Sabine Chabrillat 1
Alexander F.H. Goetz 1,2
Harold W. Olsen 3
Lisa Krosley 3
David C. Noe 4

1 CSES/CIRES
2 Department of Geological Sciences
3 Colorado School of Mines
4 Colorado Geological Survey

Abstract
1.0 Introduction
2.0 Laboratory Analyses
2.1 Sampling in the Colorado Front Range Urban Corridor
2.2 Mineralogical Composition and Swelling Potential
2.3 SPECTRA: Interpretation and Correlation with Mineralogy and Swelling Potential
3.0 AVIRIS Imagery
3.1 Data Acquired and Calibration to Reflectance
3.2 Image Processing
3.3 Results and Interpretation
4.0 Conclusion
5.0 Acknowledgments
6.0 References
Figures

Swelling soils are a major geologic hazard, and expansive clays and clay-shales cause extensive damage world-wide every year. Current engineering and geologic practice for characterization of expansive clays involves time-consuming and expensive standard engineering tests for determination of swelling potential, and x-ray diffraction (XRD) analyses for mineralogical identification. Reflectance spectrometry is proposed as an alternate identification technique. Field spectrometers are used to relate spectroscopic features with mineralogical variations and degree of swelling potential. AVIRIS hyperspectral measurements are used to develop a map of swelling soil potential.

Near-infrared (NIR) reflectance spectroscopy of swelling soil field samples shows that it is possible to discriminate among pure smectite and mixed smectite/illite layer samples. The more smectite-rich the sample, the higher the swelling potential. The analysis of AVIRIS data acquired over Colorado Springs shows that exposures of expansive clays can be detected among the other components in the images. A map of expansive clay outcrops was produced and among those outcrops, spectral discrimination and identification of variable clay mineralogy (kaolinite, smectite/illite, smectite), related to variable swelling potential, is possible.

One of the worst swelling soils damage regions in the United States occurs along the 300 km long Front Range Urban Corridor in Colorado which is underlain by Cretaceous clay-shales, including the Pierre Shale. The sedimentary bedrock strata are generally flat-lying, except near the foothills of the Rocky Mountains where they have been uplifted into steeply-dipping strata. Expansive clays in the Pierre Shale and adjacent formations along the Front Range are responsible for the damage. The hazard is most severe in areas where these units dip steeply because of differential movement of adjacent beds (cm-m) which has been attributed to the abundance and composition of swelling clays (Gill et al., 1996). The three most important groups of clay minerals are smectite, illite and kaolinite. Smectite (including montmorillonite, the best-known member of the smectite group) has the greatest swelling potential and is responsible for most swelling soil damage in Colorado. Different types of clays can be identified spectroscopically thanks to their characteristic absorption bands around 2.2 mm. More detailed background related to this study and preliminary results can be found in Chabrillat et al., 1997.

Here we present additional laboratory analyses relating spectral signatures of expansive clay samples with their mineralogical composition from XRD data and their swelling potential from a refined procedure of standard engineering tests. Hyperspectral AVIRIS (Airborne Visible Infrared Imaging Spectrometer) images were acquired in 1997 in the Southern Colorado Urban Corridor area. The results obtained from the laboratory analyses are used to interpret results from AVIRIS mapping in terms of swelling potential.

Good exposures of expansive clays along the Front Range Urban Corridor are limited in size and sparse. Most of them are covered with grass, especially in the northern and Denver areas. South of Colorado Springs and around Pueblo and Canon City, exposures are more common and covered with desert-type vegetation. A total of 104 samples were collected at 30-50 cm depth from different locations shown in figure 1 on Hart’s (1974) generalized geologic map. They were all taken from Upper Cretaceous Formations, Pierre, Niobrara, and Benton (younger to older). The great majority of samples obtained were clay-shales, constituting predominantly of mixed-layer illite/smectite with frequent thin bentonite beds (almost pure smectite, 10-30 cm thick), and are classified as having moderate to very high swell potential. On each sample three types of analyses were performed: (a) geotechnical index tests including the Atterberg limits, grain size analysis, and one-dimensional swell test to determine the swelling potential; (b) XRD patterns on oriented samples to determine the mineralogical composition; and (c) reflectance spectra that were acquired in the VNIR-SWIR region.

Figures 2a and 2b show the mineralogy and swell potential for two suites of samples respectively from Perry Park and Canon City. These two suites of samples were chosen for their variability and encompass the range that we have found along the Front Range.

Quantitative mineralogy was derived from XRD patterns of oriented preparations using a millipore filter peel transfer onto glass slides (Drever, 1973). The calculations were carried out following the equations of Schultz (1964). When possible, mixed-layer illite/smectite (I/S) expandabilities were determined by the ° 2 q method of Moore and Reynolds (1989). When peak interferences from other phases made this determination difficult, ordering of the sample allowed an educated guess of the I/S expandability. In this case we calculated that the incertitude on the expandability value lead to an error in the percent of total smectite (smectite phase + smectite in the I/S layers) and total illite (illite phase + illite in I/S) within ± 10%. The mineralogical composition determined is expressed in clay fraction, i.e. reported to 100% clays. The samples range from pure smectite, to various degrees of mixed-layers smectite/illite and smectite content, to very low smectite content (20%). Illite content can be up to ~60%. Kaolinite content is usually less than 10% in these shale formations, but can go up to ~50%. In general, the Perry Park samples are very rich in smectite. The Canon City samples are unusually rich in kaolinite. We noticed two samples in Canon City (CC97011 and CC97013) that were composed of pure smectite phase with no illite but with a non-negligible content of kaolinite. The equations of Schultz that assume all phases are present do not produce credible results; the kaolinite content in this case is somewhat undetermined and could be up to 10% more than indicated.

The swell tests were performed initially following the ASTM (American Society for Testing and Materials) standard test method for measurements of soil potential (suction) using filter paper. As we found that the standard procedure could lead to an error in the estimation of the paper moisture content, we used a refined method to obtain a more accurate paper moisture content. The suction potential shown in figure 2 is obtained for each sample using the equation dh/dw = (6.25-h)/(0-w), where h is the total suction in pF (log 10 of the pressure in cm of water), w is the soil moisture content in %, and 6.25 (in pF) is McKeen’s (1992) intercept value at zero moisture content. The value dh/dw represents the slope in the diagram suction = f(water content), and this slope is the soil’s suction potential. The shallow negative slopes represent high suction potential and steep negative slopes are low suction potential. McKeen proposed an expansive Soil Classification System based on dh/dw. For this paper, the swell potential categories utilize only the dh/dw aspect but preserve McKeen’s boundary values as noted. Category I (dh/dw> -6) is for very high expansive potential, special cases, e.g. almost pure montmorillonite. Category II (-6 <dh/dw< -10) is for high expansive potential. Category III (-10 <dh/dw< -13) comprises moderate expansive potential, category IV (-13 <dh/dw< -20) low expansive potential. Category V soils (dh/dw< -20) are non-expansive. There is a wide range of expansive potential throughout the sample set.

Figure 3a and 3b show reflectance spectra associated with Perry Park and Canon City samples. Data were acquired in the laboratory with an ASD FieldSpec TM (www.asdi.com) and tungsten illumination. We focus on the SWIR region, as characteristic clay spectral features are around 2.2 mm. All spectra show some clay features, smectite for most of them (single well-defined absorption band), sometimes with illite (additional band at 2.35 mm, e.g. PP97011), and sometimes with kaolinite (doublet band, e.g. CC97011). They also all show a deep water absorption band at 1.9 mm, which is indicative of the presence of montmorillonite in the samples. The spectra with a prominent smectite feature are associated with Perry Park samples that are pure or almost pure smectite (99% to 100% smectite in clay fraction). Those samples are the ones with very high swell potential (e.g. PP97008). The spectra where illite can be detected with smectite are associated with the samples PP97011, PP97001, CC97012 and CC97014. The PP97011 spectrum shows a strong illite signature (illite content ~60%, low swelling potential), unlike the three other spectra that show less illite. Illite content in CC97012 and CC97014 is ~20 and 40%, and they have a moderate and low swell potential. In PP97001 however illite was barely detected in the XRD results (less than 1%) and the swell potential is only moderate, which suggests a possible XRD error (from the splits of the sample?) or that the percent clay fraction and/or low porosity have a stronger influence than its mineralogy. Kaolinite is detected in the spectra from CC97008, 009, 010. Those samples have a kaolinite content >10%, and moderate swelling potential. Kaolinite was detected also in CC97011 and 013, which are the two samples with a pure smectite phase but also some kaolinite. Those samples have a high-very high swell potential (i.e., high smectite content).

In short, relationships between mineralogy, swell potential, and reflectance are multiple and complex, but follow some simple general rules: (a) from spectral reflectance we are able to discriminate among pure smectite and smectite/illite samples. The absorption band at 2.35 mm provides a measure of the illite content. The higher the amount of smectite (the less illite), the higher the swelling potential; (b) kaolinite is detected spectrally if above 10% in clay fraction in the sample. A significant amount of kaolinite (>10-15%) is indicative of low swelling potential. In the case of a significant content of kaolinite with pure smectite (case considered rare and unusual), the kaolinite doublet masks the smectite spectral feature at 2.2 mm, but the high amount of smectite (indicative of high swelling potential) can be spectrally detected in the limb of the 1.9 mm water absorption band.

Our goal now is to be able to apply these results at the remote sensing scale. The main difficulties which usually arise at this level are: (a) water absorption bands are masked by the atmospheric water vapor features; (b) natural surfaces are seldom perfect exposures of homogeneous expansive clays; and (c) the presence of vegetation can complicate the analysis.

AVIRIS scenes were acquired over the Urban Corridor in Colorado in 1997 and 1998. We focus here on a few scenes selected over the most promising areas in terms of amount and quality of exposures of expansive clays. Those scenes were acquired in 1997 for the U.S. Geological Survey over Colorado Springs (see location in fig. 1). The amount of green vegetation cover is at a near minimum at the time of the flight in late September. Each scene was first radiometrically corrected by the Jet Propulsion Laboratory, then a preliminary calibration relative to ground reflectance was performed by the U.S. Geological Survey using the following procedure: 1- atmospheric correction using ATREM 3.0 (CSES, 1997) to model the solar irradiance; 2- dark subtraction using the darkest pixel in the image to remove path radiance and scattering; and 3- ground correction using field spectra acquired the day after AVIRIS was flown on a calibration target of a bare field. This last correction is to remove the remaining systematic effects from both the instrument and the atmospheric model.

Two parallel, adjacent N-S flight lines were flown, 4 scenes for the eastern line and 3 for the western line. The scenes in each flight line were mosaicked together, and the western flight line was registered to the eastern one using ground control points. Figure 4 shows the resulting image (1168x2050 pixels) at 0.65 mm, covering an area of ~20x37 km. Because of the average ~1600m surface elevation in the area covered, the AVIRIS pixel size is ~18x18 m, with a sample spacing of 16 m. Notice in the image that the extreme western part is mostly mountainous, unlike the eastern part which is flat. When we compare this image with Hart’s (1974) geological map (fig. 1), the mountainous part is composed of pre-Colorado group rocks, and the flat part is either Pierre Shale formation (in the northeastern part of the image), or Colorado Group formations, like Graneros Shale (in the southern part of the image). Figure 5 shows AVIRIS calibrated spectra from a vegetated field, two types of soils (Rock/Soil 2 is a clay), and a lake.

As the goal of this study is to find and map expansive clay soils in the AVIRIS images, and since we have not yet sampled any outcrop in the area described above, we applied the following procedure, using only the 2-2.45 mm spectral region:

1. A minimum Noise Fraction (MNF) transform was performed to determine and reduce the dimensionality of the data;
2. The Pixel Purity Index (PPI) method was used to find the most spectrally pure (extreme) pixels;
3. The extreme pixels were visualized in the data cloud from the five first MNF bands. We were pleased to identify one of the cluster as pixels showing characterictic spectral clay features at 2.2 mm, such as kaolinite, smectite, and possibly illite;
4. Extreme clay pixels (endmembers) were extracted from the "clay" cluster;
5. A Matched Filtering method was eventually used to locate and map expansive clays outcrops. This last analysis provides an abundance map for each selected endmember.

Figure 6 presents the three AVIRIS spectra associated with the endmembers selected in the clay cluster. From spectral identification, they are labeled as smectite (single band at 2.2 mm), kaolinite (doublet band), and smectite/illite (additional band at 2.35 mm). Figure 7 shows the abundance maps obtained with the matched filtering algorithm for the three endmembers smectite (fig. 7a), kaolinite (fig. 7b) and smectite/illite (fig. 7c). The greyscale coding ranges from black (meaning that the spectrum of the associated pixel doesn’t match with the endmember one, or abundance=0%) to white (meaning a perfect match between the spectra of the pixel and the endmember, or abundance=100%). In general, the location of all pixels coded from grey to light grey /white represents expansive clay outcrops in the image. Figure 7b shows that kaolinite is mapped only in the extreme western part of the image. Figures 7a and 7c show that variable exposures of smectite and smectite/illite can be found in the eastern part of the scene. If we apply our results from the laboratory analyses to interpret AVIRIS results in terms of swell potential, it shows that: the mountainous western slopes are low or non-expansive; the flat eastern slopes are composed of expansive soils of variable swell potential. The highest swell potential outcrops are the ones mapped with the highest abundance in the endmember smectite, and the ones more abundant in the endmember smectite/illite are less swelling. When there is some kaolinite mapped in the eastern area, this area is low to non swelling.

In summary, the AVIRIS-derived maps of clay types (fig. 7) can be divided into two main areas, each with a characteristic suite of spectral clay signatures. The western part of the scene exhibits a relatively homogeneous clay signature interpreted to be kaolinite. Geologically, this area contains Precambrian-age granite and diorite and Pennsylvanian- and Permian-age sandstone from the Fountain Formation (Scott and Wobus, 1973), which are known to contain kaolinitic clays. These rocks were considered non-expansive by Hart (1974) because they generally lack expansive clays. The eastern part of the scene contains combinations and bands of two clay signatures interpreted to be smectite and mixed-layer illite/smectite. These appear to correspond with bands of steeply to gently dipping Pierre Shale that were mapped by Hart (1974). Hart recognizes bands of moderately to highly swelling claystone that contain differing amounts of smectite. The scattered nature of clay occurrences in the AVIRIS scene is interpreted to be due to a predominance of non-clay materials at the ground surface (i.e., either vegetation cover or non-expansive, Quaternary-age alluvial deposits).

In summary, we have been able to use spectral identification to discriminate among pure smectite and mixed smectite/illite samples. The higher the smectite content, the higher the swelling potential. The absorption band at 2.35 mm provides a measure of the illite content. Spectral detection of kaolinite is possible if it comprises above 10% in clay fraction, and a high content of kaolinite will be indicative of a poor swelling potential. Spectroscopic identifications of expansive clay field samples are well correlated with mineralogical x-ray diffraction analyses and geotechnical engineering tests.

The AVIRIS images processed were over an unknown area in terms of exposures of swelling clays. The analysis of the data showed that there are some clay exposures in the image, and that we were able, using a matched filtering, to identify them among the other components in the image and despite the mixing at the sub-pixel scale (vegetation cover). A map of exposed clay material was produced. Among those exposures, spectral discrimination and identification of variable clay mineralogy, related to variable swelling potential, is possible.

From additional analyses not shown here, it appears that the spectral determination of quantitative mineralogy and swell potential from the reflectance signature would be possible only by the use of statistical multivariate analyses. The studies are currently limited as there have been very few engineering-geologic studies that include clay mineralogy (XRD) and swell behavior (suction) testing and analysis. However, hyperspectral image analysis has shown to be capable of detecting and mapping expansive clays, even at a 20m pixel size, in the presence of significant vegetation cover.

This work was carried out under NASA/JPL contract No. 960983. We acknowledge Eric Livo, at the U.S. Geological Survey in Denver, Colorado, for calibration of the AVIRIS data and letting us use their data set.

Center for the Study of Earth from Space (CSES), Atmosphere REMoval Program (ATREM), Version 3.0, University of Colorado, Boulder, 27p., 1997.

S. Chabrillat, A.F.H. Goetz, H.W. Olsen and D.C. Noe, "Field spectrometry techniques for identification of expansive clay soils", in Proceedings of the 12^th Int. Conference on Applied Geologic Remote Sensing, Denver, Colorado, Nov. 17-19 1997, I:141:148, 1997.

J. I. Drever, "The preparation of oriented clay minera specimens for x-ray diffraction analysis by a filter-menbrane peel technique", American Mineralogist, Vol. 58, pp. 553-554, 1973.

J. D. Gill, M. W. West, D. C. Noe, H. W. Olsen and D. K. McCarty, "Geologic control of severe expansive clay damage to a subdivision in the Pierre Shale, Southwest Denver metropolitan area, Colorado", Clays and Clay minerals, Vol. 44 (4), pp. 530-539, 1996.

S. S. Hart, Potentially swelling soil and rock in the Front Range Urban Corridor, Colorado, Environmental Geology 7, 4 maps, scale 1:1,000,000, Colorado Geological Survey, Denver, CO, 1974.

R. G. McKeen, "A model for predicting expansive soil behavior", in Proceedings of the 7th Int. Conf. on Expansive Soils, Dallas, Texas, p. 1-6, August 3-5, 1992.

D. M. Moore and R. C. J. Reynolds, X-ray diffraction and the identification and analysis of clay minerals, Oxford University Press, p. 327, 1989.

L. G. Schultz, Quantitative interpretation of mineralogical composition from x-ray and chemical data for the Pierre Shale, U. S. Geological Survey professional paper 391-C, 1964.

G. R. Scott and R. A. Wobus, Reconnaissance geologic map of Colorado Springs and vicinity, Colorado, U.S. Geological Survey map MF-482, scale 1:62,500, 1973.

H. B. Seed, R. J. Woodard and R. Lundgren, "Prediction of swelling potential for compacted clays", Journal of American Society of Civil Engineers, Soil Mechanics and Foundations Division, Vol. 88 (SM3), pp. 53-87, 1962.

R. W. Thompson, "Swell testing as a predictor of structural performance", in Proceedings of the 7th Int. Conf. on Expansive Soils, Dallas, Texas, p. 84-88, August 3-5, 1992.