Pixel-based and sub-pixel analysis of ASTER data for identifying lithological and mineralogical units; a case study of Tutak area, Fars Province

The Tutak area (northeastern part of Shiraz city) located in Sanandaj-Sirjan metamorphic Zone (SSZ) was chosen as study area because of having two important resources of iron and copper. Two metamorphic complexes of SSZ namely Surian and Tutak are the main host of these resources. The algorithms of Spectral Information Divergence and Adaptive Coherence Estimator were used to detect and distinguish rock units and indicator minerals. Using spectra of field samples, the pixel-based algorithm of Spectral Information Divergence enhanced greenschist and marble lithological units because of their significant spectral differences in VNIR region. Overall accuracy and kappa coefficients resulted by using this algorithm for detection of rock units were 0.83 and 0.72, respectively. Using data of USGS spectral library, the outputs of the sub-pixel algorithm of Adaptive Coherence Estimator successfully identified pixels with high abundances of chlorite and ankerite. Results showed that chlorite areas with its abundances more than 50% overlapped greenschist units, and the areas with chlorite abundances above 75% had acceptable conformity with the copper deposits, most prominent of which is Mazaijan copper deposit. The areas with more than 50% ankerite overlapped the marble units, and its abundances above 75% indicated an increase of the abundances of this mineral compared to the minerals of calcite and dolomite in this rock unit with approaching iron mineralization. So, based on study results, the Adaptive Coherence Estimator technique can be suggested to accurately detect and distinguish the minerals of the mineralogical reservoirs in study area.