LLLI
|
|
LLLI
HILL
©
=
TIR sensors are capable of detecting quartz, jasperoid,
silicification and chalcedonic/opaline sinter and caps due
toemission minima of silica varieties around 8.9 microme-
ters (Hunt, 1982; Christiansen et. al., 1986). The amor-
phous varieties of silica - chalcedony and opal - absorb over
abroad interval at 2.4 micrometers and can be detected with
scanners that measure energy in this interval. The major
iron oxide species - goethite, jarosite, and hematite - that
form from the weathering of sulfides absorb energy at
different frequencies in the VNIR/SWIR (Rowan, 1983;
Lee and Raines, 1984), providing a means of discrimina-
tion using hyperspectral scanners (e.g., Taranik et al.,
1991). Energy absorption occurs at 0.86, 0.91, and 0.94
micrometers for hematite, jarosite, and goethite, respec-
tively, and the absorption trough is steeper for hematite.
Jarosite has distinctive absorption features in the SWIR
that hematite and goethite do not share. Hyperspectral
scanners with appropriate bandpasses are capable of dis-
criminating these three important iron oxide species, the
relative amounts of which often relate to supergene enrich-
ment at porphyry copper deposits and zoning patterns at
some types of precious metal systems.
Sensors that measure narrow-band TIR energy are capable
of discriminating lithologies on the basis of quartz and
silicate mineralogy. The emission minima of silicate-
bearing rocks move to lower frequencies or longer wave-
lengths with increasing mafic composition (Vincent et al.,
1975; Christiansen et al., 1986), permitting detection of
compositional variations among volcanic and metamor-
phic units. Longer SWIR wavelengths are absorbed less by
desert varnish and other surficial products of weathering
than the shorter wavelengths, resulting in greater spectral
response from underlying host rocks (Spatz and Taranik,
1989). The evolved igneous rocks, including peralkaline
flows, units enriched in incompatible and large-ion
lithophile elements, and alkalic rocks in general often
exhibit steep spectral slopes through the 1.5 to 2.5 mi-
crometer interval (Spatz and Taranik, 1989).
Major fault controls are cited at most types of hydrothermal
ore deposits, from regional faults coincident with intrusive
rocks at porphyry ore provinces and sediment-hosted gold
belts to caldera related faults; from thoroughgoing faults at
epithermal deposits in volcanic fields to rift related faults
at alkalic centers and shears between sutured terrains.
Structural intersections are thought to control localization
of some deposits. Simple contrast enhancements as well as
more complex image processing techniques like edge
enhancement and principle component imagery, can be an
effective tool for highlighting linear structural features
based on topography, juxtaposed lithologies, linear alter-
ation patterns, and vegetation contrast. Detection of fault
linears and other structural patterns is often dependent on
Spectral contrast across the fault zone. Radar can enhance
linears, curvilinears, andother topographic expressions of
Structure.
641
Vegetation often provides an indirect indication of hydro-
thermal alteration, supergene alteration, lithologies and
structure. Stressed vegetation, growing on metalliferous
soils, and variations in plant species resulting from soil
composition can lead to anomalous reflectance values in
VNIR spectra (e.g., Raines and Canney, 1980; Milton,
1983; Collinsetal., 1983; Ageretal., 1989; Eiswerth et al.,
1989. Riparian growth may occur along important fracture
zones or within depressions related to mineralization. A
reduction in plant cover may be caused by rocky knobs and
ridges underlain by hydrothermal silicification or by toxic
soils resulting from sulfides or metals. Changes in vegeta-
tion species as well as reductions in plant density and vigor
can be caused by mineralization and poorly drained clay-
rich soils. Hyperspectral sensors that measure discrete
intervalsacross this reflectance boundary distinguish shifts
toward either shorter wavelengths (blue shift) or longer
wavelengths (red shift). A shift in either direction could be
relatedto mineralization. Radar, particularly shorter wave-
length radar, is very sensitive to vegetation density.
4.0 DEPOSIT MODELS AND REMOTE
SENSING STRATEGIES
4.1 Precious Metal Deposit Models
Renewed interest in gold exploration during the final two
decades of the 90's has lead to revised classification
schemes for precious metal deposits based on field obser-
vation (Boyles, 1979; Watson, 1980; Buchanan, 1981;
Worthington, 1981; Graybeal, 1981; Bonham, 1985, 1989;
Titley, 1987; Schafer et al., 1988; Shawe, 1988, Cox and
Singer, 1986; Sillitoe, 1993; and White and Hedinquist,
1995). Theseclassifications are similarly rootedin descrip-
tive geologic features and tectonic setting rather than
genesis or physicochemical conditions of formation. Me-
gascopic field taxonomies are convenient for discussion
and comparison of remote sensing techniques, and toward
that purpose the hydrothermal precious metal systems are
subdivided into the following types: 1) sediment-hosted
Carlin-type gold and silver deposits (Table 2); 2) volcanic-
hosted high-sulfidation deposits including hot springs,
maar, and porphyry gold deposits (Table 3); 3) volcanic-
hosted low-sulfidation veins and stockworks (Table 4); 4)
deposits related to plutonic intrusions, including veins and
shears, gold skarns, polymetallic veins, Fort-Knox type,
and deposits peripheral to porphyry copper/molybdenum
systems (Table 5); 5) deposits hosted by metamorphic
rocks, including quartz veins, exhalite deposits, and aurif-
erous iron formation (Table 6); 6) detachment related
deposits (Table 7); and 7) alkalic systems in rift environ-
ments (Table 8).
Inasmuch as the geologic characteristics of these ore
models vary so too do their remote sensing characteristics.
Taranik (1988) and Kruse (1989) have outlined remote
sensing fundamentals of gold exploration in general, and
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B7. Vienna 1996
