See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/229383008

Hydrothermal Alteration Mapping at Bodie, California, using AVIRIS

Hyperspectral Data

Article  in  Remote Sensing of Environment · September 1998

DOI: 10.1016/S0034-4257(98)00040-6

CITATIONS READS
180 393

3 authors, including:

Alvaro Penteado Crósta

University of Campinas
163 PUBLICATIONS   2,103 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Estudo geofísico das estruturas de impacto de Vista Alegre (PR) e Vargeão (SC) View project

Estudo Geofísico da Estrutura de Cerro do Jarau View project

All content following this page was uploaded by Alvaro Penteado Crósta on 29 May 2019.

The user has requested enhancement of the downloaded file.

Hydrothermal Alteration Mapping at Bodie,
California, Using AVIRIS Hyperspectral Data
Alvaro P. Crósta,* Charles Sabine,† and James V. Taranik†
A VIRIS data covering the Bodie and neighboring Par-
amount mining districts, eastern California, were used to
map hydrothermal alteration minerals. Two spectral anal-
ysis algorithms were used: Spectral Angle Mapper (SAM)
and Tricorder. The objective was to compare the perfor-
mance of the two algorithms where no a priori ground or
atmospheric information is available. AVIRIS data were
calibrated to apparent surface reflectance using a modi-
fied MODTRAN radiative transfer model in conjunction
with laboratory and in-flight calibration data. Both algo-
rithms appear to produce satisfactory results for geologic
reconnaissance and mapping applications, but Tricor-
der generally classified more pixels and identified more
mineral species than SAM. For some minerals, such as
Na-montmorillonite, results from Tricorder and SAM
matched reasonably well, but major differences appeared
in how the two algorithms classified kaolinite group min-
erals and ferric oxides. Laboratory spectra of rock sam-
ples from five localities were compared to reference spec-
tra of minerals identified by the classification algorithms
for corresponding pixels, with the results matching to
some extent. Elsevier Science Inc., 1998
INTRODUCTION
Hyperspectral remote sensing, or imaging spectrometry,
has been under development since the first experimental
sensor of this type, the Airborne Imaging Spectrometer
(AIS), was first flown in 1983. Its successor, the Airborne
Visible/Infrared Imaging Spectrometer (AVIRIS) was de-
* Geosciences Institute, State University of Campinas, Campinas,
SP, Brazil
† Desert Research Institute, Reno, Nevada
Address Correspondence to Alvaro P. Crósta, Geosciences Inst.
State Univ. Of Campinas, P.O. Box 6152, 13081-970 Campinas SP,
Brazil. E-Mail: alvaro@ige.unicamp.br
Received 3 April 1996; revised 16 January 1998.
REMOTE SENS. ENVIRON. 65:309–319 (1998)
Elsevier Science Inc., 1998
655 Avenue of the Americas, New York, NY 10010
veloped at the Jet Propulsion Laboratory (JPL) in 1987.
Data acquired by these sensors made possible a sig-
nificant number of geological applications. Among these
applications is the discrimination of rocks and minerals
important to mineral exploration, in particular, map-
ping hydrothermal alteration minerals. The lack of rou-
tine availability of operational hyperspectral sensors and
of data covering diverse geological environments in dif-
ferent regions has, however, restricted the full utilization
of this type of data. The analysis of hyperspectral data
also required the development of new image processing
algorithms specifically designed to take advantage of the
high spectral resolution and to cope with the much larger
amount of data than those used with conventional mul-
tispectral data. Some of these new algorithms have re-
cently been incorporated into commercial image pro-
cessing software.
The next step towards the operational use of hyper-
spectral data involves data availability on a global basis.
In 1994, NASA awarded TRW a contract to build the
Hyperspectral Imager (HSI). The HSI is a spaceborne
imaging spectrometer with 384 spectral bands and 30 m
spatial resolution, complemented by a panchromatic sen-
sor with 5 m resolution.
These developments, summarized by Taranik and
Crósta (1996), indicate a clear trend toward operational
availability of hyperspectral data for geological remote
sensing in the near future.
This article examines and compares two different al-
gorithms, the Spectral Angle Mapper (SAM) and Tri-
corder, developed for analyzing surface materials in im-
aging spectrometry data. The objective was to assess the
performance of these algorithms and their ability to map
alteration minerals important to precious metals explora-
tion. Unlike previous mineral mapping studies, such as
those done at well-known localities like Cuprite, Nevada,
we tried to simulate a situation often encountered in
mineral exploration, in which remote sensing is used as
the first tool for defining targets before the geologist sets
0034-4257/98/$19.00
PII S0034-4257(98)00040-6
310 Crósta et al.
foot on the ground. Therefore, techniques that require
no a priori input of ground data were assessed in their
ability to convert image data to ground reflectance and
for establishing the spectral similarity between image
pixels and reference minerals.
The area selected comprises the Bodie mining dis-
trict and the southern part of the neighboring Paramount
district located in Mono County, California, USA (Fig. 1).
AVIRIS data were acquired on 20 August 1992, at 11:22
a.m. (local time) by NASA’s ER-2 aircraft. The scene used
in this study consists of 614⫻512 pixels (10.44 km⫻9.70
km) in 224 bands at 16-bit radiometric resolution.
GEOLOGY AND MINERALIZATION OF THE
BODIE AND PARAMOUNT DISTRICTS
Bodie was an important gold-silver mining district in the
second half of the 19th century, whereas only a mercury
mine was exploited at the Paramount district. Miocene
(7.8–9.4 Ma) intermediate to basic volcanic rocks (daci-
tic, andesitic, and rhyolitic lavas, plugs, tuffs, and brec-
cias) underlie both districts. Hydrothermal alteration as-
sociated with precious metal concentration at Bodie is
probably the result of intense thermal spring activity,
which still occur along fault zones in this region (Silber-
man et al., 1995).
Figure 1. Location map of the
Bodie/Paramount study area. The
rectangle around the town of Bo-
die shows the area covered by the
AVIRIS scene. The locations of
Bodie Bluff (BB), Silver Hill (SH)
and Paramount (PA) are shown.
Bodie has characteristics of a bonanza-type lode
quartz vein deposit and a large-scale stockwork system.
Gold occurs in quartz veins and quartz–adularia–
chalcedony stockworks in the center of the district. Pet-
rographic work by Herrera et al. (1993) showed vertical
and lateral zoning in the alteration with silicification in
the center surrounded by zones of potassic, argillic, and
sericitic alteration and an outer zone of propylitic alter-
ation. Little information is available for the Paramount
district due to the lack of important mineralization, but
the district has recently been prospected for bulk ton-
nage precious metal mineralization (Herrera et al., 1993).
Geological characteristics observed at Paramount are very
similar to those in Bodie, and the origin of the alteration
is believed to be similar (Silberman et al., 1995).
DATA CALIBRATION TO
SURFACE REFLECTANCE
Identification of surface materials using hyperspectral
data is based on quantitative comparisons between pixel
and reference spectra. In order to make such compari-
sons, imaging spectrometer data must be processed to
retrieve ground reflectance values from radiance mea-
surements. A number of methods have been proposed
for such conversion, some of which require the input of

ground data collected simultaneously with the sensor’s
overflight. Radiative transfer methods rely on the use of
atmospheric models and allow the retrieval of apparent
ground reflectance from the radiance values provided by
AVIRIS. A comparison of several methods by Clark et
al. (1995) showed that a hybrid method, which combines
simultaneous ground-based data and the radiative trans-
fer method, provides the best results for reflectance re-
trieval in hyperspectral data, followed by the sole use of
the radiative transfer method developed by Green et
al. (1993a).
We used the latter method, which uses AVIRIS lab-
oratory-calibrated radiance data (Chrien et al., 1990), in
conjunction with in-flight calibration data obtained over
Rogers Dry Lake at the beginning of the 1992 AVIRIS
data acquisition season (Green et al., 1993b), and the
MODTRAN radiative transfer code. The method com-
pensates for AVIRIS-derived estimates of water vapor,
aerosol, and surface on a pixel by pixel basis. Compared
to other radiative transfer methods, it provides a better
correction for H2O and other atmospheric gases, as a
function of elevation throughout a scene (Clark et al.,
1995).
IMAGE PROCESSING METHODS FOR
MINERAL MAPPING
In this study we compared two different methods for
mineral mapping at Bodie and Paramount: Spectral
Angle Mapper (Kruse et al., 1993) and Tricorder (Clark
et al., 1990; Clark and Swayze, 1995). Both methods
compare spectra from pixels in the scene and reference
spectra from a spectral library, but use different algo-
rithms to compare and measure spectral similarity. The
two methods are relatively insensitive to illumination dif-
ferences due to topography in areas of low to moderate
relief and high sun angle. The library used for this as-
sessment was the public domain USGS Digital Spectral
Library which contains nearly 500 reference spectra of
minerals, vegetation, and other surface materials (Clark
et al., 1993).
Spectral Angle Mapper Classifier
The Spectral Angle Mapper (SAM) is a supervised classi-
fication technique that measures the similarity of image
spectra to reference spectra from a spectral library, or
from field and laboratory spectra. SAM measures similar-
ity by calculating the angle between the two spectra,
treating them as vectors in n-dimensional space, with n
being the number of bands (Kruse et al., 1993). The
angle is the arc-cosine of the dot product of the two
spectra. Smaller values for the angle indicate higher sim-
ilarity between pixel and reference spectra. This method
is not affected by gain (solar illumination) factors, since
the angle between two vectors is independent of the vec-
Mapping Hydrothermal Alteration Minerals 311
tors’ length. SAM produces an image with the pixels it
manages to classify assigned to the respective reference
minerals, constrained by a user-specified threshold, to-
gether with a set of “rule images,” one for each reference
mineral used. DN values in these “rule images” are the
expression of the angle itself (smaller DNs indicate
greater similarity to the reference mineral). SAM, on the
other hand, is not well suited in determining small spec-
tral differences between mineral species and is highly
dependent on the wavelength ranges and on the thresh-
olds selected, which are arbitrary.
We used the version of SAM in the ENVI image pro-
cessing software package (Research Systems, Inc., 1995).
USGS Tricorder Algorithm
Tricorder is still under development at the Spectroscopy
Laboratory of the U.S. Geological Survey. It was de-
signed to compare spectra of materials from the USGS
Digital Spectral Library to image spectra acquired by
hyperspectral sensors, analyzing simultaneously for multi-
ple minerals, using multiple diagnostic spectral features
for each mineral (Clark and Swayze, 1995).
Tricorder works by first removing a continuum from
spectral features in the reference library spectra and also
from each spectrum in the image data set. Continuum-
removed library and image spectra are then compared
using a least square procedure. The algorithm uses a lin-
ear correlation coefficient to determine the fit by individ-
ually fitting all diagnostic absorption features of all refer-
ence spectra in the library to each pixel spectrum. An
overall fit is then calculated, by weighing each individual
fit based on the area under its continuum-removed spec-
tra. The overall fits are then compared and the mineral
from the library with the best fit is assigned to the im-
age pixel.
One of the strengths of Tricorder is that it considers
three attributes in the analysis: the depth of particular
absorption features, the “goodness of fit,” and the reflec-
tance level of the continuum at the center of the feature.
By doing this, it can analyze feature shapes using all data
points and resolve even complex features, such as dou-
blets in minerals like kaolinite, dickite, and halloysite and
small variations in mineral species.
ALTERATION MAPPING AT BODIE
AND PARAMOUNT
The results of alteration mapping discussed in this sec-
tion uses the locations of Bodie Bluff and Silver Hill,
both in the Bodie district, and Paramount as references
(see Fig. 1).
SAM Results
Two spectral regions were processed separately: the
visible/near-infrared (VNIR) region (0.4–1.3 lm) and the
312 Crósta et al.
shortwave infrared (SWIR) region (2.0–2.4 lm). The
VNIR bands were analyzed for minerals with diagnostic
electronic transition features (hematite, goethite, jarosite,
etc.) and the SWIR bands were analyzed for hydroxyl-
bearing minerals and carbonates.
In the SWIR, 46 AVIRIS bands were used with 15
reference spectra. The threshold value used for the maxi-
mum angle between spectral vectors was established em-
pirically at 0.065 radians. Values above this produced a
large number of classified pixels in only one or two classes
and values below it produced images of mostly “nonclassi-
fied” pixels. SAM classified most of the alteration at Bodie
and Paramount as Na-montmorillonite, Ca-montmorillo-
nite, and a mixture of kaolinite⫹smectite (Fig. 2). This
mixture was concentrated predominantly at Paramount,
with smaller occurrences at three locations at Bodie: Silver
Hill, near the center of district, and Bodie Bluff. At Para-
mount, a clear zonation was revealed with most of the
western part classified as a mixture of kaolinite⫹smectite,
surrounded by pixels classified as Na-montmorillonite, and
the eastern part classified as Na- and Ca-montmorillonite.
SAM identified the two types of kaolinite used as refer-
ences (well and poorly crystalline) in a very small number
of pixels. Illite and Ca-montmorillonite were mapped
mostly as isolated pixels scattered throughout the scene
with no meaningful pattern. We treated these as misclassi-
fications and removed them from the final result.
Computation time for the SWIR 15-mineral classifica-
tion using 46 AVIRIS bands on a low-end Unix worksta-
tion (Sun IPX) was 8 min.
In the VNIR, 94 AVIRIS bands were used with nine
reference spectra of common Fe2⫹- and Fe3⫹-bearing
minerals. Initial results were not satisfactory because SAM
classified most pixels in the entire scene as a mixture of
hematite and quartz. Even after experimenting with differ-
ent values for the minimum angle and removing the he-
matite⫹quartz spectrum from the reference spectra, SAM
classified most of the pixels in the scene as a single refer-
ence mineral. We then decided to use only three refer-
ence spectra: hematite, goethite, and jarosite. Some em-
pirical work had to be done to find the threshold value
for the maximum angle, which was 0.25 radians. We no-
ticed that changes in the value of this threshold produced
significant changes in the number of pixels classified as
one of the reference minerals.
The results showed goethite in altered areas at Bodie
Bluff and Silver Hill, together with jarosite in mine dumps
and in a tailings pond (Fig. 3). A few pixels were mapped
as hematite at Silver Hill and in an area to the north of
the major alteration. SAM also mapped goethite in the
Paramount district as well as jarosite, which also appeared
scattered throughout the western part of the scene, and
along dirt roads.
It took 15 min to run SAM using 94 AVIRIS bands
and three reference spectra on a low-end Unix workstation
(Sun IPX).
Tricorder Results
Tricorder is run through command files, which are set to
analyze specific groups of minerals. The command file
used included 126 materials (minerals, vegetation, and
ice). Among the minerals analyzed were some of the
most common alteration minerals. All the minerals ana-
lyzed using SAM were included in Tricorder’s com-
mand file.
Tricorder also analyzes separately for multiple min-
erals in the VNIR and SWIR spectral regions. Therefore,
the analysis of the whole spectrum for a single pixel may
find two different minerals, one for each spectral region.
Figures 4 and 5 show the mineral maps produced
by Tricorder for the SWIR and VNIR, respectively. Com-
pared to results from SAM, these produced a greater
number of spectral classes relatively to mineral species.
In the SWIR region, Tricorder distinguished between
Na- and Ca-montmorillonite in most of the Bodie district
and at the eastern part of Paramount. It also distinguished
two types of kaolinite (well and poorly crystalline) in the
western part of Paramount, a mixture of kaolinite⫹smec-
tite in the central part of the Bodie district, halloysite in
the western part of Paramount, plus a few pixels as Al-
muscovite mostly at Silver Hill and the center of the Bo-
die district, and illite in mine tailings at Silver Hill. Dolo-
mite was given a good fit by Tricorder, with a fair number
of pixels being mapped as such, but these pixels were iso-
lated and scattered all over the image without any mean-
ingful geological pattern. We treated these as misclassifi-
cations and removed them from the final result, together
with other minerals with low fits and small number of
pixels: dolomite, phlogopite, hectorite, diaspore, clintonite,
palygorskite, elbaite, and nontronite.
In the VNIR, Tricorder mapped two different types of
hematite: the first one, a fine-grained variety, was mapped
around hydrothermally altered areas at Bodie and Para-
mount, and the other, a mixture of hematite and quartz,
was distinctively associated with the alteration at Bodie
Bluff. It also mapped goethite in the center of the Bodie
district and in a nearby tailings pond. A few pixels scat-
tered across the Bodie district were mapped as K-jarosite,
while some others were classified as Fe2⫹-bearing miner-
als (mostly cummingtonite) at the center of district and
at Silver Hill.
The algorithm took around 4 h to process the 141
MB scene for 126 minerals in the VNIR and SWIR run-
ning on a medium-performance Unix workstation (HP
9000) at the USGS Spectroscopy Laboratory.
Comparison of Results Obtained by SAM
and Tricorder
A spatial comparison between the areas of major alter-
ation identified by both methods reveals that Na-mont-
 Mapping Hydrothermal Alteration Minerals 313

Figure 2. SAM classification results for the SWIR bands (2.00–2.45 lm) of the AVIRIS scene over Bodie and Paramount.

Figure 3. SAM classification results for the VNIR bands (0.42–1.34 lm) of the AVIRIS scene over Bodie and
Paramount.
314 Crósta et al.

Figure 4. Tricorder classification results for the SWIR bands.

Figure 5. Tricorder classification results for the VNIR bands.


morillonite was identified by both methods at approxi-
mately the same locations at Bodie and the eastern part
of Paramount. SAM, however, mapped a larger number of
pixels as Ca-montmorillonite in both districts, particularly
at eastern Paramount, in areas, which Tricorder assigned
to halloysite, and both types of kaolinite. The mixture of
kaolinite⫹smectite shows good correlation between both
methods only at the center of the Bodie district, whereas
another major area of occurrence for this mineral mapped
by SAM at the western part of Paramount was mapped
by Tricorder as both types of kaolinite and as halloysite.
The separation achieved by Tricorder, between Na-
montmorillonite, halloysite, and poorly crystalline kaolin-
ite at the eastern part of Paramount, did not appear in
the SAM results.
In the VNIR, major differences show up in the com-
parison between results obtained by both methods for
those minerals that could be mapped by SAM (hematite,
goethite, and jarosite). SAM did not find much hematite
in the whole scene, whereas Tricorder mapped signifi-
cant clusters of pixels as hematite at both districts. How-
ever, some of the pixels mapped as hematite by Tri-
corder were assigned to goethite by SAM. On the other
hand, pixels that Tricorder mapped as goethite were as-
signed to jarosite by SAM.
SPECTRAL ANALYSIS
In order to assess the results, we selected five sites for
ground checking and collected samples for laboratory
spectral analysis. Sample analysis also included X-ray dif-
fraction (XRD). These sites correspond to areas where one
or more alteration minerals were mapped in the VNIR
and SWIR. Two of the sites are located at Bodie, corre-
sponding to a mine excavation at Bodie Bluff (sample
BO-11a) and to a nearby tailings pond (sample BO-05).
The other three sites are located at Paramount (BO-14a,
BO-14b, and BO-17), corresponding to natural outcrops
of altered rocks. Samples were broken off with a ham-
mer, except for sample BO-05, collected from fine-
grained sediments on the floor of a dry tailings pond. A
sixth site, located at Paramount, had been selected for
sampling, but could not be reached on the ground. The
pixel spectrum of this sixth site showed a remarkable
Table 1. Comparison of Results between SAM and Tricorder for Selected Sitesa
Pixel Location
Sample (row/line) Tricorder
BO-17 481/435 (Paramount) Na-montmorillonite
BO-11a 161/352 (Bodie) Ca-montmorillonite
BO-14b 589/445 (Paramount) Kaolinite (pxyl) & goethite
BO-14a 485/442 (Paramount) Halloysite
BO-05 78/344 (Bodie) Na-montmorillonite & goethite
* 524/352 (Paramount) Kaolinite
a
Spectra for each site are shown in Figures 6–11, respectively. *The location with an asterisk did not have a sample collected.
Mapping Hydrothermal Alteration Minerals 315
match to the kaolinite spectrum from the library, sug-
gesting a possible mineral zonation of the alteration. Table
1 shows the location of these sites on the scene, a com-
parison of the minerals assigned to them by Tricorder
and SAM and the result of XRD analysis.
Rock samples from each site were spectrally ana-
lyzed using a Beckman UV-5240 spectrophotometer. The
resulting spectra were convolved to AVIRIS bandwidths.
The results of the analyses are shown in Figures 6–10.
Each sample spectrum was plotted together with the
spectrum of the corresponding AVIRIS image pixel. The
AVIRIS spectra were edited to remove atmospheric H2O
and noisy bands. The reference spectra of the mineral
assigned to the pixel by the classification algorithm(s)
were also plotted for comparison.
Figure 6 shows the spectra of sample BO-17 (from
Paramount) and of the AVIRIS’s pixel corresponding to
its location, together with the library spectra for Na-
montmorillonite. The main hydroxyl absorption feature
near 2.2 lm is broader in the AVIRIS spectrum than in
the sample and library spectra. A secondary feature near
2.3 lm can be seem in the AVIRIS spectrum, which does
not appear in the sample spectrum, suggesting the occur-
rence of a mixture of illite/muscovite. XRD analysis showed
the presence of illite and kaolinite in sample BO-17 (Ta-
ble 1), instead of montmorillonite. Although it is difficult
to generalize for an entire pixel the results obtained from
a single sample, it appears that both classifiers were not
able to identify the correct mineralogy at this location
based on the pixel spectrum. For instance, the presence
of kaolinite, identified by XRD, did not show on the
AVIRIS spectrum for this pixel.
Sample BO-11a (Fig. 7) corresponds to the site at
Bodie Bluff classified as montmorillonite by Tricorder
and SAM. Tricorder assigned this pixel to Ca-montmoril-
lonite, whereas SAM classified it as Na-montmorillonite.
The pixel spectrum is quite similar to the one in Figure
6, suggesting also the occurrence of subpixel mixtures of
illite/muscovite and quartz. XRD analysis revealed only
the presence of cristobalite, with a very strong peak. How-
ever, the fact that XRD did not show any hydroxyl-bearing
mineral could be related to the strength of the cristobalite
peak, which may have masked other minerals.
Figure 8 shows the spectrum of sample BO-14b
SAM X-Ray Diffraction
Na-montmorillonite Illite, kaolinite
Na-montmorillonite Cristobalite
Kaolinite/smectite & jarosite Kaolinite
Na-montmorillonite Montmorillonite, kaolinite, (illite)
Na-montmorillonite & jarosite Goethite
Kaol./smectite *
316 Crósta et al.
Figure 6. Plot showing the Beckman spectrum for sample
BO-17 (from Paramount), the reference spectrum for Na-
montmorillonite extracted from the USGS Digital Spectral
Library and the AVIRIS pixel spectrum identified as Na-
montmorillonite by SAM and Tricorder.
(from Paramount), together with the AVIRIS pixel spec-
trum and the library spectrum of poorly crystalline (pxyl)
kaolinite. Both classification algorithms were in agreement
about the presence of kaolinite at this location, which is
confirmed by the distinctive doublet-type absorption
band between 2.15 lm and 2.22 lm. There were differ-
ences, however, about the type or association of this min-
eral. Whereas Tricorder assigned this pixel to poorly
crystalline kaolinite, SAM indicated the occurrence of a
mixture of kaolinite and smectite. The AVIRIS pixel
spectrum suggests a better crystallinity for the kaolinite
Figure 7. Plot showing the Beckman spectrum for sample
BO-11a (from Bodie), the reference spectrum for Ca-
montmorillonite and the AVIRIS pixel spectrum identified
as and Ca-montmorillonite by Tricorder and as Na-mont-
morillonite by SAM.
Figure 8. Plot showing the Beckman spectrum for sample
BO-14b (from Paramount), the reference spectrum for
poorly crystalline kaolinite and the AVIRIS spectrum identi-
fied as such by Tricorder (SAM indicated a mixture of kao-
linite and smectite). Spectral features characteristic of goe-
thite are also present in the AVIRIS and Beckman spectra,
identified as such by Tricorder and as jarosite by SAM.
than indicated by Tricorder. This is confirmed by XRD,
which identified kaolinite in sample BO-14b. The gen-
eral shape of the pixel spectrum between 0.4 lm and 1.0
lm shows the presence of Fe3⫹-bearing minerals at this
location, with the absorption feature at 0.92 lm indicat-
ing goethite. SAM results classified this pixel as jarosite,
whereas Tricorder classified it as goethite. XRD did not
identify any Fe3⫹-bearing mineral in this sample.
Sample BO-14a (from Paramount) corresponds to
the site classified by Tricorder as halloysite and by SAM
as Na-montmorillonite. The pixel spectrum in Figure 9
Figure 9. Plot showing the Beckman spectrum for sample
BO-14a (from Paramount), the reference spectrum for hal-
loysite, and the AVIRIS spectrum identified as such by
Tricorder and as Na-montmorillonite by SAM.
 Mapping Hydrothermal Alteration Minerals 317

minerals in this portion of the Paramount district, indicat-

ing a possible mineral zonation of the alteration.
We also examined spectral plots for AVIRIS pixels
identified as other minerals such as muscovite, illite, he-
matite, and K-jarosite in this scene and compared them
to library spectra. These minerals, however, were mapped
in much fewer pixels (Fig. 4 and 5) and with less signifi-
cant spatial distribution.

CONCLUSIONS
We managed to recognize and map a variety of minerals
associated with hydrothermal alteration processes using
AVIRIS data at the Bodie and Paramount mining districts,
California. The method selected for calibrating the data
and for retrieving reflectance from radiance was precise
Figure 10. Plot showing the Beckman spectrum for sample enough to allow mineral mapping without the use of inde-
BO-05 (from Bodie), the reference spectrum for Na-mont- pendent ground and atmospheric data. This represents a
morillonite and for goethite and the AVIRIS spectrum identi- major advantage for mineral exploration activities, in
fied as goethite in the VNIR bands and as Na-montmorillo-
nite in the SWIR bands. (SAM indicated Na-montmorillonite which ground data acquisition is usually not possible dur-
and jarosite). ing the initial phases of these activities.
Results obtained for the SWIR region showed a rela-
tively good coincidence between the two methods for
suggests the presence of a doublet near 2.2 lm, which mapping the alteration minerals of major occurrence in
is characteristic of halloysite, as seem on the library spec- the study area. Both methods showed an east–west zona-
trum. The match between the sample and pixel spectra, tion in the alteration at Paramount and also some zoning
however, is not good at this wavelength. The broadness of at Bodie. The zonation at Paramount is clearer and, as
the absorption feature on the AVIRIS spectrum does sug- shown by Tricorder, has predominantly kaolinite and hal-
gest some sort of mineral mixture at subpixel level. This loysite in the western and southwestern portions and
is confirmed by XRD, which identified the presence of predominantly Na-montmorillonite in the eastern por-
montmorillonite, kaolinite, and some illite in this sample. tion, with a dividing line coinciding with Atastra Creek.
Figure 10 shows sample BO-05 (from Bodie), col- This shows that only the eastern part of the Paramount
lected at the tailing pond where the AVIRIS spectrum in- District exhibits an alteration similar to the Bodie Dis-
dicates the presence of Fe3⫹ and O—H-bearing minerals.
This location was identified by Tricorder as containing Na-
montmorillonite and goethite. SAM also assigned the oc-
currence of Na-montmorillonite and jarosite, instead of Figure 11. Plot showing the reference spectrum for well-
goethite. A comparison of the library spectrum of Na- crystalline kaolinite and the AVIRIS pixel spectrum identi-
fied as such by Tricorder and as a mixture of kaolinite and
montmorillonite and the pixel spectrum does not show a smectite by SAM. No sample was collected at this locality.
good match for the feature at 2.2 lm in terms of shape,
although the position seems to be correct. Again, the
broadness of this feature as seem on the pixel spectrum
is likely to be related to subpixel mixtures. The main Fe3⫹
feature at approximately 0.92 lm shows a very good match
to the goethite feature. This is confirmed by XRD, which
identified goethite in this sample and no hydroxyl-bear-
ing mineral.
Figure 11 shows the AVIRIS pixel and library spectra
for the location that could not be sampled on the ground
(marked with an asterisk in Table 1). Both classification
algorithms identified the presence of kaolinite at this loca-
tion, based on the well-defined doublet absorption near
2.2 lm, characteristic of well-crystalline kaolinite. An in-
spection of other pixel spectra near this location also re-
vealed similar well-defined doublet absorption. This sug-
gests a predominance of kaolinite over other alteration
318 Crósta et al.
trict. SAM also showed the same zonation but mapped
predominantly kaolinite/smectite in the western and
southwestern portions and Na- and Ca-montmorillonite
in the eastern portion. In addition, Tricorder was able
to map two types of kaolinite, plus halloysite, muscovite,
and illite.
In the VNIR results from both methods were quite
different. Pixels mapped by SAM as goethite were as-
signed by Tricorder to hematite, whereas others mapped
as jarosite by SAM were assigned by Tricorder to goethite.
Tricorder managed to map more pixels as minerals with
diagnostic spectral features in the VNIR than SAM, in-
cluding hematite, goethite, K-jarosite, and cummingtonite.
Spectral analysis of samples from selected sites par-
tially confirmed the results obtained by image classifica-
tion. There was some coincidence, mostly of the position
and partially of shape, of diagnostic features for kaolinite
(possibly with different crystallinity, as indicated by the
shape of the doublet band), kaolinite⫹smectite, musco-
vite, and illite. However, the match between AVIRIS
pixel spectra and the respective sample spectra is some-
what poor in most of the cases. This may be the result
of subpixel mixtures of other minerals, such as illite, mus-
covite, and quartz, and also due to the fact that it is diffi-
cult to represent an entire 20 m⫻20 m pixel with only
one sample. Some sort of subpixel contamination from
other minerals, or even other cover materials (vegetation,
soils), is therefore to be expected and will certainly re-
flect on the results of any mineral mapping procedure.
Results from X-ray diffraction analysis showed some agree-
ment with the results from spectral analysis, particularly
for samples BO-05, BO-14a, and BO-14b. The same agree-
ment did not happen for samples BO-11a and BO-17, in
which both classifiers identified montmorillonite and
XRD revealed cristobalite and illite, respectively.
A comparison of the two classification algorithms used
shows that Tricorder was able to produce mineral maps
with greater variety of alteration minerals. SAM is a
much simpler method that was developed for expedite
measurements of spectral similarity. As such, it is not
able to determine small differences between mineral
species in the way Tricorder does. Some of the charac-
teristics of SAM, such as its availability in commercial
image processing packages, ease of use, and processing
speed, make it a feasible operational option for mineral
mapping in exploration activities. Tricorder, on the other
hand, produced more detailed results than SAM did, but
its complexity will demand some training, basic knowl-
edge of spectroscopy, and considerably more computing
time than SAM.
Although most of the classified areas at the Bodie
District show some sort of ground disturbance due to
mining, the same does not occur at the Paramount Dis-
trict, showing that these methods can be effectively used
for precious-metal exploration in unexplored areas.
The authors would like to thank Robert O. Green and his col-
laborators from the AVIRIS project, at NASA’s Jet Propulsion
Laboratory, for the support with atmospheric calibration of the
data and Dr. Roger N. Clark, at USGS’s Spectroscopy Labora-
tory, for the use of Tricorder and fruitful discussions on min-
eral spectroscopy. XRD analyses were carried out by Dr. Maria
Luiza M. C. Silva, from the Geosciences Institute, UNICAMP.
We also thank the two anonymous reviewers for suggestions
that contributed for improving the paper. A. P. Crósta acknowl-
edges the support of the Desert Research Institute (DRI), Uni-
versity and Community College System of Nevada, during a
sabbatical year as a visiting scientist, as well as the State Uni-
versity of Campinas, Brazil. This work was funded in part by
Fundação de Amparo à Pesquisa no Estado de São Paulo
(FAPESP) through Grant No. 94/3474-0 and by Conselho Nacio-
nal de Desenvolvimento Cientı́fico e Tecnológico (CNPq) through
grant No. 320229/84-0.
REFERENCES
Chrien, T. G., Green, R. O., and Eastwood, M. (1990), Accu-
racy of the spectral and radiometric laboratory calibration of
the Airborne Visible/Infrared Imaging Spectrometer (AVIRIS).
In Proc. Second Airborne Visible/Infrared Imaging Spec-
trometer (AVIRIS) Workshop, 4–5 June, JPL Publ. 90-54,
Jet Propulsion Laboratory, Pasadena, CA. pp. 1–14.
Clark, R. N., and Swayze, G. A. (1995), Mapping minerals,
amorphous materials, environmental materials, vegetation,
water, ice and snow, and other materials: the USGS Tri-
corder algorithm. In Summaries of the Fifth Annual JPL
Airborne Earth Science Workshop, 23–26 January, JPL
Publ. 95-1, Jet Propulsion Laboratory, Pasadena, CA, Vol.
1, pp. 39–40.
Clark, R. N., Gallagher, A. J., and Swayze, G. A. (1990), Mate-
rial absorption band depth mapping of imaging spectrome-
ter data using a complete band shape least-square fit with
library reference spectra. In Proc. Second Airborne Visible/
Infrared Imaging Spectrometer (AVIRIS) Workshop, 4–5
June, JPL Publ. 90-54, Jet Propulsion Laboratory, Pasadena,
CA, pp. 176–186.
Clark, R. N., Swayze, G. A., Gallagher, A., King, T. V. V., and
Calvin, W. M. (1993), The U.S. Geological Survey Digital
Spectral Library: Version 1: 0.2 to 3.0 lm, USGS Open File
Report 93-592, U.S. Geological Survey, 1340 pp.
Clark, R. N., Swayze, G. A., Heidebrecht, K., Green, R. O.,
and Goetz, A. F. H. (1995), Calibration to surface reflec-
tance of terrestrial imaging spectrometry data: comparison
of methods. In Summaries of the Fifth Annual JPL Airborne
Earth Science Workshop, 23–26 January, JPL Publ. 95-1, Jet
Propulsion Laboratory, Pasadena, CA, Vol. 1, pp. 41–42.
Green, R. O., Conel, J. E., and Roberts, D. A. (1993a), Esti-
mation of aerosol optical depth, pressure elevation, water
vapor and calculation of apparent surface reflectance from
radiance measured by the Airborne Visible/Infrared Imag-
ing Spectrometer (AVIRIS) using a radiative transfer code.
In Proceedings SPIE, Orlando, FL, April 14–15, SPIE Vol.
1937, pp. 2–11.
Green, R. O., Conel, J. E., Helmlinger, M., Bosch, J., Chovit,
C., and Chrien, T. (1993b), Inflight calibration of AVIRIS

in 1992 and 1993. In Summaries of the Fourth Annual JPL
Airborne Earth Science Workshop, 25–29, October, JPL
Publ. 93-26, Jet Propulsion Laboratory, Pasadena, CA, Vol.
1, pp. 69–72.
Herrera, P. A., Closs, L. G., and Silberman, M. L. (1993), Al-
teration and geochemical zoning in Bodie Bluff, Bodie min-
ing district, eastern California. J. Geochem. Exploration 48:
259–275.
Kruse, F. A., Lefkoff, A. B., Boardman, J. W., et al. (1993), The
Spectral Image Processing System (SIPS)—interactive visual-
ization and analysis of imaging spectrometer data. Remote
Sens. Environ. 44:145–163.
View publication stats
Mapping Hydrothermal Alteration Minerals 319
Silberman M. L., Breit, F., and Lawrence, E. F. (1995), Geology
and Ore Deposits of Bodie Hills, Northern Mono Basin Re-
gion, Special Publ. 22, Geological Society of Nevada, 49 pp.
Research Systems, Inc. (1995), ENVI User’s Guide, Version 2.0.
RSI, Boulder, CO.
Taranik, J. V., and Crósta, A. P., (1996), Remote sensing for
geology and mineral resources: an assessment of tools for
geoscientists in the near future. In XVIII Congress of the
International Society of Photogrammetry and Remote Sens-
ing, Vienna, Austria, 9–19, July, International Archives of
Photogrammetry and Remote Sensing, Vol. XXXI, Part B7,
pp. 689–698.