|
|
SIR 2004-5241
Remote Sensing for Environmental
Site Screening and Watershed Evaluation in Utah Mine Lands—East Tintic
Mountains, Oquirrh Mountains ,
and Tushar Mountains
By Barnaby W. Rockwell, Robert R. McDougal, and
Carol A. Gent
Prepared in cooperation with the United States Environmental Protection Agency
Version 1.2
Printable PDF (22 MB)
Keywords: remote sensing, imaging spectroscopy, reflectance spectroscopy, AVIRIS, Landsat 7 ETM+, mineral mapping, abandoned mine lands, acid rock drainage, Superfund Program, geoenvironmental assessment, environmental impacts of mining, Utah mines, mining in Utah, Bingham Canyon mine, Bingham mine, International Smelter, Camp Floyd mining district, Mercur mine, Stockton mining district, Bauer Mill, Tintic mining district, Eureka, Dragon mine, Burgin mine, Trixie mine, Marysvale volcanic field, Deer Trail mine, Big Rock Candy Mountain, jarosite, alunite, goethite, halloysite
Table of Contents
AVIRIS Data Acquisitions, Reflectance Calibration, and Georectification
The East Tintic Mountains and the Tintic Mining District
The Tushar Mountains/Marysvale Region
Appendix. X-Ray Diffraction Results
Abstract
Imaging spectroscopy—a powerful
remote-sensing tool for mapping subtle variations in the composition of
minerals, vegetation, and man-made materials on the Earth's surface—was applied
in support of environmental assessments and watershed evaluations in several
mining districts in the State of
The Landsat 7 ETM+ data were used for initial site screening and the planning of AVIRIS surveys. The AVIRIS data were analyzed to create spectrally defined maps of surface minerals with special emphasis on locating and characterizing rocks and soils with acid-producing potential (APP) and acid-neutralizing potential (ANP). These maps were used by the United States Environmental Protection Agency (USEPA) for three primary purposes: (1) to identify unmined and anthropogenic sources of acid generation in the form of iron sulfide and (or) ferric iron sulfate-bearing minerals such as jarosite and copiapite; (2) to seek evidence for downstream or downwind movement of minerals associated with acid generation, mine waste, and (or) tailings from mines, mill sites, and zones of unmined hydrothermally altered rocks; and (3) to identify carbonate and other acid-buffering minerals that neutralize acidic, potentially metal bearing, solutions and thus mitigate potential environmental effects of acid generation.
Calibrated AVIRIS surface-reflectance data were spectrally analyzed to identify and map selected surface materials. Two maps were produced from each flightline of AVIRIS data: a map of iron-bearing minerals and water having absorption features in the spectral region from 0.35 μm to 1.35 μm and a map of minerals (including clays, sulfates, micas, and carbonates) having absorptions in the spectral region from 1.45 μm to 2.51 μm. Several methods were used to verify the AVIRIS mapping results, including field checking of selected locations with a portable spectrometer, visual inspection of the AVIRIS reflectance spectra, and X-ray diffraction (XRD) analysis of field samples.
The maps of iron-bearing minerals
derived from analysis of the visible (VIS) and near-infrared (NIR) regions of
the electromagnetic spectrum were shown to be more consistently reliable in
indicating the presence of jarosite than were the maps generated from analysis
of the short-wave infrared (SWIR) region. When present in abundance,
phyllosilicate minerals tend to dominate the SWIR and mask the spectral features
of jarosite in that wavelength region. The crystal field absorptions of
jarosite in the
Large exposures of unmined hydrothermally altered rocks occur throughout the three study areas. These rocks commonly contain sulfide or sulfate minerals that produce sulfuric acid upon subaerial oxidation. The acid may be introduced into local surface and ground water and thus lower the baseline (that is, the premining) pH for a watershed.
The three study areas also have widespread exposures of rocks with acid-neutralizing potential. Lithologies containing carbonates and (or) other acid-buffering minerals—such as sedimentary limestones and dolomites and propylitically altered igneous rocks—were mapped with the AVIRIS data throughout the Oquirrh and East Tintic Mountains and locally in the Antelope Range and Tushar Mountains.
Because elevated levels of various
heavy metals in local soils and tap water have been identified by previous
USEPA studies, parts of the town of
In the
In the
Analysis of Landsat 7 ETM+ data can provide a very cost-effective screening tool for identifying mineralized and (or) mining-affected areas and guiding the planning of low-altitude imaging spectrometer surveys or field investigations. Then, when coupled with a geological understanding of a study area, the interpretation of mineral maps derived from imaging spectroscopy data can be an effective means of (1) evaluating potential environmental impacts associated with hydrothermally altered rocks and mine waste on a watershed or regional scale and (2) focusing field-sampling and remediation programs.
Introduction
This report is a summary of the
results obtained from analysis and interpretation of spectroscopic imagery
collected by the Airborne Visible/Infrared Imaging Spectrometer (AVIRIS) over
several mining districts in
This geophysical and mineralogical research was undertaken as a part of the United States Environmental Protection Agency (USEPA) and U.S. Geological Survey (USGS) Utah Abandoned Mine Lands (AML) Imaging Spectroscopy Project (U.S. Environmental Protection Agency and U.S. Geological Survey, 2002). An index map showing the study areas for this project is available on the project Web site (http://speclab.cr.usgs.gov/earth.studies/Utah-1/utahproj_large.jpg). The project had three primary goals:
- to remotely identify
and map natural and
anthropogenic sources of acid generation in and around several historic mining
districts in
- to seek evidence for downstream or downwind
transport of minerals associated with acid generation from mine sites and zones
of unmined, altered rock; and
- to identify and map carbonates and other
acid-buffering minerals that neutralize acidic, potentially metal bearing
solutions, thus mitigating their environmental effects.
Remotely sensed image data were used to screen and evaluate watersheds containing multiple sources of mining-related heavy metals because ground surveys using traditional methods of multimedia sampling and analysis are costly and time consuming. The geologic analysis of spectroscopic image data such as those acquired by AVIRIS enables the detection of specific materials and mixtures of materials on the land surface based on quantitative comparisons of spectral absorption features in the image data to libraries of standard reference spectra of minerals, water, vegetation, and man-made materials. Such detailed mapping allows an evaluation of the critical geochemical regimes and processes in an area, thus providing an objective, scientific means of prioritizing potential environmental hazards for the purposes of streamlining and focusing subsequent field-sampling and remediation programs.
Scientific Background
Mapping the Acid-Drainage Geochemical System
Using Imaging Spectroscopy
The iron sulfide mineral pyrite (FeS2) is a common gangue, or waste, mineral in precious and base metal deposits such as those in the mining districts studied in the Utah AML project. Because pyrite contains sulfur and is commonly unstable in moist, subaerial conditions, the mineral plays a key role in determining future geochemical regimes when it is exposed either by erosion or by mining. During mining operations, broken-up waste rock containing pyrite, carbonate minerals (for example, calcite and dolomite), and (or) phyllosilicate minerals (for example, clays and micas) associated with hydrothermal alteration was commonly dumped near the shafts, adits, and open pits of the mines. At ore-processing mills, slurries of tailings material containing pyrite and other gangue minerals were released into impoundments or directly into drainages. Through time, the pyrite will oxidize in the presence of atmospheric oxygen and water to form sulfuric acid (H2SO4) and various ferric and (or) ferrous iron sulfate-hydrate minerals including copiapite (Fe2+Fe3+4(SO4)6(OH)2·20H2O), and melanterite (Fe2+SO4·7H2O). As a part of the reaction process, thin coatings of sulfate salts such as copiapite may be precipitated on waste-rock surfaces as water from rain events evaporates. With time, most of the pyrite in the waste rock will oxidize, leaving behind coatings of fine-grained jarosite ((K,Na,H3O)Fe3+3(SO4)2(OH)6) that are more stable and less soluble than the hydrated iron sulfate salts that precipitate early in the process. These coatings may in turn break down to the metastable mineral ferrihydrite (approximately 5Fe3+2O3·9H2O), then to the ferric iron hydroxide mineral goethite (α-Fe3+O(OH)), and, with additional time, possibly to the ferric iron oxide mineral hematite (Fe2O3) (Swayze and others, 2000). In natural and anthropogenic exposures of jarosite formed from the oxidation of pyrite, a zoning pattern of iron-bearing minerals is commonly observed that reflects the pH of the waters from which the minerals precipitated (Swayze and others, 2000; Rockwell and others, 1999, 2000). This pattern consists of a central core of unaltered pyrite and (or) copiapite formed under low-pH conditions that grades outward into subconcentric and commonly discontinuous zones of jarosite, jarosite + goethite, goethite, and hematite formed under progressively more neutral pH conditions. The metastable secondary mineral schwertmannite (Fe3+16O16(OH)12(SO4)2) may also form in environments affected by acid drainage from undisturbed rocks, mine waste, and tailings (Ferris and others, 1989; Bigham and others, 1992; Desborough and others, 2000).
The sulfuric acid-bearing solutions generated by the oxidizing reactions can infiltrate downward through the waste-rock pile. The weathering process can also produce clay minerals such as smectites (for example, montmorillonite, (Na,Ca)0.33(Al,Mg)2Si4O10(OH)2·nH2O) and kaolinite (Al2Si2O5(OH)4) as alteration products of feldspars and micas in the waste rock; kaolinite forms under the most acidic conditions. If Ca concentrations in the acidic solutions are sufficiently high, gypsum (CaSO4·2H2O) may precipitate. The acidic solutions can also mobilize heavy metals (lead, cadmium, zinc, arsenic, etc.) present in the waste rock and potentially transport them into ground and surface water. If present in sufficient concentrations, these heavy metals can pose a health hazard, and they have been found to preferentially adsorb onto amorphous iron hydroxide minerals contained within mine waste in near-neutral pH environments where residence times can be long (Bowell, 1994).
Imaging spectroscopy has been used
since the mid 1990s to map surface minerals in abandoned mine lands for the
purposes of environmental site characterization (Farrand and Harsanyi, 1997;
Smith and others, 1998; Swayze and others, 2000; King and others, 2000; Dalton
and others, 2000). As copiapite, jarosite, goethite, and hematite are
characterized by distinct and diagnostic spectral absorption features in the
region of the electromagnetic spectrum measured by AVIRIS and other imaging
spectrometers (fig. 1) (
Figure 1. Reflectance spectra of minerals associated
with acid drainage.
with a high degree of accuracy by using imaging spectroscopy. In contrast, pyrite is difficult to detect through the use of remote spectroscopic mapping techniques because of its low overall albedo, weak (saturated) absorption features, and frequent masking by coatings of secondary iron sulfate minerals. Pyrite in very high concentrations, however, has been successfully identified and mapped by using AVIRIS data: both the Leadville, Colorado, mining district (Swayze and others, 2000) and the Bauer Mill site near Stockton, Utah (this report), have sufficient waste-rock pyrite to be identified and mapped by AVIRIS. In these cases, the mapped pyrite was surrounded by pixels in which jarosite was identified. Jarosite is thus an important indicator of the presence of rocks bearing pyrite, possibly other sulfide minerals such as chalcopyrite (CuFeS2), and (or) other sulfate minerals that are sources for the generation of acidic solutions. Copiapite is also an important indicator of sulfide minerals that can be reliably mapped with imaging spectroscopy data, but is highly soluble and is therefore less common than jarosite. Other highly soluble, sulfate-bearing salts such as alunogen (Al2(SO4)3·17H2O) and epsomite (MgSO4·7H2O) are also detectable with AVIRIS data, although the lack of narrow diagnostic absorption features in the spectra of these minerals makes remote detection less accurate. These soluble salts may be precipitated as thin, temporary crusts on exposed, pyrite-bearing rock after rain events (Cunningham and others, 2005). As the crusts dry, they can change in color (from yellowish orange to gray or white), suggesting possible mineralogic changes that could be identified and monitored by using spectroscopic data. Although ferrihydrite also has diagnostic electronic absorption features conducive to spectral identification and has been found to occur abundantly in acid-mine-drainage environments (Ferris and others, 1989), it has yet to be definitively identified through the use of imaging spectrometer data because of the apparent spectral dominance of jarosite, goethite, and hematite at the pixel scale and (or) the tendency of ferrihydrite to dissolve and reprecipitate as goethite (Bigham and others, 1992). Schwertmannite is stable at the Earth's surface only in low-pH and (or) aqueous environments, yet was spectrally identified in ferricretes associated with acid rock drainage in the Animas River watershed of the San Juan Mountains of Colorado (Dalton and others, 2000; Desborough and others, 2000).
The just-described background indicates that site-characterization plans developed on the basis of maps derived from remote-sensing surveys should focus subsequent field-sampling efforts on areas in which spectra signifying pyrite-, copiapite-, schwertmannite-, or jarosite-bearing mineral assemblages were identified, as it is likely that rocks in these areas contain the highest concentrations of acid-producing sulfide or sulfate minerals. However, it should be noted that goethite coatings on weathered rocks on the surface may mask abundant pyrite occurring in underlying rocks. On the basis of field studies and an understanding of the geochemical regimes discussed above, however, it can be generally assumed that there will be less pyrite at and near the surface in areas where goethite is prevalent than in areas where jarosite is the spectrally dominant mineral and, therefore, that surface runoff from goethite-coated areas will be of more neutral pH than that derived from jarositic areas.
Other Minerals with Acid-Producing Potential in
Unmined Mineralized Areas
Acid-producing minerals also occur on the Earth's surface in the natural environment, usually in rocks that have been altered and mineralized by hydrothermal solutions. Most sulfide and hydrous iron sulfate minerals have some acid-producing potential (APP), and the most commonly occurring of these can be ordered from high to low APP as follows: pyrite, copiapite, schwertmannite, and fine-grained secondary jarosite formed via pyrite oxidation. The APP of such jarosite is discussed by Desborough and others (1999). Alunite ((Na,K)Al3(SO4)2(OH)6) is a common mineral formed by acid-sulfate hydrothermal alteration and is found in abundance in the Tintic and Marysvale districts in Utah, the Silverton and Lake City calderas and the Summitville deposit in Colorado, and in the Goldfield and Cuprite districts in Nevada, among many others. Alunite (Rye and others, 1992) and some spectrally identifiable clay minerals, such as dickite (Al2Si2O5(OH)4) and pyrophyllite (Al2Si4O10(OH)2), commonly occur with pyrite in magmatic hydrothermal acid-sulfate systems. Therefore, although these minerals have never been associated with acid generation themselves, they can also be regarded as indicators of sources of potentially strong acid generation.
Not all jarosite is associated
with the presence of pyrite. Primary hypogene jarosite can form in steam-heated
acid-sulfate hydrothermal systems such as those associated with the Miocene
replacement alunite deposits in the Marysvale volcanic field,
When performing watershed-based environmental evaluations, researchers should consider that ground and surface runoff from unmined, but hydrothermally altered rocks might have a lower pH than runoff from unaltered rocks and that such low-pH runoff could effectively lower the premining pH baseline of a watershed. However, it can be generally assumed that pyrite-bearing waste-rock piles at mine sites and tailings near mill sites will have significantly higher APP than similarly sized exposures of naturally occurring, sulfur-bearing altered rock because (1) the acid-producing minerals in waste-rock piles and tailings are commonly in chemical disequilibrium with atmospheric conditions, (2) the surface area of sulfur-bearing minerals available for oxidation is increased because of the generally small grain size of the blasted or processed waste rock, and (3) the waste-rock piles are highly permeable, allowing for rapid penetration of oxidizing precipitation.
Acid-Buffering Minerals
The carbonate minerals calcite (CaCO3) and dolomite (CaMg(CO3)2) buffer acidic solutions on contact, causing an increase in fluid pH and the precipitation of some dissolved metals. "Free-hydroxyl" minerals such as chlorite ((Mg,Al,Fe)12[(Si,Al)8O20](OH)16) and the sorosilicate mineral epidote (Ca2(Al,Fe3+)3(SiO4)3(OH)) have more limited capacities for acid neutralization. The buffering capacity, or acid-neutralizing potential (ANP), of chlorite has been estimated to be an order of magnitude weaker than that of calcite (Desborough and others, 1998). If rocks bearing these minerals are located downstream from sources of acid generation, they can provide some buffering capacity.
AVIRIS Data Acquisitions, Reflectance Calibration, and Georectification
AVIRIS Data Acquisitions
Spectroscopic image data covering the East Tintic Mountains (fig. 2), the Oquirrh Mountains (fig. 3), and the Tushar Mountains/Marysvale region (fig. 4) of Utah were acquired on August 5, 1998, by the AVIRIS sensor from the high-altitude National Aeronautics and Space Administration (NASA) ER-2 aircraft flying at an altitude of ≈20 km (Vane, 1987). These data have a GIFOV (ground instantaneous field of view), or ground spatial resolution, of ≈17 m/pixel. Nongeorectified quicklook images of the high-altitude AVIRIS flightlines, or "runs," are accessible from the online quicklook index of 1998 data (http://aviris.jpl.nasa.gov/ql/list98.html) on the NASA Jet Propulsion Laboratory (JPL) AVIRIS Web site. The satellite-borne Landsat 7 ETM+ sensor acquired multispectral image data of these areas on October 17, 1999. The Landsat 7 data have a ground resolution of ≈30 m/pixel.
In 1999, a second phase of the
project focused more detailed mapping on intensely mined and (or) mineralized areas
identified by the high-altitude 1998 survey. On October 17-19, 1999, additional
flightlines (17 total) of low-altitude AVIRIS data were acquired over selected
parts of the study areas. These data were acquired from a Twin Otter aircraft
flying at ≈5.33 km altitude and have a ground resolution of 2-3 m/pixel.
Georectified quicklook images of these low-altitude AVIRIS flightlines are
accessible from the online quicklook index of 1999 low-altitude AVIRIS data (http://aviris.jpl.nasa.gov/ql/listla99.html) on
the NASA JPL AVIRIS Web site. Results obtained from the analysis of three of
these flightlines of low-altitude AVIRIS data are presented in this report.
Figure 4 shows the location of the flightline covering the Big Rock Candy
Mountain area near Marysvale, and figure 5 shows the locations of the two
flightlines covering the Tintic mining district. Results from analysis of the
Big Rock Candy Mountain flightline are also discussed by Cunningham and others
(2005).
Figure 2. Location map of East Tintic
Mountains-Cedar Valley region,
Figure 3. Location map of
Figure 4. Location map of Tushar Mountains/Marysvale region,
Figure 5. Location map of low-altitude AVIRIS data coverage over the
Tintic mining district,
Calibration of High-Altitude AVIRIS Data
The high-altitude AVIRIS data are
calibrated to reflectance by using a two-step process (Rockwell and others,
2002; King and others, 2000). In the first step, the data are corrected by
using an algorithm (ATREM, Gao and Goetz, 1990; Gao and others, 1992) that
estimates the amount of atmospheric water vapor in the spectrum of each pixel
independently, as compared with an atmospheric model. The algorithm uses this
information on a pixel-by-pixel basis to reduce the effects of absorptions
caused by atmospheric water vapor. This step also includes characterizing and
removing the effects of Rayleigh and aerosol scattering in the atmosphere (path
radiance) and a correction for the solar spectral response relative to
wavelength. The second step requires the on-site spectral characterization of a
ground-calibration site that is present within the AVIRIS data coverage. Table
1 lists the sites used for ground calibration of the high-altitude AVIRIS data
covering each of the three study areas. The spectra of these field sites are
used to smooth the AVIRIS data by removing residual atmospheric absorptions and
sensor artifacts. AVIRIS spectra smoothed in this way may be directly and
quantitatively compared to libraries of standard reflectance spectra. The
reflectance calibration of the 1998 high-altitude AVIRIS data covering the
Oquirrh and
Table
1. Ground sites used for reflectance calibration of high-altitude AVIRIS data.
Reflectance data derived from the
ground calibrations shown in table 1 contained substantial spectral artifacts
related to either residual absorptions of atmospheric gases and particulates
that were not removed by the ATREM and path-radiance corrections or sensor
noise in the 2.0- to 2.5-μm spectral region. Residual artifacts related to
atmospheric water (mainly 0.94 and 1.13 μm) and CO2 (2.01 and 2.06
μm) may become more pronounced for areas at elevations different from that of
the ground-calibration site. This effect is caused by the fact that the
ground-calibration process corrects the entire AVIRIS data coverage relative to
atmospheric conditions at the calibration site. As absorptions related to CO2
increase in depth with increased atmospheric path length, reflectance spectra
of pixels sampled from high elevations will show smaller CO2
absorption-feature depths than pixels sampled from lower elevations.
Overcorrection for CO2 will occur at elevations higher than the
calibration site, resulting in positive "humps" at the CO2
absorption-feature locations. Conversely, undercorrections for CO2
will occur for areas at lower elevations than the calibration site. The
reflectance data derived from the calibration site at the
To alleviate these deficiencies in the reflectance data, additional areas of known composition located near the average elevations for a study area were used to verify and further refine the accuracy of the calibrations and derive any residual corrections for path radiance. Reflectance spectra of bright (high surface albedo) areas of known composition were sampled from the calibrated high-altitude AVIRIS data and edited, or "polished," to identify and remove artifacts related to residual absorptions of atmospheric gases, particulates, and sensor noise. Corrections for the subtle artifacts identified in this way were incorporated into the data used for the original reflectance calibrations, and the AVIRIS radiance data were recalibrated to reflectance format by using this refined calibration data. Sites used for this secondary reflectance-based spectral polishing are also listed in table 1.
Calibration of
Low-Altitude AVIRIS Data
As no field spectra were obtained
during the low-altitude overflights and the flightlines did not cover the
calibration sites used for the high-altitude data, reflectance-calibrated
high-altitude AVIRIS spectra were used to simulate field spectra for the low-altitude
data calibration. The process of reflectance calibration described in Rockwell
and others (2002) was applied to the low-altitude AVIRIS data with the
exception that edited high-altitude AVIRIS spectra were used as simulated field
spectra of ground-calibration sites. Spectra of areas of bright soil and rock
covered by both the high- and low-altitude AVIRIS data were sampled from the
reflectance-calibrated high-altitude AVIRIS data, averaged, and edited to
remove residual atmospheric absorptions. This "boot-strapping"
procedure of using high-altitude AVIRIS data to calibrate overlapping
flightlines of low-altitude data is further described in Rockwell and others
(1999). For the two flightlines acquired over the Tintic mining district (figs.
6 and 7), a patch of bright soil in the
Figure 6. True-color composite image generated from the low-altitude
AVIRIS flightline over the
Figure 7. True-color composite image generated from the low-altitude
AVIRIS flightline over the East Tintic subdistrict,
Figure 8. False-color composite image generated from the low-altitude
AVIRIS flightline over the Big Rock Candy Mountain area of the Marysvale
volcanic field,
Georectification of AVIRIS Data
The high-altitude AVIRIS data were collected from a NASA ER-2 aircraft flying at an altitude of ≈20 km. Although the ER-2 was designed to simulate conditions on a stable satellite platform and is equipped with a roll-compensation system, geometric distortions related to variations in aircraft roll, pitch, yaw, and velocity are present in the AVIRIS data. These distortions must be removed prior to image georeferencing to a map projection if positional errors are to be minimized, especially in areas of significant terrain relief. The USGS AVRECGEN and AVRECTFY algorithms were used to remove these distortions; the algorithms are based on modeling the look-point equation for each AVIRIS pixel using the engineering and navigation data that are recorded simultaneously with the spectral image data (Clark and others, 1998).
The low-altitude AVIRIS data described here were acquired from a propeller-driven Twin Otter aircraft flying at 5.33-km altitude. The distortions caused by roll, pitch, yaw, and velocity variations are much more pronounced in data acquired by the low-altitude platform than in data from the ER-2. Therefore, a different algorithm was used to remove these distortions from the low-altitude data (Boardman, 1999). This method does not remove topography-induced image distortions, but does remove the aircraft-induced and scan mirror-induced distortions that dominate AVIRIS low-altitude data.
After distortion removal, the high-altitude data were georeferenced to the Universal Transverse Mercator map projection by using a second-order polynomial transformation with control points selected from USGS 1:24,000-scale Digital Raster Graphics. The low-altitude data were georeferenced by using rubber-sheeting functions (Watson, 1992).
Spectral Analysis
The USGS Tetracorder expert system
was used for spectral analysis of the AVIRIS data (Clark, Swayze, Livo, and
others, 2003). This semiautomated software system independently compared the
spectrum of each pixel in the AVIRIS data to a digital library of standard
laboratory reference spectra of minerals, mineral mixtures, water, snow,
man-made objects, and vegetation. The library reference spectra used by the
Tetracorder software are available in published spectral libraries (Rockwell,
2002; Clark, Swayze, Wise, and others, 2003). One or more diagnostic spectral
absorption features were analyzed according to a detailed set of rules for each
reference material. This analysis generated quantitative digital image maps of
(1) absorption-feature depth in the image spectra and (2) modified
least-squares fit of image spectra to library reference spectra across defined
spectral intervals (continua) for each reference material. In general,
absorption-feature depth is proportional to the spectral abundance of a
material in a pixel, given a constant grain size (
The spectrum of each pixel of AVIRIS data was analyzed separately for several groups of surface materials. These can be detected independently of each other because they have diagnostic absorption features in different wavelength regions of the electromagnetic spectrum. For every pixel, modified least-squares fit values were generated for each reference material belonging to a particular material group. The material with the highest fit value for that group was selected as the spectrally identified material within that group. The reliability of the mapping results is directly proportional to both high feature depths in the image spectra and high degrees of fit. Therefore, the image maps showing feature depth and feature fit are multiplied to generate a "fit x depth" image for each identified material (Clark, Swayze, Livo, and others, 2003). Pixels with high fit x depth values are most likely to be an accurate identification of a given material. Pixels not identified as a particular material in a group (that is, their fit and (or) depth values were below a user-defined threshold) were assigned a fit x depth value of zero for that group. Therefore, for each group of surface materials (for example, the iron-bearing mineral group or the clay, sulfate, mica, carbonate, and hydrous silica mineral group), a given pixel may have a positive value (representing an identification) for only one material, or it may not be identified as any material in that group. The fit x depth image map is used for the final interactive analysis of the mapping results. The Tetracorder system identifies only the material or mixture of materials that is spectrally dominant in a pixel, meaning that the absorption feature of the identified material is sufficiently unobscured by features of other materials to allow its recognition by spectral analysis. Therefore, identification of the spectrally dominant material in a spectrum does not imply that other materials do not also exist in that pixel.
A separate map can be generated for each material group. For this report, two types of maps were generated to show the distribution of the following materials: (1) those having absorption features in the 0.35- to 1.35-μm spectral region, such as iron-bearing minerals, snow, ice, and water; and (2) those having vibrational absorption features in the 1.45- to 2.50-μm spectral region, including such minerals as phyllosilicates (micas and clays bearing Al-OH or Mg-OH), sulfates, carbonates, amphiboles, hydrous quartz (chalcedony and opal bearing Si-OH bonds), and epidote (a sorosilicate mineral bearing calcium and Al-OH and (or) Fe-OH bonds). The AVIRIS-derived maps of surficial materials presented in this report consist of color-coded pixels identified as specific materials on a grayscale background image of a single AVIRIS band. In generating the final maps, each material is assigned a discrete color. The fit x depth image corresponding to a particular material may be digitally stretched so that pixels of all fit x depth values will be represented by a single color ("hard stretch"), or the image can be stretched so that the fit x depth values will be represented by a range of brightness levels for a given color ("continuous stretch"). For example, in the case of a continuous stretch, pixels with the highest fit x depth values will be represented by the color chosen for that material, and pixels with decreasing fit x depth values will be represented by successively darker shades of that color. Hard stretches are used more frequently than continuous stretches in making maps showing many different materials, as maps showing many shades of colors can be difficult to interpret. Minerals and mineral assemblages for which continuous stretches were used are marked with a "C" in the map explanations (legends).
Map Explanations
The explanations (legends) with the high-altitude AVIRIS mineral maps presented here have been designed to facilitate interpretation of the imagery. These explanations relate identified minerals and mineral assemblages to associated acid-producing potential (APP) and acid-neutralizing potential (ANP). The explanations for the maps of iron-bearing minerals are organized in order of decreasing APP from top (high APP) to bottom (low APP). APP can be considered to be inversely proportional to pH. In the explanations for the maps of clay, sulfate, mica, and carbonate minerals, minerals and mineral assemblages that either may occur with pyrite (for example, dickite) or have APP themselves (for example, jarosite) are indicated with an asterisk.
Verification of Spectral Analysis Results
The results of the mineral mapping were verified by field checking and (or) interactive comparison of AVIRIS spectra with standard library spectra. Selected mapping results were also verified by using X-ray diffraction (XRD) analysis of field samples. The appendix shows the XRD results of many field samples, along with sampling locations and other information. Appendix tables A1 and A3 include the Tetracorder mapping results for AVIRIS pixels in the vicinity of the sample collection locations. Tetracorder mapping results show several different minerals for a given location, meaning that either (1) mineral mixtures were directly identified in the AVIRIS data or (2) various individual minerals were spectrally identified in the area surrounding the location and the exact AVIRIS pixel corresponding to the sampling location could not be reliably identified. In cases where field checking and (or) laboratory analysis identified errors in the Tetracorder mapping, mapping rules were reviewed and modified and (or) new standards were added to the spectral library of reference materials. In the latter case, rock samples collected in the field were analyzed by XRD, and their reflectance spectra were measured in the laboratory. Mapping rules were then developed for one or more diagnostic absorption features present in the laboratory spectra, and the spectra were added to the spectral reference library. The Tetracorder expert system was then rerun on the AVIRIS data by using the modified mapping rules and expanded spectral library. Some of the field samples that were added to the spectral library are listed in blue in the appendix. Rockwell (2002) has documented the spectroscopic properties of these samples and has defined the absorption features in each sample spectrum that were analyzed by the Tetracorder expert system.
To exemplify a Tetracorder mapping
error that was remedied as a part of this research, maps of the southwestern
Because (1) the Tetracorder expert system is experimental and under constant development and revision and (2) the Earth's surface has inherent mineralogic complexity, 100 percent accuracy of mineral identifications cannot be guaranteed for each AVIRIS pixel. For many of the more common rock-forming minerals, the Tetracorder system is very robust, especially when the minerals occur abundantly in pure or nearly pure form. Although a great effort has been made to include many common mineral mixtures in the Tetracorder spectral library, it is not currently possible to include spectra of every combination of minerals. For this reason, and because Tetracorder only identifies the mineral or minerals that are spectrally dominant in a pixel, it can be assumed that the accuracy of the mineral maps will decrease as the number of constituent minerals in a rock increases. No legal or regulatory actions should be initiated on the basis of the mineral maps alone. Targets of potential significance identified by the mineral maps should be studied in detail in the field and (or) laboratory prior to decision-making regarding a site or watershed.
The East Tintic Mountains and the Tintic Mining District
Geologic Setting
The Tintic mining district is
located ≈95 km south-southwest of Salt Lake City, Utah, in the East Tintic
Mountains (see project index map at http://speclab.cr.usgs.gov/earth.studies/Utah-1/utahproj_large.jpg).
The town of
Figure 9. Geologic map of the Tintic mining district. Refer to figure 10 for explanation showing geologic units. Blue = Swansea Quartz Monzonite (Ts) stock and related intrusive rocks. Green = Sunrise Peak Monzonite Porphyry (Tsp) stock and related intrusive rocks (contemporaneous Gough and Dry Ridge sills, Tsps, are in black and white dot pattern). Red = Silver City Monzonite (Tsc) stock and related plutons. Modified from Morris and Mogensen (1978). View full-resolution file
Figure 10. Explanation for geologic map shown in figure 9. From Morris and Mogensen
(1978).
The second volcanic event resulted
in the deposition of the Tintic Mountain Volcanic Group and culminated with the
intrusion of the Sunrise Peak Monzonite Porphyry stock and many related plugs,
dikes, and extensive latite sills at the southern and eastern edges of the
district (fig. 9). Near the end of Oligocene volcanism (≈31.5 Ma), the Silver
City Monzonite stock and related dikes and plugs were intruded along a
north-northeast-trending zone extending from the northwest boundary of the
caldera; this igneous activity initiated circulation of hydrothermal fluids
through faults, fractures, and breccia zones in the Paleozoic sedimentary rocks
and older Oligocene volcanic rocks. These hot fluids pervasively altered the
country rock and ultimately resulted in deposition of ore minerals. The
youngest intrusive rock in the district is the quartz monzonite porphyry of Diamond
Gulch, which intruded the southern part of the
Ore
Lovering (1949) recognized the sequence of events involving hydrothermal alteration and mineralization in the Tintic district: (1) pervasive dolomitization of limestone beds, (2) propylitic alteration, (3) argillic alteration (including formation of alunite, kaolinite, etc.), (4) silicification, calcification, and pyritization (post-monzonite intrusion), and (5) ore deposition (quartz, barite, sericite, orthoclase, rhodochrosite, and ore minerals). Late-stage, high-temperature (≈257°-300°C) fluids created epigenetic polymetallic base and precious metal deposits as replacements in favorable carbonate beds, replacement veins, and fissure veins (Morris, 1990). Primary ore minerals include galena, sphalerite, argentite, tetrahedrite-tennantite, enargite, sulfosalts, native gold and silver, and secondary oxides. Alteration minerals include alunite, various clay and carbonate minerals, illite (sericite), and pyrite. Secondary gypsum and jarosite (after pyrite) are also present. Primary gangue (waste) minerals include quartz, barite, calcite, dolomite, and rhodochrosite. Table 2 lists the dominant gangue minerals as a function of ore type in the district.
Table 2. Dominant ore and gangue minerals and production figures, Tintic mining district.
[Modified
from Morris (1968) and Cox and Singer (1986); production data from Morris and
Mogensen (1978)]
Most ore deposits in the Tintic
district occur as replacement bodies, replacement veins, and fissure veins
(Morris, 1968). A majority of the metals produced from the Tintic district were
derived from ore bodies that have replaced favorable horizons in folded and
faulted Paleozoic carbonate rocks. Figure 11 is a map of the Tintic district
showing the principal mines and plan views of the major underground ore bodies
in the district. Metal production revealed strong patterns of horizontal
zonation across the district. Although lead and silver ores were common
throughout the district, figure 12 shows that zinc was mainly produced from the
northernmost sections of the main replacement ore zones, whereas gold (not
shown) and copper were common mainly in the southern part of the district in
the area surrounding the Silver City Monzonite stock. The central part of the
district (located between the lines showing copper and zinc occurrence limits
in fig. 12) was known chiefly for lead and silver production. In general, Pb/Zn
and Ag/Pb ratios decrease toward the north in the district.
Figure 11. Map of Main Tintic subdistrict
showing mines and plan views of ore bodies. For description of map units, see figure 10. From Morris (1968). View full-resolution file
Figure 12. Map of replacement
ore bodies, showing generalized compositional zonation. From Morris (1968).
Drilling at the southwestern edge of the Tintic district has identified the Southwest Tintic (SWT) porphyry copper deposit (fig. 9) (Krahulec, 1996). This deposit, which had not been mined as of summer 2004, has been characterized as a high-sulfide, low-copper porphyry system associated with the quartz monzonite porphyry stock of Diamond Gulch. The deposit is associated with intense and strongly zoned hydrothermal alteration, some of which is exposed at the surface. The richest copper grades are found in stockworks of quartz ± pyrite ± chalcopyrite ± magnetite ± molybdenite veins that are largely restricted to a biotite-rich zone at the core of the deposit. A shallow supergene chalcocite blanket is present beneath alluvium in Diamond Gulch 1-3 km to the southwest of Horseshoe Hill (fig. 2). Phyllic, or quartz-sericite-pyrite (QSP), alteration surrounds the core and is exposed in the vicinities of Horseshoe Hill, Treasure Hill, and Ruby Hollow. Within the QSP zone, clay minerals increase in abundance relative to quartz with increased distance from the potassic core of the deposit. A propylitic envelope consisting of an inner zone of actinolite and epidote and an outer zone of calcite and chlorite surrounds the QSP zone. The alteration zones are highly elongated along a northeast-trending structure, which has been interpreted as a tear fault formed during the Jurassic.
Mining History
The Tintic district was discovered
in 1869, and production of rich polymetallic ores steadily increased, peaking
in 1921. Production declined from 1921 until the mid 1950s, when new
discoveries in the East Tintic subdistrict prompted another burst of mining
that lasted until the 1990s. Metal production figures from 1869 to 1976 are
shown in table 2. Virtually no metallic ores have been produced from the Main
Tintic subdistrict since 1960. Natural caverns formed by dissolution of
carbonate rocks were used for mine dewatering, most notably in the Gemini and
Chief No. 1 mines adjacent to
The most recent mining activity in the Eureka/Tintic area has taken place in the East Tintic subdistrict. The main ore body of the Burgin mine was discovered in 1958 and was mined for Pb, Zn, Ag, minor Au and Cu, and high-silica flux ores until the late 1970s. The Trixie mine, located 2.5 km southwest of the Burgin, was discovered in the mid 1950s and mined for gold, copper, and silver until the 1990s (Morris, 1990). New gold discoveries were made in the late 1990s in the vicinity of the Trixie mine.
Because of elevated levels of Pb,
As, Sb, Cd, Hg, Ag, and other metals in the soil, parts of the town of
Mapping and Characterization of Mine Waste in the
Tintic Mining District
Maps of minerals and water were
generated from the high- and low-altitude AVIRIS data (figs. 13-18). Minerals
were identified in only a small percentage of processed AVIRIS pixels. Most of
the region within and surrounding the
Table 3 lists mines in the Tintic
district at which jarosite-bearing rocks were mapped by using AVIRIS data. For
the locations of these mines, refer to figures 11 and 13-19. The largest
exposures of jarosite-bearing tailings and waste-rock piles are marked with an
asterisk in table 3. For example, figure 20 shows the size of the waste-rock
piles at the
Table
3. Mine sites at which
jarosite-bearing waste rock was mapped by using AVIRIS data.
[Asterisks indicate the largest mapped
exposures of acid-generating minerals; mine locations shown in figs. 13-19]
Figure 13. Map of iron-bearing minerals and water in the
Figure
14. Map of clay,
carbonate, sulfate, and mica minerals in the
Figure 15. Map of iron-bearing minerals and water in the Silver City-Dragon
mine area, Main Tintic subdistrict,
Figure 16. Map of clay, carbonate, sulfate, and mica minerals in the Silver
City-Dragon mine area, Main Tintic subdistrict,
