Summaries of the 6th Annual JPL Airborne Earth Science Workshop March 4-8, 1996

(View Press Release from March 13, 1996)

Mapping Acid-Generating Minerals at the California Gulch Superfund Site in Leadville, Colorado using Imaging Spectroscopy


Gregg A. Swayze, Roger N. Clark, Ronald M. Pearson*, and K. Eric Livo

U.S. Geological Survey, Box 25046 MS964 DFC Denver, CO 80225

*U.S. Bureau of Reclamation, P.O. Box 25007 D-8321 DFC, Denver, CO 80225


The Leadville mining district, located at an elevation of 3000 m in the Central Colorado Rockies, has been mined for gold, silver, lead and zinc for over 100 years. This activity has resulted in waste rock and tailings, rich in pyrite and other sulfides, being dispersed over a 30 sq. km area including the city of Leadville. Oxidation of these sulfides releases lead, arsenic, cadmium, silver, and zinc into snowmelt and thunderstorm runoff, which drains into the Arkansas River, a main source of water for Front Range urban centers and agricultural communities. The U.S. Environmental Protection Agency (EPA), U.S. Bureau of Reclamation (USBR), contractors, and responsible parties are remediating the mined areas to curtail further releases of heavy metals into various drainage tributaries of the Arkansas River. The USBRs' current work is directed toward characterizing and pinpointing localized sources of acid mine drainage and contamination from waste rock piles found throughout an area of 900 hectares.


The sulfide oxidation process is biologically driven along complex chemical pathways with feedback reactions that enhance the speed and magnitude of oxidation (Nordstrom, 1982). Release of heavy metals is facilitated by sulfide oxidation, since many of the sulfides contain the heavy metals (e.g. Pb, As, Cd, Ag, and Zn). The oxidation-weathering process produces low pH water in which the heavy metals dissolve as aqueous phases that are then transported by runoff into nearby streams. Secondary minerals such as jarosite, ferrihydrite, schwertmannite, goethite, and hematite are formed by sulfide oxidation or precipitation from metal-rich water. These secondary minerals are Fe-rich and usually hydroxyl-bearing, making it possible to identify them from their unique spectral signatures. As the pH of the stream water increases, from dilution with higher pH sources, the secondary minerals precipitate out as streambed coatings. Because the heavy metals can substitute for Fe, they are also precipitated from solution as constituents of secondary minerals or as contaminants adsorbed onto the surfaces of the secondary minerals (Smith and Macalady, 1991; Alpers et al., 1994). Subsequent pulses of low pH water may dissolve the secondary minerals and remobilize the heavy metals and transfer them downstream.

The sulfide mineral pyrite is a primary source of acid drainage. Because direct spectral detection of pyrite is hampered by its low reflectance level, its broad Fe-absorption, and usual opaque coating of oxidation products, pyrite can only be detected when it is locally concentrated. However, our study shows that pyrite weathers first to copiapite, a mixed Fe2+- Fe3+ -sulfate, and then to jarosite, and eventually to hematite or goethite (Table 1), forming a sequence where the degree of oxidation and weathering is indicated by the species of secondary mineral that forms. Therefore, an indirect way to find oxidizing pyrite is to look for areas where the secondary minerals grade through the oxidation-weathering sequence (e.g. those areas with copiapite or jarosite surrounded by goethite or hematite). Quite fortuitously, the presence of heavy metals and low pH associated with the mine waste prevent the growth of vegetation over most waste piles leaving them exposed.

Table 1. Pyrite and Secondary Minerals

	Pyrite		FeS2 *

	Copiapite	Fe2+Fe3+(SO4)6(OH)2.20H2O

	Jarosite	(Na,K)Fe3+3(SO4)2(OH)6

	Goethite	alpha-FeO(OH)

	Hematite	alpha-Fe2O3

* Formulas from Fleisher (1980)

The size of the Leadville mining district and presence of spectrally detectable secondary minerals from the pyrite oxidation-weathering process makes imaging spectroscopic analysis effective for locating those minerals related to the acid mine drainage sources. AVIRIS data was collected over Leadville on July 27, 1995 as part of the NASA SIR-C Data Integration Project (lead by F. Kruse). Data were calibrated from radiance to reflectance in two stages: 1) removal of solar flux, atmospheric absorption, and path radiance using the ATREM program (Gao et al., 1992), and 2) further corrections of the ATREM reflectance data using ground reflectance targets. One of the additional corrections was needed because ATREM overcorrected the AVIRIS scene's path radiance. We derived this correction by first extracting an average spectrum from the ATREM cube over fir trees in a shaded ravine, and then dividing this into a fir tree spectrum from our library. Next we used the low points in this spectral ratio to cubic spline a new curve which was then added back into the ATREM data cube to compensate for the overcorrection. Since the terrain around Leadville is heavily vegetated we used the Sugarloaf earthen-fill dam as our high reflectance ground calibration target. The dam surface consists of granite boulders and smaller fragments of metamorphic gneiss piled to form an irregular surface dipping uniformly away from the reservoir. Because of the heterogeneity of this site, a portable field spectrometer was used to record 300 reflectance spectra, each of a 0.4 meter diameter roving spot of the dam surface, relative to a spectralon standard. These spectra were averaged and then used to derive a multiplier which was applied to the ATREM cube to removed the remaining artifacts. Given the heterogenous composition of the dam's surface and the need to produce a correction relatively free of noise (to preserve the high S/N of the AVIRIS data), use of a portable field spectrometer was critical to producing the final AVIRIS reflectance calibration.

Calibrated AVIRIS reflectance data was spectroscopically mapped using the Tricorder algorithm (Clark et al., 1995; Clark et al., in prep.). Tricorder is an expert system which is capable of simultaneously analyzing spectra of solids, liquids, and gases. It can preprocess data, then do an initial analysis, and from this decide on a number of alternate analysis paths. Tricorder's primary subroutine is a least-square shape-matching algorithm which compares spectra of unknown materials to hundreds of reference library spectra and picks the best spectral match. This subroutine uses thresholds on band depth, fit, and continuum slope to help constrain the spectral matching. The spectral library used to map AVIRIS Leadville data contains reference spectra of the individual secondary minerals and some mixtures of these, along with 160 spectra of other materials. Because the visible to 1.0-um spectral region is dominated by electronic absorptions due to transition metals, including Fe, and because the 2 to 2.5-um region is dominated by molecular vibrational absorptions, maps of the dominant materials in each spectral region were produced for the Leadville AVIRIS scene.


Examination of the electronic-absorption map (Figures 1, 2, and 3) shows the waste piles as having small zones of jarosite surrounded by jarosite-goethite zones, in turn surrounded by goethite and hematite zones. Field checks of these oxidation-weathering zones with a portable spectrometer and hand lens indicates that the jarosite zones are pyrite- rich. Coarse pyrite weathers more slowly than fine pyrite and enough had survived weathering that it was mapped in the Oregon Gulch tailings. At the Apache pile, concentrations of finer grained pyrite (used as a pigment to color beer bottles), had an oxidation coating of copiapite which was also identified and mapped. Reprocessed tailings in the Stray Horse Gulch waste piles mapped as hematite, providing a way to aerially separate these materials from the goethite- rich waste rock. Areas surrounding the site of an old smelter spectrally mapped as an amorphous Fe-hydroxide. Field investigation in these areas showed rocks coated with a grayish-coating of the condensed arsenic-rich effluent from the smelter smoke stacks.

Figure 1: 58K GIF

Figure 1: Leadville iron-bearing mineral map. Map of the mineral distribution of the waste rock and tailings piles at the California Gulch Superfund Site near Leadville, Colorado (130 km SW of Denver). This map was produced by the U.S. Geological Survey for the U.S. Bureau of Reclamation and U.S. Environmental Protection Agency using the NASA/Jet Propulsion Laboratory Airborne Visible/Infrared Imaging Spectrometer (AVIRIS). Each color identifies iron-bearing minerals in each 17 by 17 meter sqaure area (pixel) on the ground. Blue colors show minerals which cause acid mine drainage high in dissolved metals like cadmium, zinc, and lead. Areas in green have minerals that are more neutral but are still of concern. Other colors are minerals not contributing to water contamination. No iron-bearing minerals were found in areas shown in black. Area shown is 10.5 km wide and 17 km long. North is toward the lower right of the image.

Figure 2: 389K GIF

Figure 2: Leadville iron-bearing mineral map on aerial photographic base. Map of the mineral distribution of the waste rock and tailings piles at the California Gulch Superfund Site near Leadville, Colorado (130 km SW of Denver). This map was produced by the U.S. Geological Survey for the U.S. Bureau of Reclamation and U.S. Environmental Protection Agency using the NASA/Jet Propulsion Laboratory Airborne Visible/Infrared Imaging Spectrometer (AVIRIS). Each color identifies iron-bearing minerals in each 17 by 17 meter sqaure area (pixel) on the ground. Blue, purple, and yellow colors show minerals which cause acid mine drainage high in dissolved metals like cadmium, zinc, and lead. Areas in green have minerals that are more neutral but are still of concern. Other colors (red and orange) are minerals contributing less to water contamination. The mineral colors are overlain on an aerial photo base to aid in locating the potential acid generating areas. Note bright areas of waste rock and tailings piles where no detected acid-generating minerals were found. Image is about 6 km across.

Figure 3: 11K GIF

Figure 3: Idealized Acid-Mineral Assemblage Bull's Eye. Secondary minerals form roughly concentric zones around pyrite-rich areas at the surface. They have pyrite or jarosite in the center surrounded by a goethite-jarosite zone, in turn surrounded by either goethite or hematite forming bull's eyes around the hot spots of pyrite oxidation. These patterns can be used to locate point-sources of acid mine drainage at Leadville and probably at other mined and unmined areas by marking areas of low pH and potential trace metal mobility.

Examination of the vibrational absorption map shows that the waste piles abruptly change composition from Na-montmorillonitic to kaolinite-sericitic alteration across a series of interconnected NW-trending graben faults. Since mineralization was hosted in limestones on both sides of the fault, the differences in alteration may be related to changes in the chemistry and temperature of the hydrothermal fluids or the composition of the waste rock. These spectrally observed compositional differences may, nevertheless, fingerprint waste rock piles as originating from specific mines or ore bodies. At present, no clear relationship can be drawn between the distribution of clays and the location of acid drainage point-sources.


The mineral maps created from AVIRIS data are being used to guide USBR's investigation, characterization, and remediation efforts for the EPA at the California Gulch Superfund (NPL) Site. Based on the pattern of secondary oxidation-weathering minerals that form above pyrite-rich waste rock and tailings, a map of the electronic absorption spectral region can be used to locate potential sources of acid mine drainage. Those areas covered by jarosite and jarosite-goethite mine-waste are believed to have a high acid generating capacity and are surface point-sources for acid water and heavy metals. Areas covered by goethite and hematite can still contain heavy metals (adsorbed on or as components of the secondary minerals) but they will not be mobile in the absence of low pH water, normally found in the jarosite-rich areas. It is possible that the minerals at the surface of the waste piles and tailings may not reflect the pile's composition at depth. Depending on the amount of precipitation and level of the groundwater table, unoxidized sulfides may be mantled by a layer of non-acid generating secondary minerals. In these cases, surficial mineral maps will not show the acid- generating capacity of the waste rock and tailings beneath the surface (Figure 4). In most instances, we believe that these oxidation rinds will prevent water from penetrating to the sulfides, and in effect, isolate them from the aqueous environment. As part of our ongoing investigation, we plan to correlate surface mineralogy with soil metal-concentration and pH information from drill cores and geoelectrical profiling, to better address this possibility.

Significant cost savings in investigation and remediation design have already been realized using AVIRIS data at this National Priority List Site.

Figure 4: 11K GIF

Figure 4: Highly idealized profile of a waste rock pile. In places a thick rind of oxidized material composed of secondary minerals mantles unoxidized pyrite and other sulfides. If the rind is impermeable to water then the underlying pyrite will be effectively isolated from the aqueous environment. On south facing or steep slopes where freeze-thaw cycles cause debris flows, erosion will expose unoxidized pyrite, which in gaining access to the atmosphere can readily oxidize and produce a low pH zone rich in jarosite. We think most jarosite anomalies at Leadville are this type. These "hot spots" may be the dominant sources of acid mine drainage because oxidation proceeds most rapidly at the surface. It is possible that acid waters can enter the ground water through old covered shafts that would go undetected by AVIRIS, however, their locations can be mapped with other geophysical tools and drilling.


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, (ed.) R.O. Green, Jet Proplusion Laboratory Publication 95-1, p. 39-40.

Clark, R.N., G.A. Swayze, and T. V.V. King, in prep., Imaging spectroscopy: a new tool for identifying and mapping materials: minerals, amorphous materials, environmental and man-made materials, vegetation species, health and water content, water, ice, snow, and atmospheric gases: the USGS Tricorder algorithm: to be submitted to Science.

Alpers, C.N., D.W. Blowes, D.K. Nordstrom, and J.L. Jambor, 1994, Secondary minerals and acid mine-water chemistry: The Environmental Geochemistry of Sulfide Mine-Wastes, Short Course Handbook, (eds.) D.W. Blowes and J.L. Jambor, vol. 2, Waterloo, Ontario, Canada, Mineralogical Association of Canada, May 1994, p. 247-270.

Fleischer, M., 1980, Glossary of Mineral Species, Mineralogic Record, Tucson, Arizona, 192 p.

Gao, B.C., K.B. Heidebrecht, and A.F.H. Goetz, 1992, ATmospheric REMoval Program (ATREM) User's Guide, version 1.1, Center for the Study of Earth from Space document, University of Colorado, 24 p.

Nordstrom, D.K., 1982, Aqueous pyrite oxidation and the consequent formation of secondary iron minerals: in Acid Sulfate Weathering, eds J.S. Kittrick et al., Soil Society of America, p. 37-56.

Smith, K.S., and D.L. Macalady, 1991, Water/sediment partitioning of trace elements in a stream receiving acid-mine drainage: Second International Conference on the Abatement of Acidic Drainage, NEDEM, Montreal, Canada, Sept. 16-18., p. 435-450.

Links to related information:

U.S. Bureau of Reclamation Engineering Geology Team

Phone email and regular mail addresses of spectroscopy lab personnel for further information.

U.S. Geological Survey, a bureau of the U.S. Department of the Interior
This page URL=
This page is maintained by: Dr. Roger N. Clark
Last modified December 26, 1998.