http://speclab.cr.usgs.gov

Distribution of Acid-Generating and Acid-Buffering Minerals
in the Animas River Watershed
as Determined by AVIRIS Spectroscopy

by

J. Brad Dalton, Trude V.V. King, Dana J. Bove,
Raymond F. Kokaly, Roger N. Clark, J. Sam Vance and Gregg A. Swayze

U.S. Geological Survey, Mail Stop 973,
Box 25046, Denver Federal Center,
Denver, Colorado 80225

Derived From:
Dalton, J.B., T.V.V. King, D.J. Bove, R.F. Kokaly, R.N. Clark, J.S. Vance and G.A. Swayze, Distribution of Acid-Generating and Acid-Buffering Minerals in the Animas River Watershed as Determined by AVIRIS Spectroscopy Proceedings if the ICARD 2000 Meeting, May 21-24, 2000, Denver Colorado.

ABSTRACT

Visible-wavelength and near-infrared multispectral image cubes for the Animas River Watershed from Hermosa, Colorado to the headwaters at Animas Forks, Colorado were acquired on June 18, 1996 using the Jet Propulsion Laboratory's AVIRIS (Airborne Visible and InfraRed Imaging Spectrometer) instrument (Green et al., 1998). These image cubes have been analyzed using the USGS Tetracorder V3.4 implementation (Clark et al., 1995), an expert system which utilizes a database of more than 250 laboratory spectra of endmember minerals and mineral mixtures to generate maps of mineralogy, vegetation coverage, and other material distributions. Major iron-bearing, clay, carbonate, and other minerals were identified along with several minerals associated with acid-generating hydrothermal systems including pyrite, jarosite, alunite and goethite. Additionally, distributions of alkaline minerals such as calcite and chlorite indicate a relation between acid-buffering assemblages and stream geochemistry within the watershed.

INTRODUCTION

The Animas River Watershed is the site of a coordinated effort by several federal, state and local agencies to characterize the extent and severity of environmental effects from acid mine water drainage. This water originates both from numerous abandoned mine sites that date back up to a century or more, and from extensive areas of natural altered and mineralized outcrops. The headwaters of the Animas River are within the San Juan and Silverton volcanic calderas, which were responsible for creating large fractures and faults suitable for later mineralization. As part of the Abandoned Mine Lands (AML) Project of the United States Geological Survey (USGS), AVIRIS data were obtained over the San Juan Mountains and Animas Watershed and are being used in conjunction with field geologic mapping, geochemistry and geophysics to determine the relative extent of natural and anthropogenic sources of acid water runoff, and their effect on water quality.

THE AVIRIS DATASET

The Animas Basin AVIRIS data set is comprised of 14 AVIRIS scenes, approximately 10.5 km x 8 km apiece, extending from Hermosa, Colorado to the headwaters of the Animas River north of Silverton, near the ghost town of Animas Forks. The 224 spectral channels cover a range from .37 to 2.51 µm, encompassing visible and near-infrared wavelengths suitable for mapping a wide variety of vegetation and minerals, including but not limited to the hydrothermally altered rocks of relevance to the San Juan Mountains. The 14 scenes are arranged in two overlapping lines parallel to and bounding the Animas river as it flows from its headwaters toward Durango, Colorado. This includes most of the Silverton caldera which is situated within the watershed. The data were acquired under cloud-free late morning conditions on June 18, 1996 in a 17-minute data collection pass. While yearly precipitation is normally quite high in the San Juan Mountains, the resulting thick vegetation still leaves significant outcrops of exposed rock available for analysis; in addition, at the time of acquisition of this dataset, the snowpack was greatly reduced due to melting.

REGIONAL GEOLOGY AND MINING ACTIVITY

The Animas watershed is located in the western part of the mid-late Tertiary age San Juan volcanic field (Lipman et al., 1976; Bove et al., in press). The local geology is largely comprised of lavas and related volcanic rocks associated with the San Juan caldera and the younger Silverton caldera (Steven and Lipman, 1976). Collapse of these calderas created ring fractures which provided conduits for later episodes of intense hydrothermal alteration and mineralization associated with dacitic to rhyolitic intrusion (Casadevall and Ohmoto, 1977; Lipman et al., 1976). More recent geologic activity has been dominated by uplifts during the Neogene (Steven et al., 1995), followed by down-cutting and the formation of the Animas River and its tributaries, with brief episodes of sediment deposition. Gold was discovered in the San Juan Mountains in 1871 and upwards of a thousand mining claims were staked in the upper Animas River above Silverton within the next two decades. The Denver and Rio Grande railroad was extended from Durango to Silverton in 1882 and ore production continued at various levels until 1991 when the Sunnyside Mine directly upstream from Silverton was closed (Church et al., 1997). Remediation of private holdings in the region continues today, and the AML project is partly concerned with evaluating remediation needs of the area, including the thousands of abandoned prospects and mines on both federal and private properties.

MINERAL MAPPING PROCEDURES

The AVIRIS data for the Animas Watershed were converted to apparent surface reflectance using the radiative transfer methods of ATREM (Gao et al., 1993; Gao et al., 1997) followed by calibration to ground reflectance using a path radiance correction (Clark et al., 1993a). The reflectance data were then analyzed using Tetracorder (Clark et al., 1990, 1991, 1993b, 1995), testing for the presence of over 250 minerals, mineral mixtures, water, vegetation, and other materials of interest. The most significant materials detected were then assembled into color-coded mineral maps. For this investigation, attention focused on the pyrite weathering sequence, wherein pyrite weathers to jarosite, which in turn weathers to goethite, and then hematite. Pyrite oxidation is the primary source of mining-related acidic runoff because sulfuric acid (H2SO4) is a product of this weathering reaction (Swayze et al., 1996). Pyrite itself is difficult to detect directly using AVIRIS because of its extremely low reflectance; however its oxidation products are spectrally bright and have strong 1-µm absorptions due to Fe3+ in the matrix. Infrared spectra of the pyrite weathering sequence minerals (Figure 1) show a broad Fe3+ absorption band centered near 0.8 µm. The exact center of this band (arrows) and its shape are different for jarosite, hematite, and goethite, allowing them to be mapped by the Tetracorder algorithm. Jarosite can be further differentiated by a strong absorption near 2.3 µm. These minerals are abundant in intensely mineralized regions such as in the Red Mountain area (see Bove et al., this volume) north of the Silverton mining district, and elsewhere throughout the watershed. Mine tailings in the San Juan Mountains are typically weathered to jarosite and goethite, and the surface expressions of these minerals readily pinpoint mines and prospects in the AVIRIS scenes. The oxidation process can create distinctive patterns such as bull's eyes of mineral zones on waste rock piles (Swayze et al., 1996). Creeks carrying runoff from tailings piles and mineralized outcrops may have pH values as low as 1.8, and this high acidity level enables these streams to carry high concentrations of trace metals (Church et al., 1997) such as zinc, copper, cadmium, arsenic, and lead. Stream beds and banks typically display thick rinds of precipitated iron oxide materials, which also are quite evident in the Tetracorder-processed AVIRIS mineral maps.



12K GIF

Figure 1 Infrared spectra of primary pyrite weathering products jarosite, goethite, and hematite. Each has a strong Fe absorption centered near .8 µm (arrows.) Due to changes in crystal structure, each mineral exhibits different band positions, strengths, and shapes. Note also that jarosite has a distinctive absorption in the 2.3 µm range that is useful for discrimination.


For this study, the AVIRIS data were analyzed in two separate groups based on spectral range. The Group 1 minerals were those with strong absorptions in the 1-µm wavelength range, such as the pyrite sequence, and other iron-bearing minerals. The Group 2 minerals similarly have their most distinct signatures in the 2-µm wavelength range. This group includes most clay minerals, as well as chlorites, epidote, carbonates, and muscovites. Figures 2 and 3 display the preliminary mineral maps for an AVIRIS scene which encompasses the town of Silverton. Although Tetracorder in its current implementation can identify over 250 different materials, only a relative few were considered relevant to this paper. The dominant material in each 17.5 m AVIRIS pixel was selected by Tetracorder on the basis of a least squares comparison of absorption features in the AVIRIS data to those in the reference library. Each map includes a subset of the minerals considered by Tetracorder; black regions in images correspond to unmapped pixels or materials not under consideration, such as snow cover, minor mineral phases, or the vegetation cover which is prevalent in the Animas Basin.The Silverton street grid is expressed by the distribution of mineral signatures near the left center of the images. The Animas River enters the scene from the top right and meanders past the town before exiting on the lower right. Cement Creek and Mineral Creek enter the image from the left; Cement Creek is above the town to the northwest and Mineral Creek below. A typical outcrop of goethite and jarosite was mapped just west of the town and north of Mineral Creek. Although located on a steep slope, the bull's eye pattern (jarosite at the center, goethite surrounding) indicative of pyrite weathering can be clearly distinguished. Such outcrops are major sources of acidic runoff. In the Group 2 map (Figure 3) this location is also characterized by a significant concentration of alunites, another indication of the hydrothermal alteration responsible for local ore formation. In this case, the outcrops are related to abandoned mine adits, but significant natural occurences are also widespread elsewhere in the watershed. In the bottom left of both maps, a series of sedimentary outcrops are highlighted; these outcrops are outside the caldera and are composed primarily of limestones, and are mapped by Tetracorder as calcite.

CALCITE, EPIDOTE AND CHLORITE ASSEMBLAGES

Recent work has shown that carbonate minerals in the San Juan Mountains may play an important role in the buffering of acidic stream waters, reducing metal concentrations and carrying capacities (Church et al., 1997; Bove et al., this volume). Calcite occurs in the San Juan Mountains both as sedimentary limestone, and as alteration products of primary minerals within volcanic rocks. While calcite within propylitically altered rocks may have a reduced role due to limited availability for buffering reactions (Runnels and Rampe, 1989), the importance of calcite in buffering reactions has been established for limestones (ibid.). Magnesium-bearing chlorites such as clinochlore may also provide some level of buffering of acidic stream reaches, however this has not yet been fully quantified (Desborough et al., 1998; Kwang and Ferguson, 1997). Calcite has been identified and mapped in the Animas River watershed using imaging spectroscopy methods, and its presence in particular stream reaches correlates with increased pH levels and attendant lowered dissolved trace metal concentrations (Church et al., 1997). Unequivocal identification of calcite in the AVIRIS data is complicated by the spectral absorption bands of epidote and chlorite which overlap the 2.3-µm diagnostic calcite absorption. Calcite-epidote-chlorite-bearing lavas are widely distributed in the western San Juan Mountains.



81K GIF

Figure 2. Preliminary mineral map of the Silverton scene from the 1996 AVIRIS overflight, showing "Group 1" (1-µm range) minerals as mapped by Tetracorder V3.4. Major alteration and weathering phases are included, as well as five broad classes of iron-bearing minerals. Black regions are unmapped pixels; in the San Juans this is predominantly vegetation though many of the black pixels here are simply not under consideration, such as standing water bodies, snow cover, and minor mineral phases.




100K GIF

Figure 3. Preliminary mineral map of the Silverton scene from the 1996 AVIRIS overflight, showing "Group 2" (2-µm range) minerals as mapped by Tetracorder V3.4. Clays, carbonates, and alunites are of primary interest in this map. Potentially acid-buffering assemblages of calcite, epidote, and chlorite are shown together in one spectral class.


Sample infrared spectra for calcite, chlorite, and epidote are shown in Figure 4. The diagnostic absorption bands for each of these minerals are centered near 2.3 µm, and are difficult to distinguish from one another. Based on shapes and band positions, the Tetracorder algorithm can separate the three end members; however when they occur as a fine-grained mixture the individual absorptions combine in nonlinear ways to produce a new absorption spectrum. During the 1998 field season, over 150 samples of calcite-epidote-chlorite lavas were collected in the Animas Basin, and work is now underway to incorporate the complex spectral behavior of these mixtures into the Tetracorder analysis database. The lower spectrum in Figure 4 is of one such sample, a mixture of all three minerals in a finely-grained propylitic lava rock from the Burns formation. Using the full spectral resolution of the AVIRIS instrument, the Tetracorder mapping technique should be able to use such information to define the calcite, epidote and chlorite concentrations in the basin. Areas presently mapped as simply calcite-epidote-chlorite in Figures 3 and 6 will ideally be separated into groups dominated by one of the three minerals, enabling a more accurate assessment of acid-buffering potential. Defining the extent of this buffering capability, however, and improved mapping of these lavas, is an ongoing aspect of our current work.



12K GIF

Figure 4 Infrared spectra of calcite, epidote, and chlorite, along with the spectrum of sample SJ98-74D, which is a rock comprising a fine-grained mixture of all three in a propylitically altered matrix.

PRELIMINARY AVIRIS MINERAL MAPPING RESULTS

According to Figure 5, pyrite-weathering products are abundant in the mountains surrounding the town of Silverton; particularly to the north and west, where hydrothermal activity was more intense. Below the confluence of Cement Creek and the Upper Animas, amorphous iron hydroxides have precipitated from the stream and deposited along the river margins. This is due to an increase in the pH in the river: the highly acidic water of Cement Creek is diluted by the more neutral waters of the Upper Animas. Although acid-generating minerals are evident throughout the upper portion of the scene, Figure 6 reveals a much higher concentration of calcite-epidote-chlorite assemblages along the Upper Animas. Aqueous geochemistry has demonstrated (Church et al., 1999) that stream pH is indeed higher along several tributaries to the Upper Animas where these buffering assemblages occur. Along lower stretches of the Animas, water quality continues to improve, especially after the section near Molas Lake (Figure 6, lower left) which contains abundant sedimentary limestone beds, mapped as calcite in Figure 6 and confirmed during the 1998 field season. The absence of buffering mineral assemblages along Mineral and Cement Creeks is one factor



337K GIF

Figure 5. Single-band AVIRIS image of Silverton scene with pyrite-weathering products superimposed in colors, as mapped by Tetracorder V3.4. Mineral assemblages indicative of acid runoff superimposed on a single-plane AVIRIS band image to highlight effects of acid-generating minerals on stream quality. Outcrops of pyrite-weathering sequence minerals do not fully correlate with poor water quality. Iron hydroxide precipitates are apparent lining stream edges in this image.




260K GIF

Figure 6. Single-band image of the Silverton scene, with acid-buffering calcite, epidote, and chlorite assemblages superimposed in color as mapped by Tetracorder V3.4. Potentially acid-neutralizing mineral assemblages of calcite, epidote and chlorite superimposed over a single-plane AVIRIS band image. Improved stream quality in the upper Animas River (top right) is probably due not to the absence of acid-generating minerals, but to the presence of calcite and chlorite in the watershed. Due to spectral overlap, calcite-epidote-chlorite mixtures are not separated in this image. Calcite present as limestone, however, is easily identified in the lower left of the image.


contributing to their high metal loads. More accurate discrimination between the calcite, epidote and chlorite constituents will play an important role in assessing the natural and artificial factors influencing stream water quality.

CONCLUSION

Mineral mapping via imaging spectroscopy, as performed using the AVIRIS instrument and USGS Tetracorder algorithm, provides a powerful technique for locating sources of acidic and acid-buffering drainage of either natural or anthropogenic origin. Diagnostic features in absorption spectra of pyrite weathering products provide for accurate identification. Acid-buffering minerals such as calcite and chlorite are also distinguishable in AVIRIS spectra, and work is underway to more accurately characterize them in their host phases and in combination with each other. Field studies have confirmed the presence of all the mineralogies presented in this study, and geochemistry studies within the basin corroborate the findings and indicate a strong link between acid-buffering assemblages and water quality. This approach will be extremely useful to property owners interested in assessing and remediating lands affected by acidic drainage.

This preliminary study, appropriate images, follow-on studies and related research can be found at our web site: http://speclab.cr.usgs.gov.

REFERENCES

Bove, D.J, M.A. Mast, W.G. Wright, G.P. Meeker, D.B. Yager, and P.C. Verplanck, (in press), Geologic Control on Acidic and Metal-rich Waters in the Southeast Red Mountain Area, Near Silverton, Colorado, Proceedings if the ICARD 2000 Meeting, May 21-24, 2000, Denver Colorado.

Casadevall, T. and H. Ohmoto, 1977, Sunnyside mine, Eureka mining district, San Juan County, Colorado: Geochemistry of gold and base metal ore deposition in a volcanic environment: Economic Geology, v. 92, pp. 1285-1320.

Church, S.E., B.A. Kimball, D.L. Fey, D.A. Ferdener. T.J. Yager, and R.B. Vaughn, 1997, Source, Transport, and Partitioning of metals between Water, Colloids, and Bed Sediments of the Animas River, Colorado, U.S. Geological Survey Open File Report 97-151.

Church, S.E., D.L. Fey, E.M. Brouwers, C.W. Holmes, and R. Blair,1999, Determination of Pre-Mining Geochemical Conditions and Paleoecology in the Animas River Watershed, Colorado, U.S. Geological Survey Open File Report 99-0038.

Clark, R.N., A.J. Gallagher, and G.A. Swayze, 1990, Material Absorption Band Depth Mapping of Imaging Spectrometer Data Using a Complete Band Shape Least-Squares Fit with Library Reference Spectra, Proceedings of the Second Airborne Visible/Infrared Imaging Spectrometer (AVIRIS) Workshop. JPL Publication 90-54, 176-186.

Clark, R.N., G.A. Swayze, A. Gallagher, N. Gorelick, and F. Kruse, 1991, Mapping with Imaging Spectrometer Data Using the Complete Band Shape Least-Squares Algorithm Simultaneously Fit to Multiple Spectral Features from Multiple Materials, Proceedings of the Third Airborne Visible/Infrared Imaging Spectrometer (AVIRIS) Workshop, JPL Publication 91-28, 2-3.

Clark, R.N., Swayze, G.A., Heidebrecht, K.B., Goetz, A.F.H., and R.O. Green, 1993a, Comparison of Methods for Calibrating AVIRIS Data to Ground Reflectance, Summaries of the 4th Annual JPL Airborne Geosciences Workshop, Volume 1: AVIRIS Workshop, JPL Publication 93-26, pp. 31-34.

Clark, R.N., G.A. Swayze, and A. Gallagher, 1993b, Mapping Minerals with Imaging Spectroscopy, U.S. Geological Survey, Office of Mineral Resources Bulletin 2039, pp. 141-150.

Clark, R.N. and G.A. Swayze, 1995, Mapping Minerals, Amorphous Materials, Environmental Materials, Vegetation, Water, Ice and Snow, and Other Materials: The USGS Tricorder Algorithm. Summaries of the Fifth Annual JPL Airborne Earth Science Workshop, January 23- 26,, Vol. 1: AVIRIS Workshop, R.O. Green, Ed., JPL Publication 95-1, p. 39-40.

Desborough, G.A., P.H. Briggs, and N. Mazza,1998, Chemical and Mineralogical Characteristics and Acid-Neutralizing Potential of Fresh and Altered Rocks and Soils of the Boulder River Headwaters in Basin and Cataract Creeks of Northern Jefferson County, Montana, U.S. Geological Survey Open File Report 98-40.

Gao, B.C., K.B. Heidebrecht, and A.F.H. Goetz, 1993, Derivation of Scaled Surface Reflectances from AVIRIS Data, Remote Sensing of Environment, v. 11, 99. 1775-1795.

Gao, B.C., K.B. Heidebrecht, and A.F.H. Goetz, 1997, Atmosphere Removal Program (ATREM) Version 3.0 User's Guide, Center for the Study of Earth from Space, University of Colorado at Boulder.

Green, R.O., D.A. Roberts, and J.E. Conel, Characterization and Compensation of the Atmosphere for the Inversion of AVIRIS Calibrated Radiance to Apparent Surface Reflectance. Summaries of the Sixth Annual JPL Airborne Earth Science Workshop, March 4-8, 1996, JPL Publication 96-4, Vol. 1, AVIRIS Workshop, pp. 135-146.

Green, R.O., M.L. Eastwood,C.M. Sarture T.G., Chrien., M. Aronsson, B.J. Chippendale, J.A. Faust, B.E. Pavri, C.J. Chouit, M. Solis, M.R. Olah, and O. Williams, 1998, Imaging Spectroscopy and the Airborne Visible Infrared Imaging Spectrometer (AVIRIS), Remote Sens Environ, v.65;(3), pp. 227-248.

Kwang, Y.T.J., and K.D. Ferguson, 1997, Mineralogical Changes During NP Determinations and their Implications, Proceedings of the Fourth International Conference on Acid Rock Drainage, Vancouver, B.C., May 31-June 6 1997, pp. 435-447.

Lipman, P.W., H.H. Mehnert, C.W. Naeser, R.G. Luedke, and T.A. Steven, 1976, Multiple Ages of mid-Tertiary mineralization and alteration in the Western San Juan Mountains, Colorado, Economic Geology, v. 71, pp. 571-588.

Runnells, D.D., and J.J. Rampe, 1989, Natural Clean-up of a Stream Contaminated by Mine and Mill Wastes: A Reaction Analogous to Liming in Water Treatment, Proceedings of the 6th International Symposium on Water-Rock Interaction, Malvern, 3-8 August 1989, Ed. D.L. Miles, pp. 597-598.

Steven, T.A., K. Hon, and M.A. Lanphere, 1995, Neogene Geomorphic Evolution of the Central San Juan Mountains Near Creede, Colorado, U.S. Geological Survey Map I-2504, Denver Colorado.

Steven, T.A., and P.W. Lipman, 1976, Calderas of the San Juan Volcanic Field, Southwestern Colorado: U.S. Geological Survey Professional Paper 958.

Swayze, G.A, R.N. Clark, R.M. Pearson, and K.E. Livo, 1996, Mapping Acid-Generating Minerals at the California Gulch Superfund Site in Leadville, Colorado using Imaging Spectroscopy. Summaries of the 6th Annual JPL Airborne Earth Science Workshop, JPL Publication.

Swayze, G.A., R.N. Clark, K.S. Smith, P.L. Hageman, S.J. Sutley, R.M. Pearson, G.S. Rust, P.H. Briggs, A.L. Meier, M.J. Singelton, and S. Roth, 1997, Using Imaging Spectroscopy to Cost-Effectively Locate Acid-Generating Minerals at Mine Sites: An Example From the California Gulch Superfund Site In Leadville, Colorado. Geological Society of America Abstracts with Programs, October 20-23, 1997, v. 29, no. 6, pp. A-322.


U.S. Geological Survey, a bureau of the U.S. Department of the Interior
This page URL= http://speclab.cr.usgs.gov/PAPERS/animas1999/animas.html
This page is maintained by: Roger N. Clark and J. Brad Dalton rclark@speclab.cr.usgs.gov
Last modified Jan. 14, 2000.