Supervised Classification for Lithologic Discrimination in Shaikh Ibrahim Area, NW Iraq Using Landsat Images

In order to evaluate the supervised classification in discrimination of geological materials, the study intends to deal with Landsat ETM+ images to define the spectral behaviors of lithologic components and determine the adequacy of Maximum-Likelihood Classification (MLC) in discriminating various sedimentary rocks including limestone and clay deposits in Shaikh Ibrahim area, northwestern Iraq. The proposed strategy is based on color composite ratio, Principal Component Analysis (PCA), preliminary classification of Spectral Angle Mapper (SAM), and MLC. The combination of image ratios mostly cannot differentiate the sensitivity variations of weathering products of parent materials. Whereas, PCA works toward better result in categorization of most components occurrences, SAM identifies accurately the responses of several endmembers including spectral perceiving of two types of limestone. MLC discriminates significantly the configuration and lithologic-contents of various geologic formations, where they are dominated spectrally by considerable areas of carbonates (42.54 %) in addition to clay and scree deposits (22.89 %). The spectral behavior of carbonate, clay, gypsum, and iron-bearing materials are the main spectral component types in the study area which meet a reliable lithologic discrimination by using Landsat ETM+ imagery. The study result compares favorably with the available geologic map of the study area and the reconnaissance field trip's views of lithological classes supported the discrimination result. MLC shows considerable power in discrimination of sedimentary lithologies with 100 % accuracy when interactive preliminary classification of SAM is considered.