08
-99.00
0.14
0.32
0.12
-99.00
0.49

Chapter 4 Importing Data from External Files

Text Input File Formats: GeoEas Format
The standard GeoEas input file format is comprised of header records and data records :
Data records are comma-delimited XYZ type data. This means that each
data record contains at least 3 fields: an x-coordinate location, a ycoordinate location, and the value for at least one z variate measured at that
x-y location (single-dimension transects will have only x-coordinate and z
variate data values). Additional fields can hold a sample ID value and multiple z-variates for a particular x-y location.
Header records precede the data records and contain specific information
about the data records:
Record 1 contains text of the user's choice, usually a data set title or
file name.
Record 2 contains the number of fields (values) in each data record.
Record 3 contains the name of the first field.
Records 4+ contain the names of the second, third, etc. fields (one
name per record)
There are no missing value indicators in the standard GeoEas format
Any of these parameters (field delimiters, number of header records, missing value indicators, etc.) can be changed to a custom format from the Import Text File window.
The following listing is the first 11 records of a standard GeoEas input file
that has fields for sample ID, x-coordinate, y-coordinate, and one z variate.
Note the four variate names in records 3-6:
File for field 55
4
ID
X meters
y meters
Z (mm)
13,-14036,-3097,11
14,-13621,-1266,25
22,-12384,-911,79
23,-121276,978,91
24,-12674,190,14

37

Chapter 4 Importing Data from External Files

Text Input File Formats: ArcView XYZ Format
The ArcView input file format is comprised of header records and data records :
Data records are comma-delimited XYZ type data. This means that each
data record contains at least 3 fields: x-coordinate location, a y-coordinate
location, and the value for at least one z variate measured at that x-y location (single-dimension transects will have only x-coordinate and z variate
data values). Additional fields can hold other variates for that location, e.g.
sample ID, other measured z-variates. The ArcView XYZ input format is
not the same as the ArcInfo Ascii Raster File Format, which is a Kriging output format that can be read directly by Arcinfo or Arcview GIS software.
A single header record precedes the data records and contains field (variate) names for the data record fields. Names are within quotes and commadelimited.
Missing values are indicated by blank fields.
Any of these parameters (field delimiters, number of header records, missing value indicators, etc.) can be changed to a custom format from the File
Import Properties window.
The following listing is the first 9 records of a standard ArcView input file
that has fields for an x-coordinate, y-coordinate, and one z variate. Note the
three variate names in record 1 and the missing values in records 5 and 9:
X data,Y data,Z data
4.5,11.9,0.42
2.7,29.4,0.45
1.6,32.6,0.08
4.1,,0.65
0.6, 64,0.14
2.4, 71.8,0.32
7.8, 3.5,0.12
6.7,10.2,
6.7,16.3, 0.49

38

Chapter 4 Importing Data from External Files

Text Input File Formats: Surfer XYZ Format
The standard Surfer input file format is comprised of header records and data records :
Data records are space-delimited XYZ type data. This means that each
data record contains at least 3 fields: an x-coordinate location, a ycoordinate location, and the value for at least one z variate measured at that
x-y location (single-dimension transects will have only x-coordinate and z
variate data values). Additional fields can hold other variates for that location, e.g. sample ID, other measured z-variates.
A single header record precedes the data records and contains field (variate) names or column titles for the data record fields. Names are spacedelimited so they must be single words (e.g. "mEast mNorth Nitrate") in order that they be properly assigned to their columns. You can allow names
to be delimited by commas or other characters by changing this to a Custom
Format in the File Import Properties window (e.g. "meters East, meters
North, Nitrate (ug/L)" ).
Missing values are indicated by blank fields. For files where there is more
than one z-variate per record, a missing value for any field in the record
means that the entire record will be treated as missing. (To avoid this problem use comma-delimited data records.)
Any of these parameters (field delimiters, number of header records, missing value indicators, etc.) can be changed to a custom format from the Import Text File window. Note that some Surfer files are comma-delimited
rather than space-delimited.
The following listing is the first 9 records of a standard Surfer input file that
has fields for an x-coordinate, a y-coordinate, and one z variate. Note the
three variate names in record 1 and the missing values in records 5 and 9:
Xdata
4.5
2.7
1.6
4.1
0.6
2.4
7.8
6.7
6.7

Ydata
11.9
29.4
32.6
44.5
64
71.8
3.5
10.2
16.3

Zdata
0.42
0.45
0.08
0.14
0.32
0.12
0.49

39

Chapter 4 Importing Data from External Files

Spreadsheet and Database Input Files

Choosing to import a spreadsheet, database, or HTML files from the File Import Dialog brings up a window similar to the one for Excel spreadsheets below. Multiple
worksheets or tables within the larger spreadsheet workbook or database are available through a pull-down listbox. The contents of the spreadsheet or data table are
displayed in the large preview area. When the OK button is pressed the contents of
the file are read into the GS+ Worksheet according to rules specified by the Properties command. A different dialog is used to import text files as described earlier.

Worksheets Available
The worksheets or tables within individual spreadsheet or database files are listed
here and can be chosen via the pull-down listing.
Properties
The Properties Command brings up a File Import Properties dialog window within
which you can specify how GS+ should identify missing values. When importing
spreadsheets and databases, other file import properties are limited to single predefined choices e.g. spreadsheet columns are used to denote different GS+ Worksheet fields.
Change Column Assignments
The Change command brings up a Field or Column Assignment window from which
you can assign variates (e.g. X coordinate) to columns or fields in the data file. Columns can be reassigned later from the Data Worksheet window.
OK/Cancel
Press OK to import the specified files data into the GS+ Worksheet and close the
dialog window. If data already exists in the GS+ Worksheet, you will be asked what
to do with it. Press Cancel to exit without importing data.
40

Chapter 4 Importing Data from External Files

Input File Formats (File Import Properties)

Choosing to import files from the Import Text File or Import Spreadsheet dialogs
brings up the File Import Properties window below.

Data Records
The Field (Column) Delimiter specifies how individual values within the data records
are formatted. For text files values can be comma separated, tab separated, space
separated (free format), character separated, or binary data. When spreadsheets
are imported, columns are delimited by Table Cells only.
The Missing Value Indicator specifies the value or character within the file that indicates that a value is missing and that the record should be ignored during importing. A missing value that is imported becomes a blank cell in the GS+ Data Worksheet. The indicator can be absent, a decimal point, a numeric value, or a character.
Header Records
The Number of Header Records indicates whether the first records in the file contain header records descriptive text that should be ignored as the file is read into
the Data Worksheet. For text files choose None, Varies, or Fixed Number. Header
records will be read into the Data Description field of the Data Worksheet Window.
Spreadsheet files are assumed to contain a single header row containing column
names.
The Column Title Separator refers to whether column titles (variate names) appear
in the second record of the file, and if so, how names are separated from one another within the record. For text files choose No Field Names, Same as for Data Records, Quotes, Brackets, Comma, Tab, Space, or Character. When spreadsheets
are imported, columns titles appear in different cells of the first spreadsheet r ow
OK/Cancel
Press OK to accept the settings shown and return to the Import Text File or Import
Spreadsheet dialog. If the settings are different from the standard settings for the
format originally defined, the format will change to Custom with the new settings defining the new Custom format. Press Cancel to close without importing.
41

Chapter 4 Importing Data from External Files

Viewing Files (File View Window)

The file view window is used to display the contents of selected files. It is available
from windows in which input and output files are specified. At the top of the window
is the file name and size. The contents of the window can be highlighted and copied
to the Windows clipboard using the Edit Copy menu command but cannot be
changed. The example below was brought up from the View Output File command
in the Interpolation window.

Data Append Dialog

When importing data into the Data Worksheet, if the worksheet already contains
data then you will asked how to import the new data whether the imported data
should Replace the existing data, be placed in columns to the right of the existing
data (Append to Side), or be placed at the bottom of the worksheet (Append to
End). Cancel will cancel the import with no changes to the worksheet.

42

Chapter 5 Summary Statistics

Chapter 5
Summary Statistics
Data Summary Window - Z Tab
The Data Summary window provides standard descriptive statistics for the variates
defined in the Data Worksheet window. Information is provided for both the Zvariate (as below), for the coordinate variates in a separate X,Y Coordinates tab,
and when a covariate is being analyzed, for the regression of Z vs. the Covariate Z2
in a separate Regression tab.
For the Z-variate it is also possible to specify a lognormal (base e) or square-root
transformation in order to better normalize the variates distribution prior to geostatistical analysis. If you do transform the variate, you may choose to have GS+ report
the interpolated (Kriged) values either in transformed form or backtransformed to the
original measurement domain. The backtransformation occurs after all analyses
have been performed, and it is not applied to autocorrelation results.
Also from the Data Summary window you can access a full-window frequency distribution, a cumulative probability distribution, or a normal probability distribution by
clicking on the small frequency distribution image.

Transformation
It is often helpful to apply a lognormal or a square root transformation to a Z variate
in order to normalize for skewed frequency distributions. It can also be useful to
scale data to a range of 0-1 if the values are extremely large. The transformation
specified is applied to every Z value in the data set prior to geostatistical analysis;
43

Chapter 5 Summary Statistics

the values in the data worksheet are not transformed. View the effectiveness of the
transformation by viewing the Frequency Distribution or a probability distribution and
the values for skewness and kurtosis in the data summary.
Offset
If your z-variates span the range of <1 to >1 (e.g. 0.3 to 20.1) and you decide to apply a lognormal or square root transform, you should make all values >1 prior to
transformation by adding an offset value such as 1 (e.g. ln[z+1]). This is because of
the discontinuous nature of the lognormal transformation across the <1 to >1 range.
Backtransformation
When a transformation is chosen, after analysis of the transformed data the output
data are customarily (but not necessarily) back-transformed to the original data domain when reported. You may choose among three potential backtransformation
choices: None, Standard, or Weighted. The standard backtransformation is simply
the converse of the transformation: scaled values are rescaled to the original range,
logn-transformed values are raised to the natural exponent e, and squared values
are raised to the 0.5 (square root). Offset values are subtracted from the backtransformed values.
The Weighted backtransformation is a complex backtransformation that more
closely approximates true population statistics than simple backtransformations.
See Haan (1977) and Krige (1981) for further details.
Backtransformations are applied only to final data. These include statistics on the
Data Summary screen (mean, standard deviation, etc.), and all kriging results. Individual semivariance values are not backtransformed prior to display.
Frequency Distribution
Click on a frequency distribution image to view an enlarged version of the frequency
distribution. From the Frequency Distribution window you will also be able to view a
normal or cumulative probability distribution of the data (see below).
Summary Statistics
Descriptive statistics appear in the box noted. Note that n refers to the number of
active data items currently in the analysis (128 in the screen above); n missing refers to the number of worksheet records that were excluded from the analysis because they contained a missing x, y, or z value or because they were excluded from
the analysis by the Filter command (4 values, in the example above). If duplicate
values were averaged (see the Duplicate Values dialog), then all the duplicates for a
given location count as a single record, and the duplicates will not be counted as
missing.

44

Chapter 5 Summary Statistics

Frequency Distribution Histogram

The Frequency Distribution window contains a bar graph of the frequency distribution for the Z variate. If the data are transformed, two graphs will appear, with the
distribution for the transformed data to the right of the distribution for the nontransformed data (as below). If the data are not transformed, only the left-hand graph will
appear. Data can be transformed in the in the Data Summary window.
The number of frequency classes (bars) can be changed using the Edit Graph
command. In addition to a frequency distribution bar graph, you may also choose to
view a Cumulative Frequency Distribution or a Normal Probability curve (described
below).

List Values
Press List Values to bring up a window containing the data used for the frequency
distribution.
Edit Graph
Press Edit Graph to bring up the Graph Settings Dialog Window, which will allow
you to make changes to the graph including changes in the number of distribution
classes (bars).
Print Graph
Press Print Graph to bring up the Graph Print dialog window.
Plot
Type of plot to graph. Choose either Frequency (as displayed above), Cumulative
Frequency, or Normal Probability curves (shown later).

45

Chapter 5 Summary Statistics

Frequency Distribution Values

The Frequency Listing window provides a listing of the values used to create the
frequency distribution graph. The number of classes is set from Frequency Distribution window using the Edit Graph command. This is a read-only worksheet.

Actions
You may Print or Copy the contents of the worksheet using the menu commands of
the main GS+ window. You may also change the Decimal Places reported by highlighting a column and pressing the Increase or Decrease Decimals icon (or use the
Data | Change Decimals menu command). Clicking on the top of a column will Sort
the worksheet based on the column selected in alternating ascending or descending order. You may also change Column Widths by placing the cursor over the line
between two columns and dragging to a new location.

46

Chapter 5 Summary Statistics

Cumulative Frequency Distribution

The Cumulative Frequency Distribution window contains a graph of the cumulative
frequency distribution for the Z variate. If the data are transformed, two graphs will
appear, with the distribution for the transformed data to the right of the distribution
for the nontransformed data (as below). If the data are not transformed, only the
left-hand graph will appear. Transformations are performed in the Data Summary
window.
If the data are normally distributed, the curve will describe an S-shape frequency
distribution. E.G. in the graph below, transforming the data acts to normalize the
distribution.

List Values
Press List Values to bring up a window containing a listing of cumulative frequency
distribution values.
Edit Graph
Press Edit Graph to bring up the Graph Settings Dialog Window, which will allow
you to make changes to the graph.
Print Graph
Press Print Graph to bring up the Graph Print Dialog Window.
Plot
Type of plot to graph. Choose either Frequency, Cumulative Frequency (as displayed above), or Normal Probability curves (displayed below).

47

Chapter 5 Summary Statistics

Cumulative Frequency Values

The Cumulative Frequency Listing window provides a listing of the values used to
create the Cumulative Frequency Distribution graph. All values in the data set are
ranked by value (ascending order); any given row is the frequency of the value of
that row plus all preceding frequencies (expressed as percent); the final cumulative
frequency value is 100%. For example, the first value in the table is the lowest value
in the data set and has a cumulative frequency of 100 (1/ n).

Actions
You may Print or Copy the contents of the worksheet using the menu commands of
the main GS+ window. You may also change the Decimal Places reported by highlighting a column and pressing the Increase or Decrease Decimals icon (or use the
Data | Change Decimals menu command). Clicking on the top of a column will Sort
the worksheet based on the column selected in alternating ascending or descending order. You may also change Column Widths by placing the cursor over the line
between two columns and dragging to a new location.

48

Chapter 5 Summary Statistics

Normal Probability Distribution

The Normal Probability Distribution window contains a normal probability curve for
the Z variate. If the data are transformed, two graphs will appear, with the distribution for the transformed data to the right of the distribution for the nontransformed
data (as below). If the data are not transformed, only the left-hand graph will appear. Transformations are performed in the Data Summary window.
The normal probability curve is the same as a Cumulative Frequency distribution
with the y-axis logn transformed. If the data are normally distributed, the curve will
describe a straight line. E.G. in the graph below, transforming the data acts to normalize the distribution.

List Values
Press List Values to bring up a window containing a listing of normal probability distribution values.
Edit Graph
Press Edit Graph to bring up the Graph Settings Dialog Window, which will allow
you to make changes to the graph.
Print Graph
Press Print Graph to bring up the Graph Print Dialog Window.
Plot
Type of plot to graph. Choose either Frequency, Cumulative Frequency, or Normal
Probability (as displayed above) curves.

49

Chapter 5 Summary Statistics

Normal Probability Values

The Normal Probability Listing window provides a listing of the values used to create
the Normal Probability distribution. All values in the data set are ranked by value
(ascending order); any given row is the frequency of the value of that row plus all
preceding frequencies (expressed as percent lognormally transformed).

Actions
You may Print or Copy the contents of the worksheet using the menu commands of
the main GS+ window. You may also change the Decimal Places reported by highlighting a column and pressing the Increase or Decrease Decimals icon (or use the
Data | Change Decimals menu command). Clicking on the top of a column will Sort
the worksheet based on the column selected in alternating ascending or descending order. You may also change Column Widths by placing the cursor over the line
between two columns and dragging to a new location.

50

Chapter 5 Summary Statistics

Data Summary Window - X,Y Coordinates Tab

The Data Summary provides simple descriptive statistics for the variates defined in
the Data Worksheet window. Information is provided for both the X,Y Coordinates
(as below) , for the Z variate in a separate Z-Variate tab (described earlier), and in a
Regression tab when a covariate is assigned (described later).

Coordinates Range and Name

This is the range over which the x-direction and y-direction data vary, and the name
of the variates as defined in the Data Worksheet.
Posting
The posting is a map of the location of each x,y coordinate point within the range of
X and Y coordinate values. For 1-dimensional transects the posting appears as
points along a straight line. Press Expand or click on the map image to bring up a
larger Covariate Postings (Quantile Plots) window that provides access to legends
and different ways of grouping quanitle data (see below). Data postings of kriged
data are also available from the Map window.

51

Chapter 5 Summary Statistics

Coordinate Postings (Quantile Plots)

2d Scattergrams
The Coordinate Postingwindow provides a map of sample locations that can be
coded to show sample values at that location. Values are grouped into as many as
10 symbol classes, with classes defined using one of three different strategies.

Data may be grouped in one of three distributions:

Quantiles in which Z data are placed into groups of equal frequency, i.e.
into groups that have the same number of sample values. Quartile distributions are divided into 4 groups, percentiles into 100 groups, etc.
Intervals in which Z data are placed into groups based on even intervals of
the data range, without regard to how many sample values are placed in
each group.
Custom in which you can define how groups are formed.
Note that this coordinate posting is different from the posting available from the Map
Window. The Map Window posting uses data saved with the interpolation file; this
posting uses nonmissing data in the current Data Worksheet.
Levels
Set the Number of levels (corresponding to different symbols) to any value be52

Chapter 5 Summary Statistics

tween 1 and 10. For values of 1, the legend will be suppressed.
Set the Type of level to correspond to how symbols are assigned to locations.
Quantiles divide the domain into an equal number of sorted values per level. In the
example above (quartiles), there are as many values greater than the median of
0.34 as there are less than 0.34.
Choose Intervals to divide the range into even intervals, e.g. in the example above,
in which values range from 0 to 1.25, choosing Intervals would divide the data range
of 0 1.25 into 4 levels: <0.3125, <0.625,< 0.9375, and <1.25.
Define provides the opportunity to define custom intervals and set symbol size,
color, and shape, as described later.
Edit Graph
Press Edit Graph to bring up the Graph Settings Dialog Window, which will allow
you to make changes to the graph.
Print Graph
Press Print Graph to bring up the Graph Print Dialog Window.

53

Chapter 5 Summary Statistics

1d Scattergrams
The 1d Coordinate Posting window provides a map of coordinate locations for 1dimensional data. The location of each data point in the active data set is marked
by a symbol corresponding to its relative value. Both the number of levels displayed
as well as the range for each level is user-defined. A similar window provides postings for 2D data, as described above.

Commands in this window are the same as for 2d Scattergrams, above.

54

Chapter 5 Summary Statistics

Defining Posting Intervals

The Define Posting Intervals window allows posting intervals to be custom-defined,
and provides the opportunity to set symbol size, color, shapes, and legend names.
This window is accessed from the Coordinate Posting window.

Value Range
Specify the upper threshold of each value range. Changing a Percentile or Intervals
value will change the Type of Level (defined in the Coordinate Posting window) to
Custom. The number of values displayed is based on the number of Levels chosen
in the Coordinate Posting window.
Symbol
Choose a symbol for the specified interval.
Name
Choose a name to appear in the chart legend.
Color
Choose a color for the specified symbol by clicking on the colored box.
Size
Choose a size for the specified symbol
Legend title
Specify the title to appear at the top of the legends. No legend will appear if only 1
interval level is chosen.

55

Chapter 5 Summary Statistics

Data Summary Window - Regression Analysis Tab

The Data Summary Window Regression Tab is available when a covariate is assigned in the Data Worksheet Window via the Field Assignment Dialog. Information
for the Z variate or Covariate is provided in a separate Z-Variate tab, and for the coordinate variates in a separate X,Y Coordinates tab, as described earlier.

Enlarge
Press Enlarge to view a larger version of the Cross-Variate Regression scatterplot.
From the Regression window you will be able to edit and print the graph as well as
list graphed data values, as described below.

56

Chapter 5 Summary Statistics

Cross-Variate Regression
The best test of whether a covariate is related to the primary variate a prerequisite
for cokriging is to perform a regression analysis. In GS+ the results of the regression of the primary variate Z vs. the Covariate Z2 is presented in the Data Summary
Regression tab. Clicking the Enlarge command in that window brings up the CrossVariate Regression window:

List Values
Press List Values to bring up a window containing a listing of the values used in the
regression, as described below.
Edit Graph
Press Edit Graph to bring up the Graph Settings Dialog Window, which will allow
you to make changes to the graph.
Print Graph
Press Print Graph to bring up the Graph Print Dialog Window.

57

Chapter 5 Summary Statistics

Cross-Variate Regression Values

The Regression Listing window provides a list of the values used to create the regression. Regression statistics appear in the Data Summary Regression tab.

Actions
You may Print or Copy the contents of the worksheet using the menu commands of
the main GS+ window. You may also change the Decimal Places reported by highlighting a column and pressing the Increase or Decrease Decimals icon (or use the
Data | Change Decimals menu command). Clicking on the top of a column will Sort
the worksheet based on the column selected in alternating ascending or descending order. You may also change Column Widths by placing the cursor over the line
between two columns and dragging to a new location.

58

Chapter 6 Semivariance Analysis

Chapter 6
Semivariance Analysis - Overview
Semivariance is an autocorrelation statistic defined as

(h) = [1 / 2N(h)] [ zi zi+h ] 2

where

(h) = Semivariance for interval distance class h;

zi = measured sample value at point i;
zi+h = measured sample value at point i+h; and
N(h) = total number of sample couples for the lag interval h.
Semivariance is evaluated in GS+ by calculating (h) for all possible pairs of points in
the data set and assigning each pair to an interval class h. For uniform interval
classes , GS+ makes interval class assignments for any given pair of points using
the following formula:
class = INT(D/DI) + 1
where
D = distance separating the pair
DI = lag class distance interval
INT = Integer function
Where the first lag is offset by interval in order to better resolve values close to
the origin, the formula is:

class = INT([D/DI]+0.5) + 1

This option is available by checking a box in the Preferences Analyses dialog. It

can provide better resolved variograms when there are sufficient pairs of points at
shorter separation distances. The disadvantage is that if there are few pairs of
points for the shortest distance classes when the first lag is not offset (a common
problem), there will be even fewer pairs available with the first lag offset. If this is
the case, there will be little if any improvement to the variogram.
For individually-specified lag class intervals, pairs of points are assigned to interval
59

Chapter 6 Semivariance Analysis

lag classes based on values in the Define Lag Class Intervals window.
GS+ calculates a semivariance statistic for each interval class; the graph of all hs vs.
all semivariances for each interval class in the analysis constitutes the variogram
(more properly called the semivariogram).
For cokriging, semivariance analysis must be performed for the primary variate (Z),
for the covariate (Z2), and for the cross variate (Z x Z2) situation.

Autocorrelation Window
The Autocorrelation Window is where various options for variogram calculations and
the end results appear; see the Semivariance Analysis summary for a definition of
semivariance and formulas for lag class distance intervals.

Active Lag Distance

The Active Lag Distance specifies the range over which semivariance will be calculated. The minimum distance for this field is the minimum distance between adjacent
points in the data set, while the maximum distance is the maximum distance between points. Specifying a value too large or too small will assign respectively the
largest or smallest possible values to this distance.
For example, a 1200 m transect will have a maximum lag of 1200 m; specifying an
Active Lag of 300 m will limit the variogram to lag intervals less than or equal to 300
60

Chapter 6 Semivariance Analysis

m along the entire 1200 m length of the transect.
The default active lag is 80% of the maximum lag. This is not likely to be the most
appropriate active lag for your data but rather will provide a starting point.
Variograms typically decompose at large lag intervals because of decreasing numbers of couples per lag class as the maximum lag interval is approached.
GS+ allows 1 million lag classes to be specified with up to 1 billion pairs per class.
Lag Class Distance Interval
The Lag Class Distance Interval defines how pairs of points will be grouped into lag
classes. Each point in a variogram represents the average semivariance for a single
lag class, which is a group of pairs separated by a certain Lag Class Distance Interval, sometimes called a step size. This interval can either be calculated by GS+, in
which case it will be uniformly distributed across the active lag distance, or it can be
manually set by the user:

Use individually-specified points

With this option you may use the Define command to bring up a window to
Define Lag Class Intervals, i.e. to specify individual break points for the lag
intervals (described below).

Use a uniform interval

With this option, the value specified is the size of the interval, applied uniformly across the active lag distance. E.G. an interval of 2 units with an active lag distance of 10 units will create 5 lag classes, each 2 units wide. The
minimum interval allowed is the smallest distance separating any two sample point locations in the data set. The maximum interval is the greatest distance separating any two sample point locations. The default value is 10%
of the active lag or, if 10% of the active lag is smaller than the minimum allowed, the minimum allowed. This default may not be appropriate for any
given data set; you should try different steps for every set.
The number of lag classes (and therefore plotted points) in a variogram is a
function of values for the active lag and the active step; a 300 m active lag
with a 15 m active step will have ca. 20 lag classes. Note, however, that the
lag distance for a given class will be the average distance separating points
within the class and not necessarily the midpoint for the class. For a 10-20
m lag class, e.g., the average lag distance may be 12.3 m rather than 15 m
if more pairs of points are separated by 10-15 m intervals than by 15-20 m
intervals.
Changing the distance interval will clear results on the screen from previous
analyses calculated with a different step. Results based on the new step
61

Chapter 6 Semivariance Analysis

must be re-generated with Calculate command.
Anisotropic Axis Orientation
Anisotropy refers to a direction-dependent trend in the data. Consider data collected
from a two-dimensional grid on a mountain slope: elevation will be autocorrelated
differently in the upslope-downslope direction than in a cross-slope direction, and
thus an isotropic (all-direction) analysis may hide much of the autocorrelation that in
fact is present. Anisotropic analysis allows you to see if your data have a directional
component that might arise from a variety of unforeseen factors. Anisotropic analysis is irrelevant for single-dimension data such as a transect.
GS+ evaluates geometric anisotropy, i.e. anisotropy which is expressed as variograms with different ranges in different directions. The Principal Anisotropic Axis
(the Major Axis of the anisotropic model) is the direction with the longest range, i.e.
the direction of major spatial continuity.
The best way to evaluate anisotropy is to view the Anisotropic Semivariance Surface
(Variogram Map), and use the Azimuth function to define and then set the Principal
Anisotropic Axis to the direction aligned with the lowest semivariance values (the
direction of maximum spatial continuity, or major axis of the anisotropic variogram
model). The map is accessed by pressing the Surface command at the bottom of
the Semivariance Analysis window.

Principal Axis (degrees N)

The Principal Axis is the direction of maximum spatial continuity, or base axis from
which the offset angles for anisotropic analyses are calculated. Offset angles are 0,
45, 90, and 135 clockwise from the base axis; points aligned sufficiently close to
one or another of these angles (see Offset Tolerance below) are included in the anisotropic analysis for that angle.
The axis orientation should correspond to the axis of maximum spatial continuity, i.e.
the major anisotropic axis. The default axis is 0 from the north-south (y) axis.
Choose the appropriate axis value by using the Surface command to bring up the
Anisotropic Variogram Surface window.

Offset Tolerance (degrees)

In anisotropic analyses, the Offset Tolerance determines how closely the alignment
between any two points needs to be for those points to be included in the analysis
for a given offset angle. Two points will be included in the analysis for a given offset
angle if the angle between them is within the offset tolerance from the offset angle.
62

Chapter 6 Semivariance Analysis

For example, if the angle between two points is 59.3 and the offset tolerance is
15.0, the points will be included only in the 45 angle class, which would include all
angles between 30 and 60. The default tolerance is 22.5.

Variogram Options
Show Sample Variance
Check this option to show the sample variance for the data as a dashed line on the
variogram graphs.
Show Variogram Model
Check this option to show a model for the variogram points. If the model has already been defined, either automatically or manually, the variograms will be redrawn
with the model now graphed. If a model has not yet been defined, or upon executing the Calculate command, a best-fit model will be calculated and graphed.
To see the model parameters and to change the model, use the Model command at
the bottom of the variogram image.
Expand
The Expand command brings up a separate variogram window, from which the
variogram can be printed or formatted. Variance Cloud Analysis, the ability to view
individual semivariance values, and the number of pairs per variogram class interval
are also available from these windows.
Model
The Model command brings up a Model Dialog window within which you may
change the variogram model. The Model command is enabled only when the Show
Model Variogram Option is selected.
Surface
The Surface command brings up the Anisotropic Variogram Surface Map window.
The surface map is useful for visualizing anisotropic autocorrelation when present,
as described below.
Calculate
The Calculate command causes the semivariogram to be calculated.

63

Chapter 6 Semivariance Analysis

Define Nonuniform Lag Class Intervals

Use this dialog window to specify individual lag classes that are not necessarily uniform. In the cells of the spreadsheet you can specify the upper bound of the distance interval classes desired. Zero is always the lowest bound and the Active Lag
Distance is always the upper bound, regardless of values entered in this dialog. In
the example below, note that classes are separated first by 2 distance units, then by
4, 18, and 70 units.

Clear
Clear the worksheet.
Import
Import a text file containing the lag class interval bounds. The default extension for
interval files is .int. The format of the file to be imported is numeric-only records following a variable number of alphanumeric header records. For example:
line 1:
line 2:
line 3:
line 4:
line 5:
line 6:
line 7:
line 8:
64

Optional header record 1

Optional header record 2
2.0
4.0
8.0
12.0
30.0
100.0

Chapter 6 Semivariance Analysis

This file describes 8 lag classes:
0 to <2.0,
2.0 to <4.0,
4.0 to <8.0,
8.0 to <12.0,
12.0 to <30.0,
30.0 to <100.0,
100.0 to maximum lag distance.
You can adjust the active lag to be any value up to and including the maximum distance separating points in the input file, i.e. the maximum lag distance.
Actions
You may Print or Copy the contents of the worksheet using the menu commands of
the main GS+ window. You may also change the Decimal Places reported by highlighting a column and pressing the Increase or Decrease Decimals icon (or use the
Data | Change Decimals menu command). Clicking on the top of a column will Sort
the worksheet based on the column selected in alternating ascending or descending order. You may also change Column Widths by placing the cursor over the line
between two columns and dragging to a new location.

65

Chapter 6 Semivariance Analysis

Variograms Window
The Variograms window presents a full-window variogram that can be edited and
printed. Different tabs hold variograms for the different anisotropic directions. Additionally, the semivariance values that were used to produce the variogram can be
listed, and Variance Cloud Analysis can be performed. Variance Clouds provide a
means for detecting outlier pairs of points that may be artificially skewing the variogram.
You can use the mouse cursor to identify the number of pairs in specific lag classes
(reported at the bottom of the window; in the example below, the cursor (not shown)
was over the 2nd symbol in the variogram), and to begin variance cloud analysis.
Note also that parameters for the variogram model (if present) are presented in a
graph footnote. If you do not want parameters to appear as a footnote you must use
the Edit Graph command to bring up the Graph Settings dialog and in the Footnote
field remove the notation (model results).

List Values
Bring up a Semivariance Values Window, including for each lag class the average
separation distance for pairs of points in that class, the average semivariance for
those points, and the number of pairs of points upon which the average distance and
semivariance are based.
66

Chapter 6 Semivariance Analysis

Graph Cloud
Create a Variance Cloud window.
Edit Graph
Bring up a Graph Settings dialog window for editing the graph.
Print Graph
Print the graph via a Graph Print dialog window.

Semivariance Values
In this worksheet are listed for each lag class the average separation distance for
pairs of points in that class, the average semivariance for those points, and the
number of pairs of points upon which the average distance and semivariance are
based. This is a read-only worksheet.

Actions
You may Print or Copy the contents of the worksheet using the menu commands of
the main GS+ window. You may also change the Decimal Places reported by highlighting a column and pressing the Increase or Decrease Decimals icon (or use the
Data | Change Decimals menu command). Clicking on the top of a column will Sort
the worksheet based on the column selected in alternating ascending or descending order. You may also change Column Widths by placing the cursor over the line
between two columns and dragging to a new location.

67

Chapter 6 Semivariance Analysis

Isotropic Variogram Models

GS+ provides five types of isotropic models, each of which can be described based
on three parameters:
Nugget Variance or C0 the y-intercept of the model
Sill or C0+C the model asymptote
Range or A the separation distance over which spatial dependence is apparent. Sometimes this is called the effective range in order to distinguish
range (A) from a models range parameter (A0). In GS+ the Range (A) is calculated from A0 as described in the formulas for the different models, later in
this chapter.

Generalized Variogram Model

0.24

Sill (C + C0) = 0.21

Semivariance

0.19

0.12
Nugget (C0) = 0.04
0.06
Range (A) = 13.3

C0

0.00
0

20

40

Separation Distance (h)

68

60

80

Chapter 6 Semivariance Analysis

+

GS calculates default values for each parameter of the five models. You may
change any of these three model parameters from the Isotropic Variogram Model
dialog window:

Variogram Model Type

Choose one of four isotropic models (Linear, Spherical, Exponential, and Gaussian).
As a model is chosen the variogram graphs will be updated to denote the change.
Model Terms
Any of the three model parameters for each model may be changed within the
ranges allowed for individual parameters. To change a value either move the slider
beneath it or type a new value directly into the text box:
Nugget Variance or C0 the y-intercept of the model; the nugget variance
can never be greater than the sill.
Sill or C0+C the model asymptote; the sill can never be less than the nugget variance.
Range (A) the separation distance over which spatial dependence is apparent. This is sometimes called Effective Range in order to distinguish
range (A) from a models range parameter (A0). The Range A is calculated
from A0 as described in the formulas for the different models, below.The
range cannot be less than 0.
Statistics This Fit
GS+ provides three statistics to aid the interpretation of model output:
Residual Sums of Squares provides an exact measure of how well the
model fits the variogram data; the lower the reduced sums of squares, the
better the model fits. When GS+ autofits the model, it uses RSS to choose
parameters for each of the variogram models by determining a combination
69

Chapter 6 Semivariance Analysis

of parameter values that minimizes RSS for any given model. The Residual
SS displayed in the This Fit box is calculated for the currently defined model.
r2 provides an indication of how well the model fits the variogram data; this
value is not as sensitive or robust as the Residual SS value for best-fit calculations; use RSS to judge the effect of changes in model parameters.
Proportion C/(C0+C) -- this statistic provides a measure of the proportion of
sample variance (C0+C) that is explained by spatially structured variance C.
This value will be 1.0 for a variogram with no nugget variance (where the
curve passes through the origin); conversely, it will be 0 where there is no
spatially dependent variation at the range specified, i.e. where there is a pure
nugget effect.
Statistics AutoFit
Statistics for the Autofit model (the model most recently calculated by GS+) apppear
in this box. In this way you can compare your changes to model terms against that
calculated automatically.
Autofit
Automatically refit the model using as starting conditions the parameters in this dialog window. Model fitting in GS+ is somewhat dependent on starting conditions (assumed model parameters prior to iterations). Sometimes you can fit a better model
by hand, in which case Refitting will further refine parameters to minimize RSS. To
return to the original model parameters you may need to exit this Model definition
window and recalculate the variogram.
Print
Print a copy of the model window.
Cancel
Exit the dialog window without applying changes.
OK
Close the dialog window and apply any changes made to individual models.

70

Chapter 6 Semivariance Analysis

Spherical Isotropic Model

The spherical isotropic model is a modified quadratic function for which at some distance A0 pairs of points will no longer be autocorrelated and the semivariogram
reaches an asymptote. The formula used for this model is:

(h) = C0 + C [1.5(h / A0) - 0.5(h / A0)3]

(h) = C0 + C

for h A0
for h > A0

where

(h) = semivariance for interval distance class h,

h = the lag distance interval,
C0 = nugget variance 0,
C = structural variance C0, and
A0 = range parameter. In the case of the spherical model,
the effective range A = A0.

71

Chapter 6 Semivariance Analysis

Exponential Isotropic Model

The exponential isotropic model is similar to the spherical in that it approaches the
sill gradually, but different from the spherical in the rate at which the sill is approached and in the fact that the model and the sill never actually converge. The
formula used for this model is:

(h) = C0 + C[1 - exp(-h / A0)]

where

(h) = semivariance for interval distance class h,

h = lag interval,
C0 = nugget variance 0,
C = structural variance C0, and
A0 = range parameter. In the case of the exponential model,
the effective range A = 3A0, which is the distance
at which the sill (C + C0) is within 5% of the
asymptote (the sill never meets the asymptote in
the exponential or gaussian models).

72

Chapter 6 Semivariance Analysis

Linear Isotropic Model

The linear isotropic model describes a straight line variogram. Note that there is no
sill in this model; the range A0 is defined arbitrarily as noted below. The formula
used is:

(h) = C0 + [ h(C / A0) ]

where

(h) = semivariance for interval distance class h,

h = lag interval,
C0 = nugget variance 0,
C = structural variance C0, and
A0 = range parameter. In the case of the linear model,
there is no effective range A spatial autocorrelation occurs across the entire range
sampled; in GS+ model windows A is set
initially to the separation distance (h) for the
last lag class graphed; in the variogram below,
this would be 77.4.

73

Chapter 6 Semivariance Analysis

Gaussian Isotropic Model

The gaussian or hyperbolic isotropic model is similar to the exponential model but
assumes a gradual rise for the y-intercept. The formula used for this model is:

(h) = C0 + C[1 - exp(-h2 / A02)]

where

(h) = semivariance for interval distance class h,

h = lag interval,
C0 = nugget variance 0,
C = structural variance C0, and
A0 = range parameter. In the case of the gaussian model,
the effective range A = 30.5 A0, which is the distance
at which the sill (C + C0) is within 5% of the
asymptote (the sill never meets the asymptote in
the guassian or exponential models).

74

Chapter 6 Semivariance Analysis

Anisotropic Variogram Models

GS+ provides four types of anisotropic or direction-dependent geometric models,
each of which can be described based on four parameters:

Nugget Variance or C0 the y-intercept of the model; this value is the same
for all directions.

Sill or C0+C the model asymptote; this value is the same for all directions.

Range or A the separation distance over which spatial dependence is apparent for the direction examined. It is the sum of:

A1 the range parameter for the major axis of variation and

A2 - the range parameter for the minor axis ( + 90),

adjusted for the angle between pairs as noted in the formulas for individual models.
Sometimes Range or A is called the effective range in order to distinguish range from a models range parameters (A1 or A2). In GS+ the Minor Ranges and Major Ranges are calculated from A1 and A2, respectively, as described in the formulas for the different models later in this
chapter.

GS+ calculates default values for each parameter of the five models. You may
change any of these four model parameters from the Anisotropic Variogram Model
dialog window:

Variogram Model Type

Choose one of four isotropic models (Linear, Spherical, Exponential, or Gaussian).
As a model is chosen the variogram graphs will be updated to denote the change.
75

Chapter 6 Semivariance Analysis

Model Terms
Any of the three parameters for each model may be changed within the ranges allowed for individual parameters. To change a value either move the slider beneath it
or type a new value directly into the text box:
Nugget Variance or C0 the y-intercept of the model; the nugget variance
can never be greater than the sill.
Sill or C0+C the model asymptote; the sill can never be less than the nugget variance.
Range Minor the separation distance over which spatial dependence is apparent. This is sometimes called Effective Range in order to distinguish
range from a models range parameter (A2). The minor range is calculated
from A2 as described in the formulas for the different models, below. The minor range can never be more than the major range.
Range Major the separation distance over which spatial dependence is apparent. In some texts this is called the Effective Range in order to distinguish
range from the models major range parameter (A1). The minor range is calculated from A1 as described in the formulas for the different models, below.
The major range can never be less than the minor range.
Statistics This Fit
GS+ provides three statistics to aid the interpretation of model output:
Residual Sums of Squares provides an exact measure of how well the
model fits the variogram data; the lower the reduced sums of squares, the
better the model fits. When GS+ autofits the model, it uses RSS to choose
parameters for each of the variogram models by determining the combination
of parameter values that minimizes RSS for any given model. The Residual
SS displayed in the This Fit box is calculated for the currently defined model.
r2 provides an indication of how well the model fits the variogram data; this
value is not as sensitive or robust as the Residual SS value for best-fit calculations; use RSS to judge the effect of changes in model parameters.
Proportion C/(Co+C) -- this statistic provides a measure of the proportion of
sample variance (C0+C) that is explained by spatially structured variance C.
This value will be 1.0 for a variogram with no nugget variance (where the
curve passes through the origin); conversely, it will be 0 where there is no
spatially dependent variation at the range specified, i.e. where there is a pure
nugget effect.
Statistics AutoFit
Statistics for the Autofit model (the model most recently calculated by GS+) apppear
in this box. In this way you can compare your changes to model terms against that
calculated automatically.
76

Chapter 6 Semivariance Analysis

Autofit
Automatically refit the model using as starting conditions the parameters in this dialog window. Model fitting in GS+ is somewhat dependent on starting conditions (assumed model parameters prior to iterations). Sometimes you can fit a better model
by hand, in which case Refitting will further refine parameters to minimize RSS. To
return to the original model parameters you may need to exit this Model definition
window and recalculate the variogram.
Print
Print a copy of the model window.
Cancel
Exit the dialog window without applying changes.
OK
Close the dialog window and apply any changes made to individual models.

Spherical Anisotropic Model

The spherical anisotropic model is a modified quadratic function in which at some
distance A1 along the major axis and A2 along the minor axis, pairs of points are no
longer autocorrelated and the variogram reaches an asymptote. The formula used
for this model is:

(h) = C0 + C[1.5(h / A) - 0.5(h / A)3 ]

for h A

(h) = C0 + C

for h > A

where

(h) = semivariance for interval distance class h,

h = lag interval,
C0 = nugget variance 0,
C = structural variance C0, and
A = {A12 [cos2 ( - )] + A22 [sin2 ( - )] }
A1 = range parameter for the major axis (). In the case
of the spherical model, the range (or
effective range) for the major axis = A1.
A2 = range parameter for the minor axis ( + 90). In the
case of the spherical model, the range (or
effective range) for the minor axis = A2.
= angle of maximum variation
= angle between pairs
77

Chapter 6 Semivariance Analysis

Exponential Anisotropic Model

The exponential anisotropic model is similar to the spherical in that it approaches the sill gradually, but different from the spherical in the rate at which
the sill is approached and in the fact that the model and the sill never actually
converge. The formula used for this model is:

(h) = C0 + C[1 - exp(-h / A)]

where

(h) = semivariance for interval distance class h,

h = lag interval,
C0 = nugget variance 0,
C = structural variance C0, and
A = {A12 [cos2 ( - )] + A22 [sin2 ( - )] }
A1 = range parameter for the major axis (). In the case
of the exponential model, the range (or
effective range) for the major axis = 3A1.
A2 = range parameter for the minor axis ( + 90). In the
case of the exponential model, the range (or
effective range) for the minor axis = 3A2.
= angle of maximum variation
= angle between pairs

Linear Anisotropic Model

The linear anisotropic model describes a straight-line variogram. Note that there is
no sill in this model; the ranges for the major and minor axes are defined in terms of
the distance interval for the last lag class in the variogram. The formula used is:

(h) = C0 + h(C / A)
where

(h) = semivariance for interval distance class h,

h = lag interval,
C0 = nugget variance 0,
C = structural variance C0, and
A = {A12 [cos2 ( - )] + A22 [sin2 ( - )] }
A1 = range parameter for the major axis (). In the
case of the linear model, there is no effective major
range spatial autocorrelation occurs across the
78

Chapter 6 Semivariance Analysis

entire range sampled; in GS+ model windows A1 is
set initially to the separation distance (h) for the
last lag class graphed.
A2 = range parameter for the minor axis ( + 90). In the
case of the linear model, there is no effective minor
range spatial autocorrelation occurs across the
entire range sampled; in GS+ model windows A2 is
set initially to the separation distance (h) for the
last lag class graphed.
= angle of maximum variation
= angle between pairs

Gaussian Anisotropic Model

The gaussian or hyperbola anisotropic model is similar to the exponential model
but assumes a gradual rise from the y-intercept. The formula used is:

(h) = C0 + C[1 - exp(-h2 / A2)]

where

(h) = semivariance for interval distance class h,

h = lag interval,
C0 = nugget variance 0,
C = structural variance C0, and
A = {A12 [cos2 ( - )] + A22 [sin2 ( - )] }
A1 = range parameter for the major axis (). In the case
of the gaussian model, the range (or
effective range) for the major axis = 30.5 A1.
A2 = range parameter for the minor axis ( + 90). In the
case of the gaussian model, the range (or
effective range) for the minor axis = 30.5 A2.
= angle of maximum variation
= angle between pairs

79

Chapter 6 Semivariance Analysis

Anisotropic Semivariance Surface (2D Variogram Map)

The Anisotropic Semivariance Surface or Variogram Map provides a visual picture
of semivariance in every compass direction. This allows one to more easily find the
appropriate principal axis for defining the anisotropic variogram model. A transect in
any single direction (e.g. 10 degrees north) is equivalent to the variogram in that direction: the surface (z-axis) is semivariance; the x and y axes are separation distances in E-W and N-S directions, respectively. The center of the map corresponds
to the origin of the variogram (h) = 0 for every direction. See Isaaks and Srivastava
(1989; page 150) and Goovaerts (1997; page 98).

Edit Graph
Change axis scales and other graph settings via the Graph Settings dialog window.
Print
Print the graph to a file, printer, or other device.
3D
Change the graph projection to 3D from 2D or to 2D from 3D.
80

Chapter 6 Semivariance Analysis

Azimuth
When the check box is marked, a transect on the map surface illustrates the azimuth
angle. The angle is specified in the text box (e.g. 330 in the window above), and
also displayed graphically in the area under the azimuth angle. The average
semivariance for the transect is displayed as A:nnn beneath the Set command.
Note that there is no difference between an angle and that angle plus 180.
Set
Make the azimuth angle displayed in this window the Principal Anisotropic Axis in
the Semivariance Analysis window. Pressing Set will reset the Principal Anisotropic
Axis; for consistency, the angle in the Semivariance Analysis window will display a
value between 0 and 180, so that a value >180 will appear as that value less 180
(e.g. 225 will be reset to 90 in the Semivariance Analysis window).
Set will force a Rebuild of the anisotropic variograms (and models, if selected) in the
semivariance window. The Principal Anisotropic Axis (the Major Axis of the anisotropic model) should be the direction of major spatial continuity (or major axis of the
anisotropic variogram model), corresponding to the direction with the longest range.
Correspondingly, A should be smaller in the major direction (lower average semivariance) and largest in the minor (90-offset) direction.
A
The average semivariance for the azimuth transect is displayed as A:nnn beneath
the Set command, where nnn is the average semivariance. A should be smaller in
the major direction (lower average semivariance) and largest in the minor (90offset) direction
Mouse Action
Off returns the mouse to normal operation.
Rotate turns the cursor into a rotator cuff when the left mouse button is
pushed, allowing the image to be rotated as desired. This option is available
only for 3D map projections.
Move allows the graph to be moved within the window by clicking the left
mouse button and dragging the cursor.
Scale shrinks the graph image with the left mouse button.
Zoom allows you to zoom in on a particular graph area by using the left
mouse button to define a rectangular zoom area. Within the zoomed area
the location of the cursor is noted on the Mouse Location panel.
Reset resets the image to the default rotation angle and scale.
Mouse Location
Provides information on the current location of the cursor when it is on the map surface. Units are map units as specified by the axes. Z is semivariance.
81

Chapter 6 Semivariance Analysis

Anisotropic Semivariance Surface (3D Variogram Map)

The 3D Anisotropic Semivariance Surface or Variogram Map is a 3D projection of
the 2D map. Use the Edit Graph command to change 3D graph formats.

Commands for the 3D map projection are the same as those for the 2D projection,
above.

82

Chapter 6 Semivariance Analysis

Cross-Autocorrelation or Cross-Semivariance Analysis

The Cross-Autocorrelation Analysis Window is identical to the Autocorrelation
Analysis window except that cross-semivariance analysis (Z x Z2) rather than
semivariance analysis (Z or Z2) is performed. Cross-semivariance analysis must be
performed prior to cokriging. For a description of the commands and options provided in this window please see the description of the Semivariance Analysis Window, above.

Commands in the Cross-Autocorrelation window are the same as in the Autocorrelation window, earlier in this chapter.

83

Chapter 6 Semivariance Analysis

84

Chapter 7 Other Autocorrelation Measures

Chapter 7
Other Autocorrelation Measures
Standardized Variograms
In standardized variogram analysis the variogram is computed as a proportion of
sample variance:

(h) = ( [1 / 2N(h)] [ zi zi+h ] 2 ) / s2

where

(h) = Semivariance for interval distance class h;

zi = measured sample value at point i;
zi+h = measured sample value at point i+h;
N(h) = total number of sample couples for the lag interval h; and
s2 = sample variance for all zs.
Standardized variograms (both isotropic and anisotropic) appear in the Autocorrelation Window, on the Standardized Variogram tab. Pressing the Expand button on
this tab brings up the Standardized Variograms window; note the sample variance
line at 1.0 along the y-axis:

The List, Edit, Print, Cloud, Scatter, and Exit buttons work here as they do for the
Variograms Window.
85

Chapter 7 Other Autocorrelation Measures

Madograms
Madograms are similar to traditional variograms (see Semivariance Analysis), except that the absolute difference between zi and zi+h is used rather than the square
of the difference as for traditional semivariance. The formula thus becomes

(h) = [1 / 2N(h)] | zi zi+h |

where

(h) = semivariance(M) for interval distance class h,

zi = measured sample value at point i,
zi+h = measured sample value at point i+h, and
N(h) = total number of sample couples for the lag interval h.

Madograms (both isotropic and anisotropic) appear in the Autocorrelation Window,

on the Madogram tab. Press the Expand button on this tab to brings up the Madograms window:

The List, Edit, Print, Cloud, Scatter, and Exit buttons work here as they do for the
Variograms Window.

86

Chapter 7 Other Autocorrelation Measures

Rodograms
Rodograms are similar to traditional variograms (see Semivariance Analysis), except
that the square root of the difference between zi and zi+his calculated rather than the
square of the difference. The formula thus becomes

R(h) = [1 / 2N(h)] [ |( zi zi+h)| ]0.5

where

R(h) = semivariance(R) for interval distance class h,

zi = measured sample value at point i,
zi+h = measured sample value at point i+h, and
N(h) = total number of sample couples for the lag interval h.

Rodograms (both isotropic and anisotropic) appear in the Autocorrelation Window,

on the Rodogram tab. Press the Expand button on this to tab bring up the Rodograms window:

The List, Edit, Print, Cloud, Scatter, and Exit buttons work here as they do for the
Variograms Window.

87

Chapter 7 Other Autocorrelation Measures

Drift
Drift (also called trend) is calculated as
F(h) = [1 / N(h)] [ (zi zi+h) ]
where
F(h) = drift for interval distance class h,
zi = measured sample value at point i,
zi+h = measured sample value at point i+h, and
N(h) = total number of sample couples for the lag interval h.

Drift graphs (both isotropic and anisotropic) appear in the Autocorrelation Window,
on the Drift tab. Press the Expand button on this tab to bring up the Drift window:

The List, Edit, Print, Cloud, Scatter, and Exit buttons work here as they do for the
Variograms Window.

88

Chapter 7 Other Autocorrelation Measures

Correlograms
Correlograms are calculated using the formula
(h) = C(h) / (-h +h)
where
(h) = correlation for interval distance class h,
+h = standard deviation of all head values (zi) for lag h, and
-h = standard deviation of all tail values (zi+h) for lag h.
Correlograms (both isotropic and anisotropic) appear in the Autocorrelation Window,
on the Correlogram tab. Press the Expand button on this tab to bring up the Correlogram window:

The List, Edit, Print, Cloud, Scatter, and Exit buttons work here as they do for the
Variograms Window.

89

Chapter 7 Other Autocorrelation Measures

Covariance Analysis
Covariance in GS+ is computed as:
C(h) = [1 / N(h)] zi zi+h m-h m+h
where
C(h) = covariance for interval distance class h,
N(h) = total number of sample couples for the lag interval h,
zi = measured sample value at point i,
zi+h = measured sample value at point i+h,
m-h = mean of all head values for lag h, or [1/N(h)] zi, and
m+h = mean of all tail values for lag h, or [1/N(h)] zi+h.

Cross-covariance is computed in the same manner but zi and zi+h represent two different variates, Z and Z2, respectively.
The covariance graph appears in the Autocorrelation Window, on the Covariance
tab. Press the Expand button to bring up the Covariance window:

The List, Edit, Print, Cloud, Scatter, and Exit buttons work here as they do for the
Variograms Window.
90

Chapter 7 Other Autocorrelation Measures

General Relative Variograms

The general relative variogram is a graph of semivariance standardized by the
squared mean of the data used for each distance interval h:

GR(h) = (h) / (m-h + m+h)2

where

GR(h) = general relative semivariance for interval class h,

(h) = semivariance for interval distance class h,

m-h = mean of all head values for lag h, or [1/N(h)] zi, and
m+h = mean of all tail values for lag h, or [1/N(h)] zi+h.

General relative variograms (both isotropic and anisotropic) appear in the Autocorrelation Window, on the General Relative Variogram tab. Press the Expand button on
this tab to bring up the General Relative Variograms window:

The List, Edit, Print, Cloud, Scatter, and Exit buttons work here as they do for the
Variograms Window.

91

Chapter 7 Other Autocorrelation Measures

Pairwise Relative Variograms

The pairwise relative variogram is a graph of semivariance normalized by the
squared average of tail and head values:

PR(h) = [1 / 2N(h)] [ (zi zi+h)2 / [0.5(zi + zi+h)]2

where

PR(h) = pairwise relative semivariance for interval class h,

N(h) = total number of sample couples for the lag interval h,
zi = measured sample value at point i, and
zi+h = measured sample value at point i+h.
Pairwise relative variograms (both isotropic and anisotropic) appear in the Autocorrelation Window, on the Pairwise Relative Variogram tab. Press the Expand button
on this tab to bring up the Pairwise Relative Variograms window:

The List, Edit, Print, Cloud, Scatter, and Exit buttons work here as they do for the
Variograms Window.

92

Chapter 7 Other Autocorrelation Measures

Morans I Analysis
The Morans I statistic is a conventional measure of autocorrelation, similar in interpretation to the Pearsons Product Moment correlation statistic for independent
samples in that both statistics range between -1.0 and 1.0 depending on the degree
and direction of correlation. The statistic is defined as:
I(h) = N(h) zi zi+h / zi+h2
where
I(h) = autocorrelation for interval distance class h,
zi = the measured sample value at point i, and
zi+h = the measured sample value at point i+h.
Note that in this analysis all of the weights in the adjacency matrix (Sokal and Oden
1978) are set to 1, i.e. I is weighted by distance (h) between sample points rather
than by simple adjacency. Calculating this statistic for a variety of lag distances
yields the Morans I autocorrelogram.
Morans I correlograms (both isotropic and anisotropic) appear in the Autocorrelation
Window, on the Morans I tab. Press the Expand button on this tab to bring up the
Moranss I window:

The List, Edit, Print, Cloud, Scatter, and Exit buttons work here as they do for the
Variograms Window.
93

Chapter 7 Other Autocorrelation Measures

Fractal Analysis
The Fractal Analysis window is used to set up the parameters for fractal analysis
and then calculate the fractal dimension D for the data set defined in the Data Definition Menu. GS+ calculates the fractal dimension D as a function of the slope of a loglog variogram plot (Burrough 1981, 1986 ):
D=2-m/2
where
D = the Hausdorff-Besicovitch statistic, and
m = the slope of a log-log variogram
Because D is based on an analysis of semivariance, it is sensitive to the same
analysis parameters that affect semivariance analysis.
Fractal variograms (both isotropic and anisotropic) appear in the Autocorrelation
Window, on the Fractal tab. Press the Expand button on this tab to bring up the
Fractal window:

The List, Edit, Print, Cloud, Scatter, and Exit buttons work here as they do for the
Variograms Window.

94

Chapter 8 Variance Clouds and h-Scattergrams

Chapter 8
Variance Clouds and h-Scattergrams
Variograms can be sensitive to outlier values in a data set, and sometimes an erratic-looking variogram can be traced to a single extreme value that inflates (h)
whenever they are compared with other values. One way to locate such outliers is
to draw a variance cloud for each separation distance interval (or lag class) h; another way is to draw an h-Scattergram for each interval.
Variance clouds and h-Scattergrams are measures of pairwise variance in autocorrelation graphs. In a variogram each point on the graph represents the average
semivariance for all pairs of points in a particular lag class. h-Scattergrams and
variance clouds let you see the individual pairs of points that are used to calcualte
the average semivariance for that class.

Variance Clouds
The formula to calculate variance for any given pair of points at locations i and j can
be reduced to the mean square difference between the points:
Varij = [ zi zj ]2
where
Varij = variance for pair (i,j),
zi = measured sample value at point i, and
zj = measured sample value at point j.
Any given pair(i,j) are separated by a specific distance; this distance is plotted along
the x axis of the variance cloud graph. All pairs on a specific graph are in the same
separation distance (lag) class.
Note that a variance cloud is specific to both direction (isotropic or a specific anisotropic direction) and to a particular lag class. In the variogram below, the cursor is on
the point representing lag class 7 of the isotropic variogram, which one might suspect contains an outlier because it is so different from the other points:

95

Chapter 8 Variance Clouds and h-Scattergrams

Clicking on this variogram point or the Cloud button brings up the variance cloud for
lag class 7 below, and it becomes apparent that a number of pairs are very different
from the others placing the cursor over each of the outlying points reveals that all
of them contain record 4 as a member of the pair (the cursor below is over the point
represented by records 4 and 96 at a separation distance of 52.77):

96

Chapter 8 Variance Clouds and h-Scattergrams

Clicking on this variance cloud point brings up the Sample Details window, which
gives us the option to temporarily mask (remove from the active data set) one of the
data records for this pair:

Since record 4 is a member of all of these outlier pairs, we can choose to mask record 4, which has a transformed value of 2.639 (if the data were not transformed,
this value would match the value in the Data Worksheet). This gives us a much
more reasonable variogram:

Re-examination of the variance cloud for lag class 7 (below) reveals that the highest
y-axis values are substantially lower than before (0.35 vs. 6.12) and, more importantly, all of the major outlier pairs are gone. This was accomplished by removing a
single data record from the analysis:

97

Chapter 8 Variance Clouds and h-Scattergrams

Lag Class
The lag class for which the variance cloud is created. The variance for every pair of
points in the lag class is plotted against the distance interval separating that pair.
List Values
Bring up a listing of Variance Cloud Pairs, including for each pair the variance, separation distance, and the identity (data record number) of each member of the pair.
Edit Graph
Bring up a Graph Settings dialog window for editing the graph.
Print Graph
Print the graph via a Graph Print dialog window.
Cloud and Scatter
Choose either Variance Cloud or h-Scattergram analysis.

98

Chapter 8 Variance Clouds and h-Scattergrams

Variance Cloud Pairs

This read-only window contains a listing of all pairs of points within a specific lag
class, including for each pair the variance, separation distance, and the identity
(data record number) of each member of the pair. These pairs are graphed in the
Variance Cloud window.
Note that in the example below, sorted by variance (with pairs of highest variance
listed first), record 4 shows up as a member of 13 high-variance pairs. This confirms
record 4 as an outlier, which was determined in the variance cloud graph, above:

Actions
You may Print or Copy the contents of the worksheet using the menu commands of
the main GS+ window. You may also change the Decimal Places reported by highlighting a column and pressing the Increase or Decrease Decimals icon (or use the
Data | Change Decimals menu command). Clicking on the top of a column will Sort
the worksheet based on the column selected in alternating ascending or descending order. You may also change Column Widths by placing the cursor over the line
between two columns and dragging to a new location.

99

Chapter 8 Variance Clouds and h-Scattergrams

h-Scattergrams
In an h-Scattergram each head value zi is plotted against each tail value zj for a
given lag class h. The formula to calculate this difference for any given pair of points
at locations i and j is
Dij = zi zj
where
Dij = difference for pair (i,j),
zi = measured sample value at point I, and
zj = measured sample value at point j.
Note that an h-Scattergram is specific to both direction (isotropic or a specific anisotropic direction) and to a particular lag class. In the variogram below, displayed earlier, the cursor is on the point representing lag class 7 of the isotropic variogramwhich looks odd:

Clicking on this variogram point or pressing the Scatter button brings up the hScattergram for lag class 7 below, and it becomes apparent that a number of pairs
are very different from the others. Placing the cursor over each point reveals that all
of the points farthest from the 45-degree line contain record 4 as a member of the
pair (the cursor below is over the point represented by records 4 and 92 with a separation distance of 52.28):
100

Chapter 8 Variance Clouds and h-Scattergrams

Clicking on this h-Scattergram point brings up the Sample Details window, which
gives us the option to temporarily mask (remove from the active data set) one of the
data records for this pair:

Since record 4 is a member of all of these outlier pairs, we choose to mask record 4,
which has a transformed value of 2.639 (if the data were not transformed, this value
would match the value in the Data Worksheet). This gives us a much more reasonable variogram:

101

Chapter 8 Variance Clouds and h-Scattergrams

Re-examination of the h-scattergram for lag class 7 reveals that the greatest differences between head and tail values of individual pairs are substantially lower than
before (0.8 vs. 2.8) and all of the major outlier pairs are gone. This was accomplished by removing a single data record from the analysis:

102

Chapter 8 Variance Clouds and h-Scattergrams

Lag Class
The lag class for which the variance cloud is created. The variance for every pair of
points in the lag class is plotted against the distance interval separating that pair.
List Values
Bring up a listing of h-Scattergram pairs, including for each pair the variance, separation distance, and the identity (data record number) of each member of the pair.
Edit Graph
Bring up a Graph Settings dialog window for editing the graph.
Print Graph
Print the graph via a Graph Print dialog window.
Cloud and Scatter
Choose either Variance Cloud or h-Scattergram analysis..

103

Chapter 8 Variance Clouds and h-Scattergrams

h-Scattergram Pairs
This read-only window contains a listing of all pairs of points within a specific lag
class, including for each pair the difference, separation distance, and the identity
(data record number) of each member of the pair. These pairs are graphed in the hScattergram window.
Note that in the example below, sorted by difference (with pairs of greatest difference listed first), record 4 shows up as a member of 13 high-variance pairs. This
confirms the record as an outlier, as was found in the h-Scattergram above:

Actions
You may Print or Copy the contents of the worksheet using the menu commands of
the main GS+ window. You may also change the Decimal Places reported by highlighting a column and pressing the Increase or Decrease Decimals icon (or use the
Data | Change Decimals menu command). Clicking on the top of a column will Sort
the worksheet based on the column selected in alternating ascending or descending order. You may also change Column Widths by placing the cursor over the line
between two columns and dragging to a new location.

104

Chapter 9 Interpolation Basics

Chapter 9
Interpolation Basics
Interpolation is the estimation of values in an area for points not actually sampled.
There are many different interpolation techniques, ranging from simple linear techniques that average the values of nearby sampled points, to more complex techniques like kriging that use in the average weights based on distance to nearby
sample points and the degree of autocorrelation for those distances.
GS+ provides three broad types of interpolation. All are nearest-neighbor techniques
in which values at locations close to the interpolation point are used to estimate the
interpolation point value. They differ in the way that nearby locations are weighted
and interpolations calculated. The three techniques are:
Ordinary Kriging, in which interpolation estimates are based on values at
neighboring locations plus knowledge about the underlying spatial relationships in a data set. Semivariograms provide knowledge about the underlying
relationships. The estimated value at a given location is a weighted moving
average of best estimates calculated to minimize local area variance.
Conditional Simulation, in which interpolations are based on a form of stochastic simulation in which data values are honored at their locations. This
means that local details are not obscured by smoothing as they are in
kriging.
Inverse Distance Weighting (IDW) and Normal Distance Weighting
(NDW), in which interpolation estimates are made based on values at nearby
locations weighted only by distance from the interpolation location. Neither
IDW nor NDW make assumptions about spatial relationships except the basic assumption that nearby points ought to be more closely related than distant points to the value at the interpolate location. IDW applies stronger
weights to nearby points than does NDW.
All three techniques are described in greater detail on the following pages.

105

Chapter 9 Interpolation Basics

The Interpolation Window

The Interpolation Window provides access to Kriging, Conditional Simulation, and
Inverse Distance Weighting techniques for estimating values for points not sampled.
Interpolation produces an output file that is used by GS+ for mapping. The output
file can also be read by other mapping programs.

Interpolation Range
Defines where to place interpolation estimates in a uniformly spaced grid across a
rectangular area or at user-specified locations, in either case with or without masks
that can define areas to include or exclude.
Uniform x-y grid (specified intervals)
A grid is defined by a rectangle that has an X-direction length, a Y-direction
length, and for each direction, intervals between the grid intersections. Interpolation locations are at every grid intersection.
The default range is defined by the minimum and maximum X-coordinate and
Y-coordinate values, and an interval based on a certain number of points in
each direction. For 1-dimensional data sets, there is no y direction.
The grid can be changed with the Define command, which will display an Interpolation Grid dialog window.

106

Chapter 9 Interpolation Basics

Nonuniform x-y grid (specified points)
A nonuniform grid is composed solely of interpolation locations specified by
the user. Press Define to bring up an Interpolate Worksheet within which locations can be defined or imported from an external text file.
Include irregular shapes (polygons)
Irregular shapes can be interpolated or excluded from being interpolated by
defining inclusive or exclusive polygons prior to kriging. Press Define to
bring up a Polygon Outlines Worksheet within which to define polygons. See
Define Polygon Outlines later in this chapter.
In inclusive polygons, the area within the polygon is interpolated.
In exclusive polygons, the area within the polygon is not interpolated or mapped.
Constrain Z estimates
The estimated Z values (interpolates) can be constrained to a specific range.
For example, if values less than zero are inappropriate you can specify that Z
estimates less than zero be reported as zero. Press Define to bring up a
worksheet to specify Z Estimate Boundaries, described later in this chapter.
Output File Name
Press Select to select an existing or new file to which kriging interpolation estimates
will be written. To examine the contents of an existing file press View.
Output Format
The format with which GS+ will write estimates to the file can be one of several
types.
GS+ Krig format (.krg) in this format a header area defines the interpolation
grid, variate names, and other information about the file needed to initiate
mapping later, and the data records include for each X and Y Coordinate location that is kriged the interpolation or Z-estimate, the standard deviation of
the Z-estimate, and the number of neighbors that were used to make the estimate. See detailed example below.
Surfer Grid format (.grd) in this format a short header area defines information needed for mapping, and the data is written as a continuous stream
of Z-estimates beginning from a specific corner of the interpolation grid. The
107

Chapter 9 Interpolation Basics

standard deviation of the estimate and the number of neighbors used for interpolation are NOT included in this format. This format is compatible with
Golden Softwares Surfer mapping program. Note that this format is not the
same as the Surfer XYZ Input file format. See detailed example below.
ArcView Format (.asc) this is similar to the Surfer format but the header
area is formatted differently and the Z-estimates are written in a pattern that
begins from a different corner of the interpolation grid. The standard deviation of the estimate and the number of neighbors used for interpolation are
NOT included in this format. Also for this format, the x and y interpolation intervals must be the same (you can set them to be the same from the Interpolation Grid dialog window). This format is compatible with ESRIs ArcInfo
Geographic Information System. See detailed example below.
Search Neighborhood
GS+ interpolates values for a specific location using nearest neighbor values
weighted by distance and the degree of autocorrelation present for that distance (as
defined by the variogram model). Searches are limited to a certain number of nearest neighbors, and can also be restricted to a particular geographic radius. The
default value of 16 nearest neighbors is usually sufficient, with no restrictions placed
on radius (in Kriging, neighbors outside of the variogram range are weighted identically and, if significant structural dependence is present, weighted minimally).
Specifying more than 16 neighbors can slow interpolation substantially.
Interpolation Type Krig vs. IDW vs. Simulate
You may choose to interpolate using either Kriging, Conditional Simulation, or IDW
(Inverse Distance Weighting) by choosing one or the other tabs on the right hand
side of the window; the options on each tab are described in the next section.
Calculate
Perform interpolation analysis. During analysis values are written to the output file in
the specified format. You can stop analysis by pressing the <Escape> key.
Cross-Validate
Perform cross-validation analysis. In cross-validation analysis each measured point
in the spatial domain is individually removed from the domain and its value estimated via kriging or IDW as though it were never there. In this way a graph can be
constructed of estimated vs. actual values for each sample location in the domain.
Cross validation is not available for Conditional Simulation.

108

Chapter 9 Interpolation Basics

Define Z Estimate Boundaries

The Z Estimate Boundaries window is used to constrain estimated Z values (interpolates) to a specific range. It is accessed from the Interpolation Window.

Range to Use for Estimated Z

Enter the lowest permissible value for estimated Z in the lefthand box, and the highest permissible value in the righthand box. In this example, if the interpolation system were to estimate a value of -0.45 for a particular point, the value will be reported
as 0.
Reset
Reset values to the lowest and highest data values, respectively.
Cancel
Close window without saving changes.
OK
Save changes and close window.

109

Chapter 9 Interpolation Basics

Defining a Uniform Interpolation Grid

With the Uniform Interpolation Grid dialog window you may define the region to be
kriged or interpolated as well as the intensity at which the interpolation is to take
place. This method for specifying interpolate locations is appropriate for interpolations at regular intervals across an area; if the interpolations are to be performed at
odd locations, use the NonUniform Interpolation Grid. If the outline of the area is a
complex polygon you should still use this dialog but then specify the shape of the
polygon from the Krig window. Note that for 1-dimensional data sets (e.g. transects
or time series) only the X direction is displayed in the dialog window.

Interpolation Range
Specify the beginning and ending values for the region to be interpolated. The region may exceed the Data Range, which is noted on the right side of the dialog window. Irregularly-shaped areas within the bounds of the interpolation range may be
interpolated by specifying polygon masks from the Krig window.
Important Note: Interpolation will begin at the minimum X and Y values, but may
not reach the maximum values if the total distance (e.g. the maximum X value less
the minimum X value) is not evenly divisible by the Distance Interval. For example,
interpolation of an X Range of 0 105 with a Distance Interval of 10 will stop interpolating at X = 100, prior to exceeding the maximum range of 105.
Data Range
This is the range covered by the actual data set; these are read-only values, they
cannot be changed by the user from this dialog window.
Distance Interval
Specify the distance interval between locations within the interpolation range. A distance interval of 2.0 over a range of 0 to 10.0 means that interpolations will be made
at points 0.0, 2.0, 4.0, 6.0, 8.0, and 10.0. Changing the distance interval will change
the number of points value.
Number of Points
The number of points within the range to be interpolated. For any given range, the
110

Chapter 9 Interpolation Basics

number of points will be the range divided by the distance interval plus 1. For a
range of 0 to 10.0 with a distance interval of 2.0, the number of points will be [(10.0
0)/2.0] +1 = 6 points at locations 0.0, 2.0, 4.0, 6.0, 8.0, and 10.0.
Optimize
The Optimize command sets the number of points to 61 for both the X and Y directions (101 points for a single-dimension transect) and calculates the appropriate Distance Interval. The Interpolation Range is not affected.
Cancel
Exit the dialog window without saving changes.
Other Actions
You may Print or Copy the contents of the worksheet using the menu commands of
the main GS+ window. You may also change the Decimal Places reported by highlighting a column and pressing the Increase or Decrease Decimals icon (or use the
Data | Change Decimals menu command). Clicking on the top of a column will Sort
the worksheet based on the column selected in alternating ascending or descending order. You may also change Column Widths by placing the cursor over the line
between two columns and dragging to a new location.

111

Chapter 9 Interpolation Basics

Defining a NonUniform Interpolation Grid

Specify in this worksheet the locations of individual points to be kriged. This method
of specifying interpolate locations is appropriate when locations are not regularly
spaced across the interpolation area. If interpolate locations are regularly spaced,
regardless of whether the area is rectangular or a complex polygon, it is usually
more efficient to use a Uniform Interpolation Grid. Note that with either this NonUniform Grid or the Uniform Grid polygons can be used to exclude or include specific
complex shapes; specify the shape of the polygon from the Krig window (see Define
Polygon Outlines later in this chapter).

Block Size
For Block Kriging, this is the size of the block around each point that will be kriged.
For uniform grids, block size is equal to the interpolation grid size. The local (discretization) grid is placed within the box. For Point or Punctual Kriging, block size is
irrelevant and is ignored during analysis.
Clear
Clears the worksheet.
Import
The Import command brings into the worksheet a text file containing interpolate locations. The text file can be formatted in a variety of ways. Header records are initial records containing non-numeric data that are ignored during import. Data records follow header records (if present) and contain a value for the X-Coordinate
followed by a value for the Y-Coordinate; additional fields are ignored (as is the Y112

Chapter 9 Interpolation Basics

Coordinate field for 1-dimensional analysis).
The file below describes 6 locations to be interpolated:
line 1:
line 2:
line 3:
line 4:
line 5:
line 6:
line 7:
line 8:

Optional header record 1

Optional header record 2
2.0, 34.
4.0, 36.
8.0, 19
64.0, 34.
30.0. 35
10.0, 12.3

Other Actions
You may Print or Copy the contents of the worksheet using the menu commands of
the main GS+ window. You may also change the Decimal Places reported by highlighting a column and pressing the Increase or Decrease Decimals icon (or use the
Data | Change Decimals menu command). Clicking on the top of a column will Sort
the worksheet based on the column selected in alternating ascending or descending order. You may also change Column Widths by placing the cursor over the line
between two columns and dragging to a new location.

113

Chapter 9 Interpolation Basics

Polygon Outlines (Interpolation Masks)

Irregular shapes can be interpolated or excluded from interpolation by defining inclusive or exclusive polygons prior to interpolating. In this window you can define these
polygons by entering coordinates of polygon vertices, that is x-y points that outline
the polygon.
Polygons must be closed, thus they must have at least 3 vertices (3 vertices describe a triangle). You may specify as many vertices per polygon as you like (up to
several billion), and polygons may be nested within one another as in the example
below. However, the vertices for each polygon must define a single, closed polygon,
i.e. no segments of the polygon may cross another segment. Also the first and last
vertex specified must connect to one another. You can check the shape of the polygon while it is being defined with the Map command.
There are two types of polygons, defined by the keyword Include and Exclude:

In inclusive polygons, the area within the polygon is interpolated.

In exclusive polygons, the area within the polygon is not interpolated.

The example below defines two polygons: the first is a 6-sided area that is excluded
from interpolation, the second defines an inclusive 4-sided area (rectangle) inside
the 6-sided area. You can use the Map command to produce a picture of these
polygons (see below). The first line in this example (Polygon) is unnecessary. The
word "Exclude" or "Include" is not optional; each set of polygon vertices must be
preceded by either Exclude or Include:

114

Chapter 9 Interpolation Basics

Clear
Clear the worksheet.
Import
The Import command brings into the worksheet a text file containing vertex locations. The text file can be formatted in a variety of ways with fields separated by
either commas or spaces.
Print
Print the worksheet.
Map
Produces an outline map of each polygon within the larger interpolation grid area.
Exclusive polygons appear in red, inclusive in blue. See the example in the next section.
Other Actions
You may Print or Copy the contents of the worksheet using the menu commands of
the main GS+ window. You may also change the Decimal Places reported by highlighting a column and pressing the Increase or Decrease Decimals icon (or use the
Data | Change Decimals menu command). Clicking on the top of a column will Sort
the worksheet based on the column selected in alternating ascending or descending order. You may also change Column Widths by placing the cursor over the line
between two columns and dragging to a new location.

115

Chapter 9 Interpolation Basics

Polygon Outline Map

This window contains an outline map of the polygons used to include and exclude
areas from kriging. It is accessed from the Define Polygon Outlines window, which
is part of Kriging analysis.
The overall area is the interpolation area specified in the Interpolation window. Polygons are defined in the Define Polygon Outlines worksheet. Exclusive polygons appear in red, inclusive in blue.

Edit Graph
Make format and text changes to the graph via a Graph Settings dialog window.
Print Graph
Print the graph to a printer or file with the Graph Print dialog window.

116

Chapter 9 Interpolation Basics

Cross-Validation Analysis
Cross-validation analysis is a means for evaluating effective parameters for kriging
and IDW interpolations. In cross-validation analysis each measured point in a spatial domain is individually removed from the domain and its value estimated via
kriging or IDW as though it were never there. In this way a graph can be constructed of the estimated vs. actual values for each sample location in the domain.
Placing the cursor over a point provides information on the estimate for a specific
location as in the example below.

Each point on the graph represents a location in the input data set for which an actual and estimated value are available. Information about individual points is provided at the bottom of the screen; points are displayed by placing the cursor on
them; in the case above the cursor was placed about the point representing record
130.
The regression coefficient at the bottom of the graph represents a measure of the
goodness of fit for the least-squares model describing the linear regression equation. A perfect 1:1 fit would have a regression coefficient (slope) of 1.00 and the
best-fit line (the solid line in the graph above) would coincide with the dotted 45degree line on the graph. The standard error (SE = 0.105, above) refers to the
standard error of the regression coefficient; the r2 value is the proportion of variation
explained by the best-fit line (in this case 39.0%; it is the square of the correlation
coefficient); and the y-intercept of the best-fit line is also provided. The SE Prediction
term is defined as SD x (1 - r2)0.5 , where SD = standard deviation of the actual data
(the data graphed on the y-axis).

117

Chapter 9 Interpolation Basics

Cross-Validation Values
In cross-validation analysis each measured point in a spatial domain is individually
removed from the domain and its value estimated via kriging as though it were never
there. Results can be graphed as well as listed in this read-only worksheet table.
Note that Estimated Z values may or may not be backtransformed depending on
settings in the Data Summary worksheet. If you have chosen a transformation without a backtransformation, then the Estimated Z values will appear very different from
Actual Z values regardless of the integrity of the interpolation system.

Table
The record number refers to the actual record in the Data Worksheet. For each record successfully kriged the actual Z value is presented in the middle column and
the estimated Z value is provided to its right.
Other Actions
You may Print or Copy the contents of the worksheet using the menu commands of
the main GS+ window. You may also change the Decimal Places reported by highlighting a column and pressing the Increase or Decrease Decimals icon (or use the
Data | Change Decimals menu command). Clicking on the top of a column will Sort
the worksheet based on the column selected in alternating ascending or descending order. You may also change Column Widths by placing the cursor over the line
between two columns and dragging to a new location.

118

Chapter 10 Conditional Simulation

Chapter 10
Conditional Simulation
Conditional simulation is an advanced interpolation technique whereby Z estimates
are based on a form of stochastic simulation in which measured data values are
honored at their locations. Other interpolation methods, including kriging and IDW,
will smooth out local details of spatial variation, especially as interpolated locations
become more distant from measured locations. This can be a problem when you are
trying to map sharp spatial discontinuities such as contamination hotspots or fault
lines. GS+ uses a sequential gaussian simulation method.
The Interpolation Window contains the Simulate tab:

Variogram Model
Variogram models for isotropic and anisotropic variograms are chosen with the
Model command in the Autocorrelation Analysis window. Here in the Simulate window you can specify whether to use the isotropic or anisotropic model for the
variogram used in the kriging system.
Output
Two types of output are available with conditional simulation: either a) the estimated
Z value and its standard deviation, or b) the probability that the estimated value for
that location is greater than some threshold value t.
An individual Z estimate and its standard deviation is the mean and standard deviation of n simulations for a specific location. Thus the number of
simulations will affect both of these values.

119

Chapter 10 Conditional Simulation

Probability is the proportion of simulations for a specific location for which
the estimated value is greater than t. Thus this value will also be affected by
the number of simulations.
Analysis tab
Number of simulations. You may choose any number of simulations >0.
Keep in mind that choosing too few simulations will produce a map that is
very rough, although a large number of simulations may require substantial
cpu time. See later in this chapter for examples. The defualt number of simulations is 1000.
Use different seed. If this box is checked each simulation will use a different
random number seed. This will slow the analysis somewhat and make each
simulation unique to a specific run. The default is to use the same random
number seed. In either case a different random path is used for each simulation.
Multigrid refinements. Checking this box forces the analysis to follow a
stepwise procedure when simulating interpolation nodes. In the first step a
coarse grid is used to allow the influence of large-scale variogram structure;
in subsequent steps the search neighborhoods are smaller. This avoids the
need for extensive conditioning. The default is 3 refinements.
Secondary data tab
Collocated secondary data can be used to refine estimations. Collocated means
that a secondary data value is available for every grid node to be estimated (not
simply for every primary data value, as is the case for regular cokriging).

Residuals. This is the same as simple kriging with a locally varying mean.
Press the Define button to define the residuals to be used.
External drift. Press the Define button to define the drift values.
Colocated cokriging. Press the Define button to define the covariate values.

120

Chapter 10 Conditional Simulation

Secondary Data for Simulations

Secondary data for conditional simulation are defined in the Simulation Secondary
Data window. Secondary data must be available for every grid node to be estimated.This window is accessed from the Secondary data tab of the Simulate tab:

Clear
Clear existing data from worksheet.
Fill X,Y
Fill the X and Y Coordinate columns with calculated values based on the uniform
interpolation grid defined in the Interpolation window. If a uniform grid is not defined
in the interpolation window this button is not available.
Import
Import values from an external text file. Each record of the file must contain a separate x, y, and secondary data value separated by commas, tabs, or spaces. You
can also import data by copying and pasting into the Windows clipboard.
Other Actions
You may Print or Copy the contents of the worksheet using the menu commands of
the main GS+ window. You may also change the Decimal Places reported by highlighting a column and pressing the Increase or Decrease Decimals icon (or use the
Data | Change Decimals menu command). Clicking on the top of a column will Sort
the worksheet based on the column selected in alternating ascending or descending order. You may also change Column Widths by placing the cursor over the line
between two columns and dragging to a new location.
121

Chapter 10 Conditional Simulation

Number of Simulations
In conditional simulation the number of simulations used to produce estimates of Z
can strongly affect the outcome of the interpolation. Presented below are maps of
an 80 x 80 m grid interpolated at a density of 0.5 m, with different numbers of simulations. Below left is a map of block kriged data; on the right is the same data set
interpolated by conditional simulation (n = 1 simulation).

Additional simulations improve clarity; the map below left represents n = 10 simulations, below right n = 100 simulations:

And still more simulations resolve further detail. The map below left represents n =
1,000 simulations; below right n = 10,000 simulations:

You may wish to set the number of simulations high enough that additional simulations reveal little further resolution.
122

Chapter 11 Kriging and Cokriging

Chapter 11
Kriging and Cokriging
Kriging provides a means of interpolating values for points not physically sampled
using knowledge about the underlying spatial relationships in a data set to do so.
Variograms provide this knowledge. Kriging is based on regionalized variable theory
and provides an optimal interpolation estimate for a given coordinate location. GS+
performs ordinary kriging.
The Interpolation Window contains the Krig tab:

Krig Methods
Block vs. Point Kriging. You may choose either block or point (punctual) kriging.
Your choice should be made on the basis of sampling design and variate characteristics. If samples were taken to represent an area around the actual sample point
(e.g. if soil samples from an area around the sampling coordinate were composited
before analysis), then block kriging may be more appropriate than punctual. If samples were taken to represent point values then punctual kriging may be more appropriate. For uniform grids, block size is equal to the interpolation grid size. For nonuniform grids, block size is defined in the
Local grid. The Discretization or Local grid describes the size of the grid
placed around the interpolation point and inside the block when block kriging.
The interpolation estimate for that point is based on the mean value of estimates for each of the discretization grid points. A single discretization point
describes punctual kriging. Larger discretization grids take longer to interpolate.
Cokrig. Cokriging provides a means for interpolating values using a secondary variate Z2 that covaries with the primary variate Z. The covariate is
sampled from the same locations as Z, and also from a number of additional
locations. The kriging estimate is based not only on distance to nearby sam123

Chapter 11 Kriging and Cokriging

ple locations for Z and the variogram for Z, but also distance to nearby sample locations for Z2, the variogram for Z2, and the cross-variogram for Z x
Z2. This can provide a more robust estimate of Z at unsampled locations so
long as Z and Z2 covary sufficiently. See below for an example.
Before cokriging you must a) define a covariate in the Data Worksheet Field
Assignment dialog, b) perform semivariance analysis (including variogram
modeling) for the primary variate and the covariate, and c) perform crosssemivariance analysis (including variogram modeling). You will also want to
check the degree of correlation between the variate and covariate in the Data
Summary Covariate window.
Variogram Model Type
Variogram models for isotropic and anisotropic variograms are chosen with the
Model command in the Variogram section of the Autocorrelation Analysis window.
Here in the Kriging window you can specify whether to use the isotropic or anisotropic model for each of the variograms used in the kriging system. If you are not
cokriging then you will be able to specify a model for the primary variate only.

124

Chapter 11 Kriging and Cokriging

Cokriging
Cokriging is an interpolation technique that allows one to better estimate map values
if the distribution of a secondary variate sampled more intensely than the primary
variate is known. If the primary variate is difficult or expensive to measure, then cokriging can greatly improve interpolation estimates without having to more intensely
sample the primary variate.
Consider the following example. After a radioactive spill, plutonium was sampled
across an 80 x 80 m area at a sample density indicated by the quartile plot below.
Block kriging (following variography) resulted in the adjacent map:

Soil carbon, easier to measure than Uranium, was sampled at the same locations as
Uranium and additionally at another 60 locations as noted in the quartile map below
left. Regression of carbon against plutonium showed that the variates were highly
correlated (right), suggesting that cokriging might improve the map of plutonium.

Using carbon as a covariate to produce a cokriged map of plutonium results in the

below-right map, plotted next to the original kriged map. Note the substantial improvement in the definition of contour (isoline) differences, especially in the upper
right quadrant of the map where the uranium was sampled most sparsely:

125

Chapter 11 Kriging and Cokriging

How do you perform cokriging? Prior to cokriging you must a) define a covariate in
the Data Worksheet Field Assignment dialog, b) perform semivariance analysis (including variogram modeling) for both the primary variate Z and the covariate Z2, and
c) perform cross-semivariance Z x Z2 analysis (including variogram modeling). You
should also check that the covariate is in fact correlated with the primary variate.

126

Chapter 12 Inverse and Normal Distance Weighting

Chapter 12
Inverse (IDW) and Normal (NDW) Distance Weighting
Inverse Distance Weighting (IDW) and Normal Distance Weighting (NDW) are interpolation techniques in which interpolated estimates are made based on values at
nearby locations weighted only by distance from the interpolation location. Neither
IDW nor NDW make assumptions about spatial relationships except the basic assumption that nearby points ought to be more closely related than distant points to
the value at the interpolate location. IDW applies stronger weights to nearby points
than does NDW. The formula used for Inverse Distance Weighting is:
Zestj = [ zi / (hij + s)p] / [1 / (hij + s)p]
where
Zestj = estimated value for location j
zi = measured sample value at point i,
hij = distance between Zestj and zi;
s = smoothing factor; and
p = weighting power.

The formula used for Normal Distance Weighting is the same as that for IDW except
that the denominator [distance plus smoothing factor] rather than [inverse of distance plus smoothing factor].
The weighting power p defines the rate at which weights fall off with hij, the distance
between the interpolated and sample locations. A value of 1-5 is typical.
The smoothing factor s reduces the likelihood that any one sample value will overly
influence an estimated value for a given interpolation location. IDW is an exact interpolator, so where an interpolation location j coincides with a sample location i,
Zestj = zi and a sharp peak or valley may result; setting s > 0 reduces this peaking effect when it occurs.

127

Chapter Chapter 12 Inverse and Normal Distance Weighting

The Interpolation Window contains the IDW tab:

Inverse vs. Normal Distance Weighting

Choose either inverse distance weighting or normal distance weighting. Inverse
weighting applies stronger weights to nearby points than does Normal Distance
Weighting.
Weighting Power
The weighting power defines the rate at which weights fall off with distance between
the interpolated and sample locations. A value of 1-5 is typical.
Smoothing Factor
The smoothing factor s reduces the likelihood that any one sample value will overly
influence an estimated value for a given interpolation location.
Reset
Restore default values to the IDW method.

128

Chapter 13 Interpolation Output File Formats

Chapter 13
Interpolation Output File Formats
GS+ Format Output Files
The GS+ format for Krig (.krg), IDW, and Conditional Simulation output files (which is
also the GS+ format for Map input files) contains a number of header records containing information about the file, followed by XYZ-style records for each interpolated
points. Each data record contains fields for an x-coordinate value, a y-coordinate
value (except 1-dimensional transects do not contain a value for the y-coordinate),
an estimated z value for that x-y location, an estimation standard deviation for the
estimated z-value (except IDW files), and the number of pairs used in the interpolation for that x-y location. Missing values are denoted by whatever value is specified
in the Preferences window. Note that this format is different from the GS+ Input File
format.
The following listing is for the first 6 coordinates for a standard GS+ krig output file. A GS+ IDW Output file would look identical except that there would
be no field for Estimated Standard Deviation in the data records.
GS+ Output: Block Kriging Interpolation File v7.0
Set: Field 54, Second Tier
Dimensions: 2; Interval source: Calculated
Interpolation interval (x; y): 1.3333; 1.3333
X-coor: m east
range: 0.00 - 80.00
Y-coor: m north
range: 0.00 - 80.00
Z-est: Pb
range: 0.151 - 0.813
Z-sd:
range: 0.0000 - 0.3655
Mean Z-estimate (sd): 0.391 (0.0186)
Valid N: 3721; Missing N: 0; Missing Value Indicator: -99
X-Coordinate
Y-Coordinate
Z-Estimate
EstStdDev
0.00
0.00
0.431
0.3608 16
0.00
1.33
0.458
0.3486 16
0.00
2.67
0.462
0.3329 16
0.00
4.00
0.466
0.3178 16
0.00
5.33
0.470
0.3034 16
0.00
6.67
0.474
0.2900 16

129

Chapter 13 Interpolation Output File Formats

ArcView and ArcInfo Format
The ArcView Ascii (.asc) format for Krig output files (which is also the Arc/Info
format for Map input files) contains a number of header records containing information about the file, followed by a string of z values for each row of the interpolated
grid beginning in a particular corner. Unlike GS+ format files, the standard deviation
of the estimate and the number of neighbors used for interpolation are NOT included
in this format. Also for this format, the x and y interpolation intervals must be the
same (you can set them to be the same from the Interpolation Grid dialog window).
This format is compatible with ESRIs ArcInfo Geographic Information System. Data
in this format can be imported directly into ESRIs ArcInfo or ArcView. Note that
this format is different from the ArcView XYZ Input File format described earlier,
which is an ArcView output format that can be read by GS+.
The following listing is for the first 200 or so coordinates in a standard
Arc/Info or ArcView format krig output file:
ncols
376
nrows
253
xllcorner -185556.375
yllcorner -127261.5234375
cellsize
1009.975
NODATA_value -999
354 314 309 301 286 264 305 306 285 285 268 314 339 306 266 248 252
240 256 282 277 289 269 285 277 258 256 282 268 246 250 249 245 266
262 287 295 335 325 323 359 369 361 357 394 450 407 409 382 409 394
486 510 502 516 546 531 542 572 579 586 594 522 654 550 615 678 709
616 533 430 576 507 638 778 769 456 432 576 635 778 721 769 869 563
589 640 588 770 833 874 1019 965 933 621 795 1131 1044 899 1072
1112 880 910 1069 1088 1071 965 975 902 800 543 350 330 376 310
384 418 352 263 205 202 200 200 201 200 198 201 201 201 201 201 201
694 695

To convert this file to an ArcGIS grid file, use ArcToolbox, either the light version that comes with ArvView or the full version that comes with ArcInfo:
a. Open ArcToolbox
b. Choose the Import to Raster tool
c. Specify the GS+ ouptut file name, the grid type (integer or floating point),
the name of the grid output file, and press OK.
d. You may now open the file in ArcMap.

130

Chapter 13 Interpolation Output File Formats

131

Chapter 13 Interpolation Output File Formats

Surfer Grid (.GRD) Format
The Surfer Grid (.grd) Krig Output format (or Map Input format) has a short header
area that provides information needed for mapping, and the data is then written as a
continuous stream of Z-estimates beginning from a specific corner of the interpolation grid. The standard deviation of the estimate and the number of neighbors used
for interpolation are NOT included in this format. This format is compatible with
Golden Softwares Surfer mapping program. Note that the Surfer Grid format is
different from the Surfer XYZ format used for input files (see Chapter 4).
The following listing is for the first 100 or so coordinates in a standard
Surfer Grid format file. The first record is a required line of code letters that
identifies the file as a Surfer grid file. Record two contains the number of
grid lines along the x axis and y axis, respectively. Record 3 contains the
minimum and maximum values of the x-coordinate. Record 4 contains the
minimum and maximum values of the y coordinate. Record 5 contains the
minimum and maximum values for the z-variate. Records 6+ contain the z
values organized in row order. Within each row the y-coordinate is constant,
and the grid row 1 corresponds to the lowest y-coordinate value and the last
row corresponds to the hightest y-coordinate. Within each row the z-variate
values are ordered from low x-coordinate to highest x-coordinate.
DSAA
376 253
-185556.375
194194.225
-127261.523
128262.152
81.000
4469.000
578 746 1097 1251 581 312 229 229 235 278 292 357 490 997 1411
1397 1373 1373 1370 1061 942 1003 1288 1014 785 963 934 797 703
924 1035 1071 1257 1775 1800 1520 1460 1520 1278 1098 915 645 495
450 378 339 342 425 613 1084 1292 1701 1682 1996 2280 2107 1662
1553 1418 1162 1225 1266 1735 1674 1550 1800 2015 2240 2158 2328
2262 2173 2046 1688 1849 1880 2023 2320 2029 2140 2240 2650 2465
2369 2180

132

Chapter 13 Interpolation Output File Formats

133

Chapter 14 Mapping

Chapter 14
Mapping
GS+ produces 2D and 3D maps of spatial data following interpolation. The data to be
mapped come from kriging or IDW analysis, and are thus contained in interpolation
output files. Maps can be displayed in a variety of ways with a variety of different
contouring schemes, 3-dimensional maps can be rotated on the fly, and both 2dimensional and 3-dimensional maps can be zoomed to more closely view a transition or other map feature. Additionally, sample postings (original data locations) can
be displayed, and estimation standard errors can be mapped for input files that have
been saved in the standard GS+ format.

Map Input File

Select allows you to choose the file with the data to be mapped. The default file extension will correspond to the selected file format (e.g. .krg, .blk,
and .pun for GS+ format), although a file with any extension can be selected.

134

View view the selected file in the View File window.

Format specify the format to use to read the input file. Files with a .krg,
.idw, .blk, or .pun extension are assumed to be in GS+ format. The format
specified in this field will override any assumed format:

Chapter 14 Mapping
+

GS Krig format (.krg) in this format a header area defines the interpolation grid, variate names, and other information about the file needed to
initiate mapping later, and the data records include for each X and Y
Coordinate location that is kriged the interpolation or Z-estimate, the
standard deviation of the Z-estimate, and the number of neighbors that
were used to make the estimate. See detailed example in Chapter 11.
Surfer Grid format (.grd) in this format a short header area defines
information needed for mapping, and the data is written as a continuous
stream of Z-estimates beginning from a specific corner of the interpolation grid. The standard deviation of the estimate and the number of
neighbors used for interpolation are NOT included in this format. This
format is compatible with Golden Softwares Surfer mapping program.
Note that this format is not the same as the Surfer XYZ Input file format.
See detailed example in Chapter 11.
ArcView Format (.asc) this is similar to the Surfer format but the
header area is formatted differently and the Z-estimates are written in a
pattern that begins from a different corner of the interpolation grid. The
standard deviation of the estimate and the number of neighbors used for
interpolation are NOT included in this format. Also for this format, the x
and y interpolation intervals must be the same (you can set them to be
the same from the Interpolation Grid dialog window). This format is
compatible with ESRIs Arc-Info Geographic Information System. See
detailed example in Chapter 11.
Map Grid
X direction the range within the file for values in the x (horizontal) direction.
Y direction the range within the file for values in the y (vertical) direction; for
1-dimensional data sets this field is blank.
Z values the range of estimated Z values in the input file.
Z SD values the range of estimation Z standard deviation values in the input file; only when input files are in a GS+ format that contain both Z values
and Z standard deviation values will this box be selectable;
N (N missing) the number of valid records in the file and (in parentheses)
the number of missing values. The default missing value indicator (e.g. 999.) is specified in the Preferences - General dialog; the active missing
value indicator for map files is indicated in the Map Contour Intervals window.
Missing values are mapped as transparent regions. Exclusive polygons are
mapped as missing values, as is any interpolate location for which there is no
kriging solution.

135

Chapter 14 Mapping
Variate to Map
Z values maps the estimated Z values in the file

Z standard deviations maps the estimation Z standard deviations in the file;

only files in the GS+ format contain both Z values and Z standard deviation
values.

Sample Posting map only the original sample locations. Requires the
presence of a posting file, which has the same name as the input file but a
.pos extension. This file is created during interpolation; if it is not present you
will not be able to map a sample posting. Coordinate Postings (Quantile
Plots) of the active data are viewed through the Data Summary X,Y Coordinates Tab window, described earlier.

Graph Type
2D displays a flat, 2-dimensional map of the data

3D displays a 3-dimensional map of the data. The height of the map can
be adjusted with the Edit Graph command of the Map window; the perspective can be adjusted with the Rotate command of the Map window.

1-d displays 1-dimensional data (e.g. a geographic transect or a time series) as an x-y graph. This choice is not available for 2-dimensional data.

Show legend displays the contour legend next to the map.

Surface
Contour lines draws lines between contour intervals

136

Color bands fills the space between contour lines with different colors; colors can be specified from the Define Contour dialog window by pressing Define Contour Levels, below.

Smoothing apply slight smoothing to the data to improve visualization.

Solid pedestal for a 3D map fill in the area beneath the surface with a solid
color.

Wireframe for 3D maps, drapes an x-y grid over the surface.

Wireframe Weave specifies the density of the wireframe grid. A weave of 0

puts a grid line at every data row and column; a weave of 1 skips one
row/column, a weave of 2 skips 2, etc.

Chapter 14 Mapping
Contour Levels
Number the number of contour levels to put on the map.

Define brings up the Map Contour Intervals dialog window (described in

greater detail below) that allows you to set break points for individual intervals and colors for contour bands.

Map Legend
Show legend displays the contour legend next to the map.
Continuous vs. Stepped display legend as continuous scale with values
next to break points (e.g. between color bands) or display legend as stepped
format, with values next to boxed colors.
Ceiling
Contour lines project contour lines above the plot surface (3D maps only).

Color bands project color bands onto the ceiling of the plot (3D maps only).

Floor

Contour lines project contour lines under the plot surface (3D maps only).

Color bands project color bands onto the floor of the plot (3D maps only).

Grid Lines
X axis place a vertical grid line along the back walls of the 3D plot x axis.

Y axis place a vertical grid line along the back wall of the 3D plot y axis.

Z axis place horizontal grid lines along the back walls of 3D plots.

Draw
Create the map in a Map Image window.

137

Chapter 14 Mapping

Map Contour Intervals

This dialog window allows one to specify the break points and colors for the contour
intervals used for mapping. The number of contour intervals is specified in the Map
window, where access to this Map Contour Levels window is provided.

Contour Intervals
Color buttons change the color of a specified interval by clicking the adjacent color button.

138

Break points change the break point between adjacent contour intervals by
providing a new value in the space provided. Note that the new break point
must be greater than the preceding point and less than the one that follows.

Missing value specify the missing value indicator for the data file. This
value is usually embedded in the input file as part of the header information,
and is automatically extracted by GS+. This value overrides the missing value
indicator specified in the Preferences General dialog or the map input file.

Chapter 14 Mapping
Intervals
Get read contour intervals from a text file; the first value of each record of
line of the file is presumed to be a separate interval.

Save save the existing contour intervals to a text file.

Reset recalculate intervals based on a uniform distribution (even intervals)

between the lowest and highest value in the file.

Color Source
Standard specifies a default color pattern of red-orange-yellow.. violet.

Blue specifies a default color pattern of different shades of blue.

Custom specifies a custom color pattern defined below.

Red specifies a default color pattern of different shades of red.

Green specifies a default color pattern of different shades of green.

Gray specifies a default color pattern of different shades of gray.

Color Source Custom

Get get a custom color pattern saved earlier. The initial custom color is the
same as the standard color source, above.

Save save the existing color pattern as the custom color pattern.

Reset reset the custom color pattern to the default Standard pattern.

Color Source Dynamic

Expands or contracts the color range when there are fewer than 15 intervals specified. For example, if 10 intervals are specified and the Dynamic box is not checked,
the colors for the intervals are the same as they would be if 15 intervals were
checked. If, on the other hand, the Dynamic box is checked, the colors for the intervals are adjusted such that interval 10 becomes the color of the original interval 15.
Checking this box provides greater color contrast on maps.
Invert
Flips the color range. This allows you to associate the highest interval values with
the darkest colors or vice versa.
Cancel
Exit the dialog window without saving changes.

139

Chapter 14 Mapping

Map Image - 3D
The 3D map image is produced by the Draw command from the Mapping window.

Edit Graph
Change axis scales and other graph formats via the Graph Settings dialog window.
Print Graph
Print the active graph to a file, printer, or other device.
Mouse Action
Off returns the mouse to normal operation.

Rotate turns the cursor into a rotator cuff when the left mouse button is
pushed, allowing the image to be rotated as desired.

Move allows the graph to be moved within the window by clicking the left
mouse button and dragging the cursor.

Scale shrinks the graph image with the left mouse button.

Zoom allows you to zoom in on a graph area by using the left mouse button
to define a rectangular zoom area. Within the zoomed area the location of
the cursor is noted on the Mouse Location panel.

Reset resets the image to the default rotation angle and scale.

Mouse Location
The current cursor location when on the map surface. Units are map units.
140

Chapter 14 Mapping

Map Image - 3D Standard Deviations

A map of standard deviations can be produced for interpolation files that contain
standard deviation values Krig or Conditional Simulation interpolations that are
saved in GS+ format. Use the Draw command from the Mapping window.

Commands in this Window are the same as for earlier graphs.

141

Chapter 14 Mapping

Map Image - 3D Rotation

Three-dimensional map images can be rotated to any viewing angle. To rotate,
choose Rotate on the Mouse Action panel and hold down the left-mouse button
while moving the mouse over the graph. The cursor will be replaced with a 3D box
(as below) of the viewing angle. Release the mouse and the map will be drawn:

142

Chapter 14 Mapping

Map Image - 2D
The 2D map image is produced when the Graph Type is set to 2D.

Map Image - 2D Standard Deviations

A 2D map of standard deviations can be created from the Map Window.

143

Chapter 14 Mapping

Map Image - 2D Sample Posting

The 2D sample posting is produced by the Draw command from the Mapping window when the Graph Type within that window is set to 2D and the Posting box is
checked. Each X symbol in the image below marks an actual sample location as
defined by the original X,Y coordinate locations in the Data Worksheet window. Coordinate Postings (Quantile Plots) of the active data (which may or may not be the
same as the mapped data) are viewed through the Data Summary X,Y Coordinates Tab window, described earlier.

Commands in this Window are the same as for earlier graphs.

144

Chapter 14 Mapping

Transects (1d Maps)

The 1D transect image with or without sample postings is produced via the Mapping
window when the variate to be mapped has only one (X) dimension:

Transect Image with Standard Deviation

A transect with standard deviations can be drawn if interpolation files contain standard deviations Krig or Conditional Simulation files in GS+ format.

Commands in this Window are the same as for earlier graphs.

145

Chapter 15 Bibliography

Chapter 15
Bibliography
The following references may be useful for those seeking further background about
geostatistics and its use in the environmental sciences.
Burrough, P.A. 1981. Fractal dimensions of landscapes and other environmental data. Nature 294:240-242.
Burrough, P.A. 1986. Principles of Geographical Information Systems for
Land Resources Assessment. Oxford University Press, Oxford.
Burgess, T.M. and R. Webster. 1980a. Optimal interpolation and isarithmic
mapping of soil properties. I. The semivariogram and punctual kriging. Journal of Soil Science 31:315-331.
Burgess, T.M. and R. Webster. 1980a. Optimal interpolation and isarithmic
mapping of soil properties. II. Block kriging. Journal of Soil Science 31:333341.
Cressie, N. 1985. Fitting variogram models by weighted least squares.
Mathematical Geology 17: 563-586.
Cressie, N. A. C. 1991. Statistics for Spatial Data. John Wiley, New York,
USA.
David, M. 1977. Geostatistical Ore Reserve Estimation. Elsevier, Scientific
Publishing Co., Amsterdam, The Netherlands.
Deutsch, C.V. and A.G. Journel 1992. GSLIB Geostatistical Software Library.
Oxford University Press, New York.
Goovaerts, P. 1997. Geostatistics for Natural Resources Evaluation. Oxford
University Press, New York.
Griffith, D.A. 1987. Spatial Autocorrelation: A Primer. Association of American Geographers, Washington, D.C. 86 p.
Haan, C.T. 1977. Statistical Methods in Hydrology. Iowa State University
Press, Ames, Iowa.
Isaaks, E.H. and R.M. Srivastava. 1989. An Introduction to Applied Geostatistics. Oxford University Press, NY.
Journel, A.G. and C.J. Huijbregts. 1978. Mining Geostatistics. Academic
Press, New York.
Krige, D.G. 1966. Two dimensional weighted moving average trend surfaces
for ore-evaluation. Journal of the South African Institute of Mining and Metallurgy 66:13-38.
Krige, D.G. 1981. Lognormal-de Wijsian geostatistics for ore evaluation.
South African Institute of Mining and Metallurgy Monograph Series. Geosta146

Chapter 15 Bibliography
tistics I. South Africa Institute of Mining and Metallurgy, Johannesburg, South
Africa.
Mandelbrot, B.B. 1982. The Fractal Geometry of Nature. W.H. Freeman,
London.
Matheron, G. 1971. The theory of regionalized variables and its applications.
Cahiers du Centre de Morphologie Mathematique, Fontainebleau, No. 5.
Robertson, G.P. 1987. Geostatistics in ecology: interpolating with known
variance. Ecology 68:744-748.
Robertson, G. P., and K. L. Gross. 1994. Assessing the heterogeneity of below-ground resources: Quantifying pattern and scale. Pages 237-253 in M.
M. Caldwell and R. W. Pearcy, eds. Plant Exploitation of Environmental Heterogeneity. Academic Press, New York, New York , USA.
Rossi, R. E., D. J. Mulla, A. G. Journel, and E. H. Franz. 1992. Geostatistical
tools for modeling and interpreting ecological spatial dependence. Ecological
Monographs 62: 277-314.
Sokal, R.R. and N.L. Oden. 1978. Spatial autocorrelation in biology. 1. Methodology. 2. Some biological implications and four applications of evolutionary and ecological interest. Biological Journal of the Linnean Society 10:199228.
Trangmar, B.B., R.S. Yost and G. Uehara. 1985. Applications of geostatistics
to spatial studies of soil properties. Pages 45-94 in N.C. Brady, editor. Advances in Agronomy Volume 38. Academic Press, New York.
Vieira, S.R., J.L. Hatfield, D.R. Nielsen, and J.W. Biggar. 1983. Geostatistical
theory and application to variability of some agronomical properties. Hilgardia
51:1-75.
Webster, R. 1985. Quantitative spatial analysis of soil in the field. Pages 170 in B.A. Stewart, editor. Advances in Soil Science Volume 3. SpringerVerlag, New York.
Webster, R. and M.A. Oliver. 1990. Statistical Methods in Soil and Land Resource Survey. Oxford University Press, NY. 316 p.

147

Chapter 16 Glossary

Chapter 16
Glossary
A0
Range parameter in variogram models. The relationship of A0 to range A (effective
range) depends on the model. See specific model definitions for more information.
A1
Range parameter for the major axis in anisotropic variogram models. The relationship of A1 to major range depends on the model. See specific model definitions for
more information.
A2
Range parameter for the minor axis in anisotropic variogram models. The relationship of A2 to major range depends on the model. See specific model definitions for
more information.
Binary Data Record Format
Binary data from many spreadsheet and database programs can be imported directly; if the file can be viewed using the View command then it can be imported.
Eligible files include Excel and Lotus 1-2-3 worksheets, dBase, Paradox, Access,
etc.
A Character as Missing Value Indicator
Specify the character to be used as a missing value indicator in the field that will appear when a character value is specified in the list box. The example below presumes that % has been specified the missing value indicator, and thus the third
value of this record will be read into the worksheet as a missing value:
13.2, 34.5, % , 0.15
Autocorrelation
Autocorrelation is the degree to which a property is related to itself in time or space.
For example, in a spatial domain values for samples taken close to one another are
more likely to be similar than are samples taken farther apart autocorrelation is a
formal measure of this self-similarity.
Backtransformation
When data values are transformed in order to make their distributions more normally
distributed, after analysis of the transformed data the output data are customarily
(but not necessarily) back-transformed to the original data domain for final reporting.
For example, the backtransformation for a ln(z) transform is exp(z-transformed); for
a z2 transform the backtransformation is the square root of the transformed value.
Offset values (if applied) are subtracted from the backtransformed values.
Binary Data Record Format
Binary data from many spreadsheet and database programs can be imported di148

Chapter 16 Glossary
rectly. Eligible files include Excel and Lotus 1-2-3 worksheets, dBase, Paradox, Access, and FoxPro.
Block Kriging
Kriging provides a means of interpolating values for points not physically sampled
using knowledge about the underlying spatial relationships in a data set to do so.
Block kriging provides an estimate for a discrete area around an interpolation point.
Punctual or point kriging provides an estimate for a precise point. In environmental
work block kriging is usually more appropriate. The block is defined as the rectangular area around a point that is not included in an adjacent block.
Brackets as Column Title Separators
Column or variate names are separated by brackets, e.g.:
<m east> <m north> <Pb (ug/g)> <pH>
C or Structural Variance
C of the variogram model represents spatially-structured variance. Compare to C0
or nugget variance, which is the portion of the variance not spatially structured.
C0 or Nugget Variance
C0 or Nugget Variance is the y-intercept of the variogram model. Nugget variance
represents variation not spatially dependent over the range examined.
Characters as Column Title Separators
Column or variate names are separated by a character specified in the field that appears when this option is selected. For example if "#" were specified as the separated:
m east # m north # Pb (ug/g) # pH
Character Separated Values
Values within data records are separated by a specific character, defined in the field
below the list box; e.g. if the character # is the delimiter:
13.2 # 34.5 # 35.6 # 0.15
Cokriging
An interpolation technique that allows one to use a more intensely sampled covariate in the estimation of values for a related variate. If the primary variate is difficult
or expensive to measure and it is correlated with a more available covariate, cokriging can greatly improve interpolation estimates. See Cokriging Analysis for further
information.
Commas as Column Title Separators
Column or variate names are separated by commas e.g.:
m east, m north, Pb (ug/g), pH
149

Chapter 16 Glossary
Comma Separated Values
Values within data records separated by commas:
13.2, 34.5, 35.6, 0.15
Conditional Simulation
An advanced interpolation technique whereby Z estimates are based on a form of
stochastic simulation in which measured data values are honored at their locations.
Covariate
The Covariate is a second Z-variate that covaries with the primary Z-variate. It is
used in cokriging.
Data Builds
Data in the GS+ Worksheet is read into the actual arrays that get analyzed during
Data Builds. Data builds occur whenever a new data set is read into GS+ (via the
Import command) or whenever a field assignment or data value is changed and the
Rebuild Button (Data Worksheet Window) is pressed. During data builds missing
values are ignored and filters and transformations are applied. Most Data Builds are
automatic; youll need to force a Data Build whenever the Rebuild Button is red (the
Data Summary Window will also be empty).
Data Fields
A data field is an individual item on the data record. In the GS+ Worksheet, records
appear as rows and fields appear as columns. Prior to analysis you must identify an
X-coordinate field, a Y-coordinate field (for 2-dimensional analyses), and the field
(column) that corresponds to the Z-variate. For cokriging you will also need to assign a field to hold the covariate (Z2) values. Field assignments are performed by
clicking on the top row of the Worksheet to bring up a Field Assignment dialog window.
Data Records
Data Records are records in text files that contain data values. Data records follow
header records that contain descriptive information about the file. Data records are
imported into the GS+ Worksheet; header records are imported into the Data Description portion of the Worksheet Window.
Decimal Point Missing Value Indicator
In the record below the third value appears as a decimal point and will be read into
the worksheet as a missing value:
13.2, 34.5, . , 0.15
Discretization Grid
The discretization grid describes the size of the grid placed around the interpolation
point when block kriging. The estimate for the interpolated block is based on the
mean value of estimates for each of the discretization grid points. A single discretization point describes punctual kriging. A 2x2 grid averages 4 grid points. Larger
150

Chapter 16 Glossary
discretization grids take longer to interpolate.
Effective Range
See Range.
Geo-referenced Data
Geo-referenced data are data that have been collected from a specific location, i.e.
any data for which there is a spatial x,y coordinate associated with each data value.
Header Records
Records at the top of a data file that precede the data records and that contain information about the data records.
Interpolate
An interpolate is an estimated value for a location in a domain based on values of
nearby measured locations. In GS+ interpolation is performed by kriging or by inverse distance weighting methods. Maps are based on interpolations at grid points
across the domain.
Kriging
An interpolation method based on regionalized variable theory that provides an optimal interpolation estimate for a given coordinate location. GS+ performs ordinary
kriging.
Lag Class Distance Interval
Each point on a variogram corresponds to the average semivariance for all values in
the spatial domain that are separated by a given distance interval h. This interval
(also called a lag distance class) is a range whose width is specified by the user in
the Semivariance Analysis window. A corresponding definition is also relevant to
Morans I, Fractal, Cross-Semivariance, and Madogram Analyses.
Missing Value
A value not present in a record. A data record can contain fields for coordinates (e.g.
X, Y) and for multiple Z variates. If any one of the fields used for an analysis is missing, the record is ignored during analyses, i.e. when the active data set is built. In the
GS+ Worksheet a value is considered missing if the cell is blank (a permanent missing value) or if the cell contents appear as red italicized text (a temporarily missing
value). Cells can be assigned temporarily missing by right-clicking on them; a second right-click returns them to non-missing. In Input and Output files missing values
are indicated by characters or values (e.g. -99) defined in the Preferences window.
Model Parameters
See Variogram Model Parameters.
No Header Records
The first record of the file contains numeric records that should be treated as data.

151

Chapter 16 Glossary
Number of Header Records are Fixed
An input field beneath the list box appears when this choice is made, and this field
specifies the number of records at the top of the file to consider descriptive text
when reading the file; the first data record is the next record.
Number of Header Records Varies
The number of header records will be assumed to be all records prior to the first allnumeric record. If all records contain an alphanumeric Sample ID field then all records will be assumed to be header records and no data records will be read. In the
example below there are 2 header records:
File GMD Test Site
<m east> <m north> <Pb> <pH>
34.5
45.6
0.231 5.8
36.7
46.5
0.241 5.9.
Numeric Value as a Missing Value Indicator
The indicator value that will appear when a numeric value is specified in the list box.
The example below presumes that -99 has been specified the missing value indicator and thus the third value of this record will be read into the worksheet as a missing value:
13.2, 34.5, -99.0 , 0.15
Parameter File
Files that store analysis parameters and data for a particular data set. GS+ parameter files typically have a .par extension.
Polygons
Polygons are Irregular shapes that can be interpolated or excluded from interpolation during kriging. Polygons are defined by at least 3 vertices (coordinate pairs) that
define the polygon outline. Polygons can be as complex as continent outlines or as
simple as building outlines.
Posting
A Coordinate Posting is a map of coordinate locations. The location of each data
point in the active data set can be marked by a symbol corresponding to its relative
value.
Punctual Kriging
A type of kriging that provides an estimate for a precise point. Block kriging provides
an estimate for a discrete area around a specific interpolation point.
Quotes as Column Title Separators
Column or variate names are separated by quotes. E.G.:
m east m north Pb (ug/g) pH
152

Chapter 16 Glossary
Range
The range (also called effective range) is the separation distance over which sample
locations are autocorrelated, i.e. over which there is spatial dependence among
sample locations. In variogram models the range is calculated from the range parameter A0 (isotropic models) or A1 and A2 (anisotropic models). See model definitions for further information.
Same as for Data Records
The delimiter for column or variate titles is the same as specified for data records.
E.G. if the data records in a file with four data columns are formatted with "#" as a
delimiter:
m east # m north # Pb (ug/g) # pH
Sample ID
The Sample ID is text or a numeric value that identifies a particular data record. It is
optional. To specify a Sample ID column click on the top row of the data worksheet.
Sill or C0+C
The Sill of the variogram model (C0 + C) represents spatially-independent variance.
Data locations separated by a distance beyond which semivariance does not
change i.e. after the model asymptote or sill are spatially independent of one
another. Theoretically the sill is equivalent to sample variance.
Space Separated (free format) Values
Values within data records are separated by spaces, e.g.:
13.2 34.5 35.6 0.15
Spaces as Column Title Separators
Column or variate names are separated by spaces; note that this option limits column titles to single words. e.g.:
east

north Pb pH

Tab Separated Values

Values within data records are separated by tab characters (denoted by [tab] in the
example below; this format is often called tab-delimited output by spreadsheet programs, e.g.:
13.2 [tab] 34.5 [tab] 35.6 [tab] 0.15
Tabs as Column Title Separators
Column or variate names are separated by tab characters, e.g.:
m east [tab] m north [tab] Pb (ug/g) <tab> pH
Transformation
A transformation is a formula applied to all Z or Z2 values in a data set, ususally to
153

Chapter 16 Glossary
better normalize its distribution. The success of a transformation can be judged by
observing its frequency distribution before and after transformation. Values may be
backtransformed prior to reporting results.
Uniform Lag Interval Classes
The Lag Class Distance Interval defines how pairs of points will be grouped into lag
classes. Each point in a variogram or autocorrelogram represents the average
semivariance or Morans I for a single lag class, which is a group of pairs separated
by a certain Lag Class Distance Interval, sometimes called a step size. This interval
can either be calculated by GS+, in which case it will be uniformly distributed across
the active lag distance, or it can be manually set by the user
Variogram Model Parameters
Model parameters for isotropic variograms include terms for nugget variance (C0),
the combined nugget plus structural variance (C0 + C, also called the model sill), and
the range parameter A0 or in the case of anisotropic variograms, A (a function of A1
and A2). The range parameter may be different from the effective range, which
should be used to compare ranges among models.
What is a valid record
A valid record is any record that contains non-missing values for the coordinate locations (both an X-Coordinate value and a Y-Coordinate value for 2-dimensional domains) AND a non-missing value for the Z-Variate. Missing values appear as blank
cells OR as red colored text; use the right mouse button to turn cells with values into
cells with missing values and vice versa.
X,Y-Coordinates
X,Y coordinates describe a physical location at which a Z variate is measured. The
coordinates are presumed to be in Cartesian space, i.e. with a 0,0 origin that increases for x in an easterly direction and decrease in a westerly direction, and for y
increase in a northerly direction and decrease in a southerly (note that values can be
less than or greater than 0). Thus a value of 10,20 means 10 units east of the origin
and 20 units north.
In GS+, if a Y-Coordinate Column is not specified the data are assumed to be 1dimensional, i.e. collected along a transect or through time. To specify (or remove)
a Y-Coordinate column, click on the top row of the worksheet.
Note also that latitude and longitude are not part of a Cartesian coordinate system
because coordinate distances (measured in degrees) represent different physical
distances (measured in meters) in different places on the globe. For a rough conversion from latitude and longitude to meters, consider that one second of latitude
equals 30.92 meters on the ground; for longitude, calculate the cosine of the latitude
then multiply by 30.92.
154

Chapter 16 Glossary
Z-Variate Data (Z)
The Z-Variate is the variate being analyzed and mapped, e.g. elevation for a topographic map, pH for a map of soil acidity, chlorophyll content for a map of lake productivity, population density for a map of rural population growth, etc. A Worksheet
can contain many variates but only one can be analyzed at a time, thus only one
column can contain the Z-Variate data. A Z-Variate Column must be defined prior to
spatial analysis. To specify a column, click on the top row of the worksheet.
Z2 Variate Data (Z2 or Covariate)
The Z2 Variate or Covariate is a second Z-variate that covaries with the primary Zvariate. It is used in cokriging.

155

Chapter 16 Glossary

Index
1D transects, 145
2D Maps, 21, 134, 143
3D
Map proportion (height to width ratio),
21
Maps, 21, 134, 140
Standard deviations map, 141
A0, 68, 69, 71, 72, 73, 74, 153
A1, A2, 75, 76, 77, 78, 79, 154
Active lag distance, 60
Adjacency matrix, 93
Anisotropy
Axis orientation, 62, 80, 82
Semivariance surface, 16, 62, 63, 80,
82
Variogram map, 80, 82
Variogram models, 75
Appending data, 42
ArcView
Input file format, 35
Map input file format, 135
Output file format, 108, 130
Assign field, 28
Autocorrelation
Autocorrelation window, 60
Correlograms, 89
Covariance analysis, 90
definition, 148
Drift, 88
Fractal analysis, 94
General relative variograms, 91
Madograms, 86
Moran's I analysis, 93
Pairwise relative variograms, 92
Rodograms, 87
Semivariance, 59
156

Standardized variograms, 85
Autocorrelation menu, 9
Axis formats, 20
Azimuth, 81
Backtransformations, 44, 118, 148
Base input file, 25
Bibliography, 146
Binary data record format, 148
Block kriging, 149
Block size, 112
C0, 68, 149
Cartesian (x,y) coordinate system, 5, 154
Ceiling, 22, 137
Character Separated Values, 149
Cokriging, 29, 60, 123, 125
Column, 26
Field assignments, 25, 28
Titles, 36, 149, 152, 153
Widths, 27
Comma separated values, 41
Commas as column title separators, 41
Conditional simulation, 105, 119
Cokriging, 120
External drift, 120
Number of simulations, 122
Residuals, 120
Secondray data, 120, 121
Contour levels, 22, 137, 138
Convert GS+ DOS File, 24
Coordinate postings, 51
1d scattergrams, 54
2d scattergrams, 52
Definition, 152
Map data, 136, 144
Posting intervals, 55
Coordinates
Assigning, 28
Cartesian, 154

Index
Latitude, Longitude, 154
Correlograms, 89
Covariance analysis, 90
Covariate
Assignment, 28
Definition, 150, 155
Icons, 6
Regression, 56
Values warning, 29
Cross-semivariance analysis, 9, 83, 124
Cross-validation
Analysis, 108, 117
Values, 118
Cross-variate regression, 56, 57, 58
Cross-variograms, 9, 83, 124
Cumulative frequency distribution
Graph, 47
Values, 48
Cumulative probability distribution, 44
Data
Append dialog, 42
Bounding, 30
Builds, 150
Fields, 28, 150
File import, 33, 41
Filter, 25, 30
Input file format, 34, 36
Menu, 8
Records, 26, 36, 37, 38, 39, 41, 112,
150
Summary window, 51, 56
Regression analysis tab, 56
X,Y coordinates tab, 51
Z tab, 43
Title, 26
Worksheet, 25
Database input files, 33, 40
Decimal places, 13, 14, 27
Define posting intervals, 55
Descriptive statistics, 44
Discretization grid, 123, 150
Distance interval, 110

DOS GS files, 24
Drift, 88
Duplicate values, 31, 44
Edit menu, 7
Effective range, 68, 69, 76, 154
Email, 12
Excel files, 40
Exclusive polygons, 107, 114, 116
Exponential anisotropic model, 78
Exponential isotropic model, 72
Export data, 8
Field (column)
Assignments, 25, 26, 28
Default assignments, 14
Delimiter, 41
Names, 26
File
Extensions, 23, 24, 64, 107, 129, 130,
132, 135, 152
Import command, 25
Import dialog, 33
Import properties, 40, 41
Menu, 6
View window, 42
Filtering data, 25, 30
Floor, 22, 137
Formulas in worksheets, 27
Fractal analysis, 94
Frequency distribution
graph, 44
Graph, 45
Values, 46
Gaussian anisotropic model, 79
Gaussian isotropic model, 74
General relative variograms, 91
GeoEas input file format, 34, 37
Geometric anisotropy, 62
Geo-referenced data, 151
Graph settings
Axis scaling, 19
157

Index
Axis titles & labels, 20
Colors, 17
Contour details, 21
Footnote, 18
General, 17
Legend, 18
Title, 18
Grid lines, 22, 137
GS+ Input file format, 34
GS+ Interpolation Output file format, 129
Header records, 36, 37, 38, 39, 41, 112,
151, 152
Help menu, 11
Histogram bars, 20, 45
h-Scattergrams, 95, 100, 104
IDW, 127
Interpolation defaults, 16
Output files, 129, 130, 132
Smoothing factor, 16, 127, 128
Weighting power, 16, 127, 128
Importing
Default file type, 14
Default filename, 14
File import dialog, 33
Import file command, 25
Spreadsheets, 40
Text files, 34
Inclusive polygons, 107, 114, 116
Input file format, 34
ArcView XYZ format, 38
GeoEas format, 37
GS+ format, 36
Spreadsheet and database formats, 40
Surfer XYZ format, 39
Insert rows or columns, 27
Installation, 2
Interpolation, 105
Cokriging, 123, 125
Conditional simulation, 105, 119
Interpolate locations, 112
Interpolate menu, 10
158

Interpolation grid, 106

Interpolation masks. See Polygons
Interpolation range, 110
Interpolation type, 108
Interpolation window, 106
Inverse distance weighting, 105, 127
Kriging, 105
Nonuniform grid, 112
Normal distance weighting, 105, 127
Output formats, 107
What is an interpolate?, 151
Z estimate boundaries, 109
Interval class, 59
Inverse distance weighting, 105, 108, 127
Irregular shapes, 107
Isotropic variogram models, 68
Kriging, 105, 123
Block vs. Point, 123
Command, 108
Krig output files, 129, 130, 132
Output file name, 107
Lag class distance interval, 59, 60, 61, 64,
151
Latitude, 154
Legend title, 55
Licensing and copy protection, 3
Linear anisotropic model, 78
Linear isotropic model, 73
Local grid, 123
Longitude, 154
Madograms, 86
Main menu, 6
Major axis, 75, 81
Map
2D image, 143
2d sample postings, 144
3D image, 140
3d rotation, 142
3D standard deviation image, 141
ArcView input file format, 130

Index

ArcView input file format, 135

Ceiling, 22, 137
Contour levels, 138
Floor, 22, 137
Grid (range), 135
Grid lines, 137
GS+ input file format, 129, 135
Input file format, 129
Input file formats, 134
Legend, 22, 137
Mapping in GS+, 134
Menu, 10
Surface properties, 21
Surfer input file format, 132, 135
Masks, 107
Minor axis, 75
Missing values
Count, 135
Definition, 151
Missing value indicator, 13, 29, 34, 36,
38, 39, 40, 41, 129, 135, 138
Permanent missing values, 27, 29
Preferences, 13
Temporarily missing values, 29, 31
Model parameters, 69, 76
Model results, 66
Morans I analysis, 93
Mouse action, 81, 140
NDW, 16, 127
Nonuniform interpolation grid, 107, 112
Normal distance weighting (NDW), 16,
127
Normal probability distribution, 44
Graph, 49
Values, 50
Nugget variance, 68, 69, 76, 149
Offset tolerance, 62
Ouput file formats, 107
Outliers, 95
Outline map, 116
Output file format

ArcView format, 130

GS+ format, 129
Surfer format, 132
Pairwise relative variograms, 92
Parameter files, 6, 152
Polygons
Defined, 152
How to create, 107
Polygon definition window, 114
Polygon map, 116
Polygon vertices, 114
Posting intervals, 55
Postings. See Coordinate postings
Preferences
Analyses, 15
Data file import, 14
Decimal places, 13
Draw semivariance surface, 16
General, 13
IDW smoothing factor, 16
IDW vs. NDW, 16
IDW weighting power, 16
Import field assignments, 14
Import file type, 14
Import filename, 14
Missing value indicator, 13
Variogram defaults, 15
Variogram offset, 15
Primary variate, 125
Principal anisotropic axis, 81
Printer setup, 7
Printing graphs, 23
Printing text, 24
Punctual kriging, 152
Quantile plots. See Coordinate postings
Quotes as column title separators, 41
Range, 68, 69, 75, 76, 153, 154
Rebuild command, 26
Records. See Data records
Regression
159

Index
Coefficient, 117
Regression analysis, 56, 57
Regression coefficient, 70, 76
Regression values, 58
Residual sums of squares, 69, 76
Rodograms, 87
Rotate graph, 140, 142
Sample ID, 28, 153
Sample postings. See Coordinate postings
Sample variance, 63
Scaling data, 43
Scattergrams. See Coordinate postings
Screen layout, 5
Search neighborhood, 108
Secondary variate (Z2), 125
Semivariance
Overview, 59
Semivariance analysis window, 60
Semivariance offset, 59
Values, 67
Variograms window, 66
Sill, 68, 69, 75, 76, 153
Single-user license agreement, 3
Site licensing, 3
Smoothing factor, 16, 127, 128
Space separated (free format) values, 41
Spaces as column title separators, 41
Spherical anisotropic model, 77
Spherical isotropic model, 71
Spreadsheet input files, 33, 40, 41
Standard deviations map, 141
Standardized variogram, 85
Structural variance, 68, 149, 153
Summary statistics, 43
Surface, 136
Surfer
Grid output file format, 107
Map input file format, 39, 132, 135
XYZ input file format, 35
Symbols, 20, 55
Tab separated values, 41
160

Tabs as column title separators, 41

Text files, 33, 34
Tick marks, 19
Tool bar, 6
Transect image, 145
Transformations
Assigning, 43
Backtransformations, 44, 118, 148
Defined, 153
Lognormal, 43
Offset, 44
Scaling 0-1, 43
Square root, 43
Uniform interpolation grid, 106, 110
Uniform lag interval classes, 154
Updates, 2, 12
Variance
Variance cloud analysis, 66
Variance cloud pairs, 99
Variate assignments, 26
Variate names, 26, 36, 149, 152, 153
Variogram
Anisotropic models, 75
Defaults, 15
Isotropic models, 68, 69
Map, 62, 80, 82
Model command, 63
Model parameters, 69, 154
Model type for kriging, 124
Nugget variance, 69, 75, 76
Offset, 15
Options, 63
Range, 69, 75, 76
Sill, 68, 69, 75, 76, 153
Structural variance, 153
Variograms window, 66
Vertices, 114
Warranty, 4
Weighting power, 16, 127, 128
Window menu, 11

Index
Worksheet
Formulas, 27
Limits, 27
Window, 25
X axis range, 19
X-coordinate
Column assignment, 28
Defintion, 154
Y axis range, 19
Y-coordinate
Column assignment, 28
Defintion, 154

Z (primary variate)
Column assignment, 28
Definition, 155
Z standard deviations, 136
Z axis range, 20
Z Estimate Boundaries, 109
Z2 (covariate)
Cokriging, 123
Column assignment, 28
Definition, 155
Values warning, 29
Zoom, 140

161