Figure 31 shows part of a West-to-East flight line as seen in two
thermal IR wavelength bands. The ratio of radiances from this particular
pair of wavelength bands is generally sensitive to Si0, content of the
rocks. The physical basis for the method depends on two properties of
silicates. First, all silicate rocks characteristically display broad
emissivity minimums (reststrahlen bands) in the 8 to 12 um wavelength
region. This minimum is caused primarily by the silicon-oxygen stretching
modes in silicate lattices. Secondly, some relationship exists between
the position of the silicate emissivity minimum and rock type. The
minimum tends to shift to longer wavelengths as the silica content de-
creases. From left to right is a dacite porphyry mountain range,
alluvium, basaltic lava flow, more alluvium, playa deposits and more
basaltic lava. The darkest parts of the ratio image have been found to
have the largest SiO, content, as expected. Note the difference between
2
the dacite mountains and basaltic lava on the ratio map; this is caused
by the difference in their chemical make-up; in this case their SiO,
content. Figure 32 shows the same region, but in different wavelength
regions. The two bands used for ratioing in this case were a visible
green band and a reflective IR band. This ratio is sensitive to the
presence of iron oxides, dark corresponding to greater contents of iron
oxide. Note the striking enhancement of the andesite dike in the
dacite porphyry mountain; this dike is hardly discernible in the single
channel images. Part of this dike was not even recorded on the best
geological maps of the area. Also note the greater iron oxide content
in the basaltic lava as compared to the alluvium. Figure 33 shows
another flight line (North to South), once again in the thermal IR
region. Pisgah Crater is located on the right-hand side of the images.
The SiO, differences are greatly enhanced in the ratio image, where the
alluvium appears dark and the basaltic lava flow appears bright.
It seems likely that these developments will lead to better methods
for rapid geologic mapping and mineralogical exploration in previously
inaccessible areas.
5. Conclusions
Despite making substantial progress in achieving improved processing
techniques, improved throughput parallel processing systems, and showing
the practical use of multispectral sensing in various earth resources
applications under limited conditions, there is still much to be
accomplished before multispectral scanner surveys provide successful,
operational user information systems for earth resources and land use
surveys. Potential benefits of such information will remain largely
"potential" until this work is done.
The most critical need is for processing systems capable of better
keeping pace with the sensors so that unprocessed data does not have
to be accumulated and stored. Greatly reduced costs of processing are
also likely to result with an attendant increase in interest in multi-
spectral scanner surveys on the part of potential users.