GRAIN SIZE ANALYSIS

(Original Article; Accepted Final Proof; Allowed for upload as self-repository - not the original Springer typeset - )
----------------------------------------------------------
GRAIN SIZE ANALYSIS
----------------------------------------------------------
pp. 341-348
Encyclopedia of Earth Science Series
Encyclopedia of Geoarchaeology
Allan S. Gilbert (Volume Ed.)
Springer
2017

Print publication under ISBN: 978-94-007-4827-9

Electronic publication under ISBN 978-1-4020-4409-0
Print and electronic bundle under ISBN 978-94-007-4828-6

Gloria I. López
Luminescence Dating Laboratory, Centro Nacional de Investigación sobre la Evolución Humana CENIEH,
Burgos, 09002, Spain
Leon Recanati Institute for Maritime Studies RIMS, University of Haifa, Haifa, Israel

Synonyms
Granulometry; Particle size analysis

Definition
Grain size analysis is an analytical technique typically conducted within the earth sciences and
implemented as a routine laboratory study. Other disciplines, such as archaeology and geoarchaeology,
also use it regularly. It is a sedimentological analysis carried out in order to determine the size of the
different particles that constitute a particular unconsolidated sedimentary deposit, sedimentary rock,
archaeological locus, or soil unit. The main goal of this procedure is to determine the type of
environment and energy associated with the transport mechanism at the time of deposition; this is done by
inference from the sizes of the sediment particles analyzed and their distributions.

Introduction
Granulometry is a basic analytical technique that has wide applications within the earth and
archaeological sciences. Particle or grain size is a fundamental attribute or physical property of particulate
samples or sediments and sedimentary rocks (Folk, 1980; Friedman and Sanders, 1978). Much can be
said from analyzing not only the size of clastic or detrital (inorganic), bioclastic (organic), or chemical
particles but also from the overall size distribution, size fraction percentages, textural maturity of the
sediment or sorting, surface texture attributes of a particle, and sphericity/angularity and shape of a
particle (Krumbein and Sloss, 1963; Syvitsky, 2007). Several sediment, soil, or material properties are
directly influenced by the size of its particles, as well as their shape (form, roundness and surface texture
or the grains) and fabric (grain-to-grain interrelation and grain orientation), such as texture and
appearance, density, porosity, and permeability.

The size of particles is directly dependent on the type of environmental setting, transporting agent, length
and time during transport, and depositional conditions, and hence it possesses significant utility as an
environmental proxy (McManus, 1988; Stanley-Wood and Lines, 1992). Grain size is related to a
multitude of external factors acting on a local or regional scale. For example, in the coastal and marine
setting, grain size is related to the bathymetry and geometry of the basin, nutrient regime, biogeochemical 1
oceanography, coastal processes, net sedimentary inputs from land sources, and outputs. The study of

pp. 341-348 | Encyclopedia of Geoarchaeology, DOI 10.1007/978-1-4020-4409-0, 2017

 GRAIN SIZE ANALYSIS

these particles can elucidate their provenance (source materials), the various processes they may have
endured during their transport (by air, land, or water), their final depositional environment, and final
burial setting (how much energy was present at that time; e.g., from waves or currents), and other
physical and chemical factors.

Traditionally, sediments were divided into three principal categories: gravel, sand, and mud. The latter is
further divided into silt and clay, mostly based on mineralogical distinction rather than (hydro-)dynamic
properties. Since the early 1900s, standardization of such size ranges has been defined (Figure 1) based
upon different grade scales constrained by particle size limits or range boundaries. The size of the
particles is based on their nominal diameter, traditionally reported in millimeters (mm), micrometers (µm)
or phi (Ø) units. The Wentworth or Udden-Wentworth scale (Udden, 1914; Wentworth, 1922) divides the
size ranges into textural classes with specific terminology, from boulders (> 200 mm) to clay (< 0.004
mm). It is a geometric scale in which each size limit is 1/2 or twice the millimetre value of the next
(Figure 1). The Krumbein Ø scale (Krumbein and Sloss, 1963) is a logarithmic scale modified from the
Udden-Wentworth one and based on conveniently calculated round values, to avoid dealing with mm
fractions (Figure 1). Classification of detrital sediments is based upon the quantification of, or relationship
between, the proportions/percentages/ratios of different particle size fractions or textural classes within a
mixed sediment (Shepard, 1954; Folk, 1980), as seen in Figure 2.

Grain size analysis is often part of the basic, initial set of analytical laboratory procedures scientists
conduct upon sediment/soil samples and/or sediment cores are recently collected in the field. The purpose
of such analysis is to (1) obtain a deeper understanding of paleo-environmental features or modern
environmental impacts, (2) reconstruct past sedimentary transport histories, depositional conditions, or
sediment provenance, or (3) analyze in detail a catastrophic event, such as a tsunami or hurricane deposit,
for example.

Depending on the thickness and regional extent of the sedimentary layer, unit, deposit, or archaeological
locus, scientists collect one or multiple samples within that unit in order to obtain representative material
and, with sufficient sampling, a more accurate statistical result. The resulting grain size analyses of
samples collected from a single unit are averaged in order to obtain the overall particle size distribution.

Several analytical methodologies can be applied in the study of granulometry and the distribution of
particle sizes within a sample or specific material (Table 1). These vary in terms of applicability,
technique, apparatus, and cost (Table 2). Nonetheless, the method selected would depend on the range of
particle size, the degree of consolidation of the material and the purpose of the analysis. Traditional
analyses include counting (individual clasts, manually), sieving, and settling, for gravelly, sandy and
muddy materials respectively. Modern analyses (Table 1) have used the same analytical principles (count,
sieve and settle), however, the methodologies are much improved today due to better instrumentation and
automatization (Table 2), including laser diffraction and imaging techniques (e.g., spectrometry), without
discarding traditional sieving and settling (hydrometer and pipette analysis) methodologies (Syvitski,
2007).

Contextual granulometry

Granulometry in geoarchaeology
Granulometry can be considered a perfect example of geoarchaeological research in which the application
of an earth sciences technique is used to understand aspects of the archaeological record. For example, in
order to obtain a better understanding of the composition of a locus or stratigraphic unit, grain size
2
analyses would be a basic parameter to estimate in order to differentiate the matrix from the
clasts/aggregates (geological terms) or inclusions (archaeological term), within the same locus/unit,

pp. 341-348 | Encyclopedia of Geoarchaeology, DOI 10.1007/978-1-4020-4409-0, 2017

 GRAIN SIZE ANALYSIS

and/or across loci/units. This information enables a more precise understanding of particles that might be
related to natural deposition versus those deriving from an anthropogenic origin – as some anthropogenic
deposits have granulometric characteristics that cannot be compared to any natural pattern since they
derive exclusively from artificial behaviors. An example is the differentiation between naturally-derived
and anthropogenically-produced submerged sedimentary units found within the same archaeologically-
rich coastal region of Caesarea Maritima, Israel (Reinhardt et al., 2006). The detailed analysis of
Glycymeris spp. (saltwater clam) bio-clast distributions found at different stratigraphic intervals indicated
two distinct naturally marine-derived tsunami phases; these were compared to an anthropogenically
induced and/or mixed ballast and pottery layer amid naturally occurring medium sand-sized siliciclastic
sedimentation related to normal marine and winter storm conditions. However, subsequent
reinterpretations by some of the authors, based primarily on particle size distributions (diameters < 2 mm)
and several sedimentary textural and structural features, now suggest a tsunami origin for the previously
identified anthropogenic layer (Morhange et al., 2014). This is an example of the complexity in weighting
one analytical proxy more others, or the difficulty of identifying and characterizing natural from
anthropogenic units in complex environmental and reworked archaeological settings.

Paleoenvironmental reconstruction
One of the goals of geoarchaeological research is to understand previous natural and anthropogenic events
that took place within an archaeological context, either happening in sequence or ongoing intermittently.
Based on such information, the environmental history of a site can be reconstructed from its beginnings.
As a primary analytical technique, particle size analysis should always be accompanied by other basic
analyses that have the potential to enhance comprehension of a targeted locus or sedimentary unit within
the archaeological context. Examples of such other proxies are micropaleontology, and isotope and
chemical analyses. The combined use of several analytical laboratory procedures increases the likelihood
of obtaining data useful in elucidating the depositional context as a whole, and with decreasing margin of
error. Underwater archaeological sites, and especially those situated along coastal areas, are among the
more complex settings within which to reconstruct the stratigraphic succession of ancient environments.
This is due not only to the submerged conditions, but also to the highly dynamic processes occurring
therein, including sea-level changes, severe storms, extreme events such as tsunamis, littoral currents, and
sediment movement, etc. In conjunction with these, prolonged and heavy human occupation of the
(ancient) coastline adds to the complexity of the setting. Nonetheless, well-documented reconstructions of
the lateral and/or vertical progression of environments due to continuous environmental evolution, abrupt
natural changes, and man-made constructions can be seen throughout ancient coastlines. Examples from
the Mediterranean include the reconstruction of the now submerged ancient city and port of Alexandria
on the Nile Delta, Egypt (Mostafa et al., 2000), the reconstruction of the harbor complex of Caesarea
Maritima, Israel – the largest constructed artificial harbor in the Mediterranean (Reinhardt et al., 2006),
and the understanding of the destruction of ancient Palaikastro in Crete following the great eruption of
Santorini in the Late Bronze Age (Bruins et al., 2008).

Analytical procedures and particle sizing techniques

One of the particularities of grain size analysis is the importance of estimating correctly the different size
fractions that constitute the ensemble of the material being analyzed. In theory, each particle constituting
the sample is to be analyzed individually, and the final result is the combination of all the individual
measurements. Particles are complex three-dimensional objects with specific lengths, widths, and
thicknesses, however (Folk, 1980; Stanley-Wood and Lines, 1992). Only perfect spheres can be
completely described by a single number, i.e., their radius or diameter. So, in order to simplify the
measurement process, particles are most commonly conceptualized as one-dimensional spheres rather
than three-dimensional objects with irregular shapes. A cylinder with X length and Y diameter can be
3
equivalent to a sphere of Z diameter that has the same volume as the cylinder. This is called the Concept
of Equivalent Spheres (Jennings and Parslow, 1988). This concept can be applied to a number of different

pp. 341-348 | Encyclopedia of Geoarchaeology, DOI 10.1007/978-1-4020-4409-0, 2017

 GRAIN SIZE ANALYSIS

measuring properties of a particle, such as maximum and minimum diameter, surface area, volume, or
weight. Based on this concept of equivalent spheres, sedimentation rates can be calculated, and simple
things such as sieve aperture sizes can be specified (Folk, 1980; McManus, 1988).

If a sample is measured using different technical means, the results of each technique may not be
equivalent because different apparatuses measure different parameters of the equivalent spheres.
Therefore, consistency and comparability must be maintained. Grain size analyses can be performed by
various means, in both dry and wet settings, depending on the type of material and its major constituent
fraction. The coarse fraction of the sediment (> 0.063 mm) is commonly separated through dry sieving,
whereas the finer sediment can be isolated by settling or sedimentation (using a hydrometer), or using
laser diffraction (see Tables 1 and 2). Preparation of the sample is also a crucial point in the
determination of a credible and accurate grain size measurement, and hence the importance of the
objective of the study. If the target material is clastic, removal of other allochthonous materials is
compulsory (e.g. organics, carbonates, oxides, salts, etc.) to avoid erroneous measurements and vice versa
(Figure 3).

Sieving analysis
Sieving is the most basic of the particle sizing techniques. It consists of having the sediment pass through
(by agitation) a series of stacked sieve meshes with defined opening sizes. Each sieve catches the size
fraction that is larger than its mesh size, so that the successive sieves break up the sample into decreasing
size fractions. The sediment fraction retained in each sieve is weighed in order to obtain its percentage
relative to the whole sample. This technique can be used under dry or wet conditions.

The advantages of sieving are that it is cheap and user friendly, useful when dealing with very coarse
samples and the physical separation of the sample is the end result. Its limitations are its low resolution
and precision, that dry particles smaller than 50 µm or cohesive materials are very difficult to separate
using this technique, and that results are influenced by the operator and the duration of agitation/shaking
used, i.e., the technique itself (Folk, 1980; Krumbein and Sloss, 1963).

Sedimentation or settling
Sedimentation is the oldest of the techniques used in particle size analysis. It measures the rate of
sedimentation of particles suspended in a liquid. Its advantages include its relatively low cost, and its ease
of applicability to soils or very fine sediments (for which it is the traditional method). Its limitations are
that it is useful only for a limited range of particle sizes, that it is not useful for sediment < 5 µm, and that
it is extremely sensitive to particle shape (geometry) (Jennings and Parslow, 1988; Stanley-Wood and
Lines, 1992).

Laser diffraction
Laser diffraction measures the angular dependence of laser light scattered by an ensemble of particles. Its
advantages are that it can handle a very wide range of particle sizes (from < 100 nm to ~ 2–3 mm), that
measurements can be made rapidly and thus large numbers of samples can be processed, and that results
are accurate and repeatable (Syvitski, 2007).

Laser diffraction measurements provide particle size distributions with great detail. This enhancement in
technical size measurement has greatly improved the ability to differentiate and compare different
environments, and sometimes even better understand their dynamics.
When using a laser diffraction particle size analyzer, sediment can be run dry or wet. If wet, however, it is
advised to pour out as much water as possible from the container to minimize errors. In either case,
homogenizing and dispersing the sample prior to insertion into the machine is always a must, in order to
4
analyze a truly representative portion of the sample.

pp. 341-348 | Encyclopedia of Geoarchaeology, DOI 10.1007/978-1-4020-4409-0, 2017

 GRAIN SIZE ANALYSIS

The limitations of laser diffraction include that it is not suitable for very coarse or nano-materials. It is a
medium resolution technique and is applied to the whole (ensemble) sample.

Dynamic light scattering

Dynamic light scattering measures scattered light intensity variations due to Brownian motion of particles
in suspension within a liquid. Its advantages are that its dynamic range is well suited to nano-materials (<
1 nm to 1 µm), its measurement speed is rapid so that it can handle larger numbers of samples, and its
results are accurate and repeatable. Its limitations include the inability to analyze dense materials, and its
medium resolution (Syvitski, 2007).

Particle size distributions

In nature, sediments do not consist of only one kind of particle, but rather an amalgamation of various
particle sizes, hence it is logical to consider grain size as a continuous variable. Only perfectly mono-
dispersed samples possess particles of exactly the same size, for example, highly sorted sand winnowed
through hydrodynamic processes. Most natural samples contain a range, or distribution, of different
particle sizes and shapes.

As a result of manual or instrumental measurement of grain size, a size frequency spectrum is obtained.
Such spectrum is determined by the count of grains, weight or volume percent of a particular particle
fraction within a specific size interval. The resulting spectrum can be deployed as a distribution of grain
sizes based on the relative frequency of their number per size fraction (Figure 4). The most common
graphic depiction used to represent the different grain sizes within a sample consists of a statistical
distribution encompassing all resulting fractions based on their relative frequency (quantity or
volume).Two common representations are used routinely: relative frequency and cumulative distribution
curves (e.g., Figure 4). The different shapes of the relative frequency distribution curve can be interpreted
as how well sorted the sample is: a narrow size range or narrow Gaussian shaped curve implies a well-
sorted sample, whereas a larger size range or ample Gaussian shaped curve implies a poorly sorted
sample (i.e., a wide range of particle sizes). Moreover, the more asymmetric the curve is within a single
distribution, or if the latter presents several frequency peaks (as seen in Figure 4), the greater modality of
the sample: poly-modality is shown by the greater frequency of distinct and different grain size range
peaks. These different frequency groups are called populations and imply an ample modality of grain size
fractions within the same sample, and vice versa.

Particle size distributions are based upon statistics, and description of the statistical parameters will
usually depend on how the data are to be used. These calculated statistical parameters may give insight
into various aspects of the environmental, depositional and transport conditions the sediment grains
endured, linking them to particular sedimentary systems. Three common parameters are the mean, the
median, and the mode (Figure 4). The mean is the average size of the entire sample, as seen in Figure 4.
The median is the diameter where 50% of the particles are below or above that threshold. It is by far the
easiest measure to determine but the least useful as it does not reflect the extremes of the curve (Folk,
1980). The mode is the particle size with the highest frequency, as seen in Figure 4, but it is not a good
proxy of the overall sediment mixture (Folk, 1980). The only instance where all of these three parameters
coincide is when the frequency distribution curve is a perfectly symmetrical Gaussian curve.

Other important statistical parameters obtained from the analysis of the distribution of particles – which
can help elucidate how uniform, symmetric or well sorted the sediment sample is – are the standard
deviation, skewness and kurtosis (Folk, 1980). The standard deviation is a precise measure of the scatter
of grain size values from the mean, corresponding then to a measure of spread or sorting of the sample. In
combination with the mean, the standard deviation is the most useful and widely applied value in
5
granulometric statistics. Three limits are useful when computing standard deviations within a single
sample: 1 standard deviation (±σ) from the mean implies that 68% of the grain size values fall within this

pp. 341-348 | Encyclopedia of Geoarchaeology, DOI 10.1007/978-1-4020-4409-0, 2017

 GRAIN SIZE ANALYSIS

limit; 2 standard deviations (±2σ) corresponds to 95% of the particles; and 3 standard deviations (±3σ) to
99% (Folk, 1980).

The skewness is used to establish the normality or symmetry of the distribution, hence to quantify the
degree of dispersion within a sample, rather than only visualizing it on a frequency histogram. The closer
the skewness value is to zero, the more symmetrical (i.e., normal or uni-modal) the distribution is.
Asymmetrical and multi-modal sediment mixtures exhibit high values of skewness, to maximums of
+1.00 and -1.00. The positive and negative sign of the skewness value indicates whether the asymmetrical
tail extends to the left or right of the curve as follows (Folk, 1980). Distribution curves highly skewed to
low grain size values show a negative value and are diagnostic of environments with higher
concentrations of silts and clays. The opposite are environments with higher concentrations of coarser
materials which show curves skewed to higher grain sizes, hence positive values. Extremely turbid
systems such as grain or turbidity flows are diagnostic of negatively skewed distributions. Tsunami,
colluvial, debris flows and torrential river deposits are diagnostic of positively skewed and multi-modal
distributions.

The kurtosis is also a quantitative measure to describe the degree of Gaussian normality of the grain size
distribution, but in terms of how acute or flat the curve is. This is a sorting relation between the end
members of the curve and its center (Folk, 1980). If the central portion of the curve is peaked, hence
better sorted than its tails, the distribution curve is said to be leptokurtic with values > 1.00. The opposite,
a flat-peaked curve with a large spread of grain size in the centre, is called platikurtic, with values < 1.00.
Normal probability curves have a kurtosis of 1.0 (Folk, 1980). Both kurtosis and skewness values are
ratios of dispersion; thus, they are dimensionless and do not have units.

Summary
Grain size analysis is a fundamental tool for classifying unconsolidated materials and sediments,
sedimentary rocks, and sedimentary environments. Quantitative analysis of the percentages of different
particulate sizes yields one of the most fundamental physical properties of clastic sediments and
sedimentary rocks. Grade scales, such as the geometric Udden-Wentworth or the logarithmic Krumbein Ø
scales, which correspond to grain size intervals with a regular relationship to one another, were created to
maintain a standardized statistical estimate of the measurement of the size of a particle, because grain size
is considered a continuous variable. The almost exclusive purpose behind sizing grains is to obtain a
frequency distribution of particle sizes.

A variety of principles and specific methodologies can be applied to differentiate and characterize the
particle sizes of unconsolidated materials and sediments or sedimentary rocks. The traditional analytical
principles behind granulometry (counting, sieving, and settling) are still frequently used today, and in
some instances, they cannot be superseded by modern techniques because certain particle ranges lie
beyond the measureable limits of sensitive minuscule sensors. For this reason, when dealing with particle
size analyses, it is also important to consider the particle size range of the material to be analyzed, how
narrow or wide this might be, in order to select the most appropriate measurement method(s). Gravels and
boulders are mostly counted manually. Pebble sizes can be determined by sieving. Sands, silts, and clays
can be measured either by sieving (wet and dry conditions) or settling (hydrometer, pipette). However,
automated modern techniques such as laser diffraction or dynamic light scattering can make the
measurement process much faster and accurate than traditional techniques, and a greater number of
samples can be analyzed at a time.

Selection of the particle size technique to use is dependent not only on the precision and accuracy
6
required for each sample measurement, but also the skills of the operator and the time that can be spent
per measurement. Another important factor to consider is the cost of the equipment and any consumables

pp. 341-348 | Encyclopedia of Geoarchaeology, DOI 10.1007/978-1-4020-4409-0, 2017

 GRAIN SIZE ANALYSIS

associated with that particular technique. The choice of method is basically dictated by the objective of
the study and the degree of consolidation of the material. This is the reason why it is imperative to have
an understanding of the complexity of the environmental system and/or archaeological site targeted for
study, as well as the natural processes that may have affected the locale, both in past and present times, in
order to have a better analytical appraisal of the environmental conditions and contributions of events to
the resulting stratigraphical signatures.

Bibliography
Bruins, H. J., MacGillivray, J. A., Synolakis, C. E., Benjamini, C., Keller, J., Kisch, H. J., and Klügel, A.,
2008. Geoarchaeological tsunami deposits at Palaikastro (Crete) and the Late Minoan IA eruption
of Santorini. Journal of Archaeological Science, 35(1): 191-212.
Folk, R. L., 1980. Petrology of Sedimentary Rocks. Austin: Hemphill Publishing.
Friedman, G.M., and Sanders, J.E., 1978. Principles of Sedimentology. New York: Wiley.
Jennings, B.R., and Parslow, K., 1988. Particle Size Measurement: The Equivalent Spherical Diameter.
Proceedings of the Royal Society of London Series A, 419(1856): 137-149.
Krumbein, W.C., and Sloss, L.L., 1963. Stratigraphy and Sedimentation, 2nd edition. San Francisco: W.
H. Freeman.
McManus, J., 1988. Grain size determination and interpretation. In Tucker, M. E. (ed.), Techniques in
Sedimentology. Oxford: Blackwell Scientific, pp. 63–85.
Morhange, C., Salamon, A, Bony, G., Flaux, C., Galili, E., Goiran, J.-P., and Zviely, D., 2014.
Geoarchaeology of tsunamis and the revival of neo-catastrophism. In Nigro, L. (ed.), Overcoming
Catastrophes: Essays on disastrous agents characterization and resilience strategies in pre-
classical Southern Levant. Rome: “La Sapienza” Studies on the Archaeology of Palestine and
Transjordan (ROSAPAT), Vol. 11, pp. 31-51.
Mostafa, M. H., Grimal, N., and Nakashima, D. (eds.), 2000. Underwater Archaeology and Coastal
Management: Focus on Alexandria. Coastal Management Sourcebooks 2. Paris: UNESCO.
Reinhardt, E.G., Goodman, B.N., Boyce, J.I., López, G.I., van Hengstum, P., Rink, W.J., Mart, Y., and
Raban, A., 2006. The tsunami of 13 December A.D. 115 and the destruction of Herod the Great's
harbor at Caesarea Maritima, Israel. Geology, 34(12): 1061-1064.
Shepard, F.P., 1954. Nomenclature based on sand-silt-clay ratios. Journal Sedimentary Petrology, 24(3):
151-158.
Stanley-Wood, N., and Lines, R.W., 1992. Particle Size Analysis. Cambridge: Royal Society of
Chemistry. Royal Society of Chemistry, Special Publication 102.
Syvitski, J.P.M., 2007. Principles, Methods and Application of Particle Size Analysis. Cambridge:
Cambridge University Press.
Udden, J.A., 1914. Mechanical composition of clastic sediments. Bulletin of the Geological Society of
America, 25: 655-744.
Wentworth, C.K., 1922. A scale of grade and class terms of clastic sediments. Journal of Geology, 30(5):
377-392.

Cross-references
Aeolian Settings: Loess
Aeolian Settings: Sand
Alluvial Settings
Archaeological Stratigraphy
Coastal Settings
Colluvial Settings
Lacustrine Settings 7
Mass Movement
Paleoenvironmental Reconstruction

pp. 341-348 | Encyclopedia of Geoarchaeology, DOI 10.1007/978-1-4020-4409-0, 2017

 GRAIN SIZE ANALYSIS

Sedimentology
Soils
Tsunamis
Underwater Settings

Tables & Figures

pp. 341-348 | Encyclopedia of Geoarchaeology, DOI 10.1007/978-1-4020-4409-0, 2017

 GRAIN SIZE ANALYSIS

pp. 341-348 | Encyclopedia of Geoarchaeology, DOI 10.1007/978-1-4020-4409-0, 2017