Mapping Geology and Structure in Hyper-Rugged Terrain Using ASTER Remote Sensing Data: A Case Study From Northern Pakistan

Extreme topography in mountainous terrane introduces radiometric distortion in remote sensing data. This effect influences the ability to correctly map lithology based on unique spectra. In this study we utilize several topographic correction regimes to accommodate spectral variation caused by topography, including the Cosine-correction and the Minnaert-correction methods. The study area is located in Chitral, northern Pakistan. In this area, the world's three greatest mountain ranges, the Himalayas (represented by Kohistan terrane), the Karakoram, and the Hindukush terrane merge together. The Himalayas are separated from the Karakoram terrane by the Shyok Suture. Whereas, the Karakoram plate is separated from the Hindukush terrane by the Reshun Fault. The area is extremely rugged; local relief is more than 2,500 meters. Many peaks are higher than 5,000 m and Tirch Mir, 12 km north of the mapped area, is 7,702 m high. ASTER imagery and a digital elevation model developed from the ASTER data is used to first apply the topographic correction then perform image analysis techniques, such as relative band-depth, to characterize the lithology. High pass filtering and textural analysis are also used to help delineate the structures in the region. This study aims to update the most current geologic map, produced by Calkins et al. (1981) in order to fully understand the interaction between the Kohistan, Karakroam and Hindukush along the Shyok Suture and the Reshun Fault. We utilized the published geological map and our field knowledge of the area to classify the ASTER data with spectral angle mapper (SAM) and Mixture Tuned Matched Filtering (MTMF).