In: Wagner W., Székely, B. (eds.): ISPRS TC VII Symposium - 100 Years ISPRS, Vienna, Austria, July 5-7, 2010, IAPRS, Vol. XXXVIII, Part 7B
261
2004). Herein this study, thermal reflectance properties of
ultramafic rocks rich in olivine minerals will be estimated.
Figure 4. Simplified stratigraphic section of the study area.
Lower Cretaceous Eldivan ophiolite, which is given
in green (dark) color, overlays Triassic basement
units, and overlaid by Tertiary sedimentary
sequences.
2.1 VNIR-SWIR Band Ratio
According to the spectral properties of mafic minerals given in
figure 1A, characteristic wavelengths can be correlated with
ASTER VNIR and SWIR bands.
There are higher reflection responses in ASTER bands 2, 5 and
8, absorption in bands 1, 3 and 6 for olivine mineral while only a
broad absorption in band 3 and a reflection in band 1 can be
noticeable for pyroxene minerals in the SWIR region of
electromagnetic spectrum. Serpentine, the alteration product of
those minerals, has also distinguishing absorptions in bands 2, 7
and 8 and reflection in bands 3 and 6. The band ratios, 2/3,
2/3*1/3 for olivine mineral and 6/8 and 6/7 for serpentine
mineral were developed.
2.2 TIR Band Algorithms
The pyroxenes exhibit visible reflections in ASTER bands 11 and
12 according to the given figure IB. Despite, olivine mineral has
higher reflection in band 13 and lower in bands 10, 11 and 12.
While showing similarities, the olivine mineral is higher in bands
12 and 14 compared to the alteration product of olivine and
pyroxenes, serpentine.
As the olivine-peridotite or dunite, composed of mainly olivine
and additionally pyroxenes and less silica, is compared with
ASTER bands in figure 5, band 13 is highly reflected while the
bands 11, 12 has absorption and 14 is relatively low reflected.
The absorption properties of ASTER bands 11 and 14 for
olivine display similarities to quartz mineral. However, the
difference between those two spectra is noticeable in bands 12
and 13. This difference was also recorded as a result of a change
in silicate content within the igneous rocks varies from felsic
through mafic to ultramafic by some researchers (Ninomiya and
Matsunaga, 1997, 2002, 2003; Hook et al, 2005).
While the higher reflection in band 13 for dunite and
neighbouring bands 12 and 14 are taking under consideration in
figure 5, the band ratios of 12/13 and 14/13 is achieved. To use
opposite band ratios instead of usual ones is the cause of the
energy emitted from the surface materials in thermal region. The
higher reflection in the spectrum corresponds to the low
emissivity. Since above ratios, (12*14) / (13 2 ) is obtained.
Figure 5. The spectral pattern comparison of olivine-peridotite
(dunite) and quartz with corresponding ASTER
bands
The formula, (14/13) / (13/12) 3 , is developed by considering the
ratio 14/13 for olivine and pyroxene mineral with comparing to
the ratio 13/12 for silicate content. The exponential parameter
(herein 3) is a number that is used for decreasing the effect of
silica content within the rocks. The number used here, 3, is
chosen for giving the best result.
2.3 Results and Discussions
The VNIR-SWIR band ratios developed to extract the
ultramafic units have mostly not efficient, except serpentine band
ratio 6/8. On the contrary, the better results were attained with
thermal infrared (TIR) region for detection of lithology like
olivine-peridotite occurrences.
The results of VNIR-SWIR band ratios, 2/3 for olivine and 6/8
for serpentine minerals are given in Figure 6. Besides the
matching localities for olivine resultant image, it is also possible
to observe the areas unmatched. The serpentine band ratio 6/8
(or even 6/7) can mostly be correlated with the serpentinized
ultramafic rock units. Besides the results of the olivine band
ratio 2/3 correlating well with the ophiolite formations inside, it
has widely wrong results outside the ophiolitic formation
boundaries.
The olivine rich ultramafic rock, olivine-peridotite or dunite was
noticeably extracted within the ophiolitic melange formation
given in simplified geological map (figure 3, 7A, 7B) with
distinct boundaries by using the spectral properties of target
ultramafic rocks. Additionally, the ground truth studies were
conducted to 15 localities of matched or unmatched with either
analyses results or geological map units. 13 of 15 collected