Clark, R.N., G.A. Swayze, and A. Gallagher, Mapping Minerals with Imaging Spectroscopy, U.S. Geological Survey, Office of Mineral Resources Bulletin 2039, pp. 141-150, 1993.
Roger N. Clark, Gregg A. Swayze, and Andrea Gallagher(1)
Imaging spectroscopy is a new mapping tool and the next generation in remote sensing technology. The narrow spectral channels of an imaging spectrometer form a continuous reflectance spectrum of the Earth's surface, which contrasts with the 4 to 7 channels of the previous generation of imaging instruments, like the Landsat Thematic Mapper (TM) and Multispectral Scanner (MSS) instruments. While systems like Landsat can distinguish general brightness and slope differences in the reflectance spectrum of the surface, imaging spectroscopy not only does that, but also resolves absorption bands in the spectrum which can be used to identify specific species. Spectroscopic analysis of imaging spectroscopy data allows any material (mineral, vegetation, man-made, water, snow, etc.) with unique absorption features in the measured spectral region to be mapped.
NASA is now flying the "Airborne Visual and Infra-Red Imaging Spectrometer" (AVIRIS) instrument. AVIRIS acquires data in the spectral range from 0.4 to 2.45 microns in 224 spectral channels. The instrument is flown in an ER-2 aircraft (a modified U-2 spy plane) at 19,800 meters (65,000 feet). The ground resolution is 20 meters, the swath width about 11 kilometers (614 pixels) and the swath length can be up to about 1000 kilometers. After initial poor performance in 1987 and 1989, the AVIRIS instrument now produces superb signal-to-noise data. (Editor's note: since this paper was published, AVIRIS has continued to improve each year, and is now a spectacular instriment. See Evolution in Imaging Spectrososcopy Analysis and Sensor Signal-to-Noise: An Examination of How Far We Have Come
In 1989 we developed a new analysis algorithm that uses a digital spectral library of known materials and a fast, modified-least-squares method of determining if a single spectral feature for a given material is present (Clark et al., 1990). We have made a major advance in the mapping algorithm: now multiple minerals using multiple spectral features are mapped simultaneously. This is done by a modified-least-squares fit of spectral features from data in our digital spectral library to corresponding spectral features in the image data. The algorithm does not force a detection like many other algorithms in use. For example, many algorithms take a set of curves and best fit them to the observed data, often requiring a set of parameters (like mineral fraction) to sum to one. Our algorithm only produces values indicating the presence of those minerals we choose to map. If the minerals do not exist in that area, the algorithm produces zeros, indicating they are not detected.
Our mapping algorithm produces for each pixel in the image, a spectral feature depth (correlated to abundance) and a fit number (least-squares correlation coefficient) to the reference spectra (giving a measure of confidence in the result) for each mineral mapped. The depth values for each pixel and each mineral form a set of images of the minerals correlating to abundance. The fit values form a set of images corresponding to the confidence level of the identification. We combine single mineral fit and depth images from several minerals, assigning a color to each mineral map. In this way we produce multi-mineral maps. For example, red might be assigned to hematite, where shades of red (from brighter red indicating a stronger spectral signature. In these maps of minerals, black indicates none of the given set of minerals were detected at that location.
We have used the algorithm on AVIRIS data of Cuprite, Nevada to illustrate some of the mapping possibilities with the new generation of sensors. The geologic and alteration maps are shown in Figures 1 and 2. A false color image of Cuprite (like one that might be produced by broad-band remote sensing instruments) is shown in Image A. Example minerals maps are shown for iron bearing minerals (Image B). A color mineral map of clays and sulfates is shown in Image C.
(higher resolution 58K GIF)
Geologic map of the Cuprite, Nevada mining district. The map was
produced by conventional field work combined with remote sensing TM
data. From Abrams and Ashley (1980) and Ashley and Evarts (1976) as
modified by Hook (1990).
Figure 2: 26K GIF.
(higher resolution 38K GIF)
Alteration map of the Cuprite, Nevada mining district.
The map was produced by conventional field work combined with remote
sensing TM data.
From Abrams and Ashley (1980) and Ashley and Evarts (1976) as modified
by Hook (1990).
IMAGE A: 320K GIF.
A color infrared image of Cuprite, Nevada from AVIRIS data. North is up in this 11 km wide by 14 km long scene. Ground resolution is 20 meters. The linear feature running north-south to the right of center in the image is Highway 95.
The area to the east of the road is the well-studied Cuprite mining district. This area consists of hydrothermally-altered volcanic rocks and contains an intensely altered central silica cap surrounded by less altered zones of opalized and argillized rock. The area west of the highway consists of altered volcanic rocks, Cambrian siltstones and limestones. It also contains silicified, opalized, and argillized zones. Altered siltstones comprise most of the altered rocks in the northern part of the Cuprite Hills with limestones further south.
(Editor's note: this image is actually from AVIRIS Cuprite 1993 data, so is slightly
to the east of the 1990 data set. The playa at east center is off the image in the
We can detect very subtle spectral differences, like degrees of kaolinite crystallinity (see Image D), the difference between Na-montmorillonite versus Ca-Montmorillonite (see Image C), and individual members of the Na-K alunite solid solution series (Image E1-E2). At Cuprite, maps of these subtle differences depict the alteration zone remarkably well. For example, poorly crystalline kaolinite or halloysite seems to be a general weathering product that occurs throughout the image, but as one moves closer to alteration zones, the kaolinite becomes progressively more crystalline. The highly crystalline kaolinite occurs just outside of the alunite zones. Alunite zones can be subdivided spectrally into areas where K-alunite occurs at the center and is surrounded by Na-alunite (see Image E1-E2).
These mineral maps have the potential to be extraordinary tools for mineral exploration because they show variations in mineral chemistry, and hence, pressure, temperature, and chemical gradients in areal detail never seen before. Coupled with further geochemical research, detection of spectral variations in mineral solid solution series may provide a means to map temperature gradients on a large scale in a matter of hours. This information can also be used to locate areas where critical relationships need investigation. The applications seem boundless.
We have developed our methodology of calibration and mapping analysis to be a near routine method. Imaging spectroscopy could now be used in USGS projects, and we believe the results illustrated here show that imaging spectroscopy could greatly enhance USGS mineral mapping. We also feel the method could be used for environmental problems as the spectral mapping algorithm will work on any material having diagnostic spectral absorption or emission features. Imaging spectroscopy could also be used in laboratory analysis of hand samples, or in the field for investigating small areas.
IMAGE B: 381K GIF.
A color mineral map of iron-bearing minerals for Cuprite, Nevada.
Hematite is RED and is restricted to the opalized and argillized volcanic rocks on both sides of the highway.
Goethite is GREEN and is probably a weathering product being a common soil component.
Jarosite is BLUE and occurs in the opalized zones on both sides of
the highway with the largest occurrences west of the highway.
IMAGE C: 357K GIF.
A color mineral map of clays and sulfates for Cuprite, Nevada.
Alunite is RED and occurs in opalized and argillic zones on both sides of the highway.
Dickite is ORANGE and is closely associated with kaolinite in the altered zones.
Kaolinite (well crystalline) is YELLOW (see Fig. 3D).
Kaolinite (medium crystalline) is YELLOW GREEN.
Kaolinite (poorly crystalline) is GREEN. (Halloysite and poorly crystalline kaolinite are spectrally indistinguishable at 2.2 microns.)
Ca-montmorillonite is LIGHT BLUE and occurs in the northeastern portion of the scene.
Na-montmorillonite is BLUE and occurs in rock units and as loess accumulations on alluvial fans and in playas. Some muscovites have a similar spectral signature and are also mapped as montmorillonite (with further research these minerals may be separated).
Buddingtonite is PURPLE, and is located only in a few pixels east of the highway, at and near "buddingtonite bump." On the HP paintjet image, buddingtonite bump can be found 34mm from the right edge and 118mm from the top edge of the image. Buddingtonite bump can be better seen in white on the expanded IMAGE E2.
Paragonite is MAGENTA, occurring mostly in the lower left center of the image. Chlorite occurs as an intimate mixture with the paragonite.
Opalized tuff is WHITE and occurs in the lower left corner as well
as in the central region of the Cuprite alteration zone.
IMAGE D: 76K GIF.
A color mineral map of KAOLINITE CRYSTALLINITY for Cuprite, Nevada.
Well crystalline kaolinite is BLUE and occurs near the outer margins of the alunite areas (which are black in this image).
Medium crystalline kaolinite is GREEN and occurs between the highly crystalline and poorly crystalline kaolinite.
Poorly crystalline kaolinite is RED and occurs furthest from the alunite
areas. Poorly crystalline kaolinite or halloysite may be due to
weathering processes. (Halloysite and poorly crystalline kaolinite are
spectrally indistinguishable at 2.2 microns.)
IMAGE E1-E2: 96K GIF.
(Editor's note: the printed article had E2 as a small section of E1 enlarged. That is not necessary with GIF images, so only one is shown.)
A color mineral map of the alunite-natroalunite solid solution series, ammonium minerals and two clays for Cuprite, Nevada.
Potassium alunite is GREEN and occurs in opalized and argillic zones on both sides of the highway.
Natroalunite (Na mole fraction 0.65) is MAGENTA and occurs next to potassium alunite, mostly to the west of the highway.
Natroalunite (Na mole fraction 0.80) is BLUE and occurs with other natroalunites, mostly to the west of the highway.
Buddingtonite is WHITE, and is located in a few pixels east of the highway, at and near "buddingtonite bump," and in a few small areas west of the highway.
Ammonium Illite-Smectite is YELLOW, and occurs in the alteration zones near alunite and buddingtonite.
Kaolinite (well crystalline) is RED and occurs near alunites in the alteration zones.
Na-montmorillonite is CYAN and occurs in rock units and loess
accumulations on alluvial fans and in playas.
Abrams, M.J., and Ashley, R.P., 1980, Alteration mapping using multispectral images - Cuprite Mining District, Esmeralda County, Nevada: U.S. Geological Survey Open File Report 80-367.
Ashley, R.P., and Evarts, R.C., 1976, Geologic Map and Alteration Map, Cuprite Mining District, Nevada: Unpublished.
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, p. 176-186.
Hook, S.J., 1990, The combined use of multispectral remotely sensed data from the short wave infrared (SWIR) and thermal infrared (TIR) for lithological mapping and mineral exploration: Fifth Australasian Remote Sensing Conference, Proceedings, Oct., 1990, vol.1, p. 371-380.
U.S. Geological Survey,
a bureau of the U.S. Department of the Interior
This page URL= http://speclab.cr.usgs.gov/PAPERS/cuprite.clark.93/mineral_map.html
This page is maintained by: Dr. Roger N. Clark email@example.com
Last modified November 18, 1998.