Detection of Alteration Induced by Onshore Gas Seeps from ASTER and WorldView-2 Data

• Main Authors: Sanaz Salati, Frank van Ruitenbeek, Freek van der Meer, Babak Naimi
• Format: Article
• Language: English
• Published: MDPI AG 2014-04-01
• Series: Remote Sensing
• Subjects: hydrocarbon seep; alteration mineral; ASTER; WorldView-2; boosted regression trees (BRT) classification
• Online Access: http://www.mdpi.com/2072-4292/6/4/3188

Hydrocarbon seeps cause chemical and mineralogical changes at the surface, which can be detected by remote sensing. This paper aims at the detection of mineral alteration induced by gas seeps in a marly limestone formation, SW Iran. For this purpose, the multispectral Advance Spaceborne Thermal Emission and Reflection Radiometer (ASTER) and the high spatial resolution WorldView-2 (WV-2) data were utilized for mapping surficial rock alteration. In addition, the potential of Visible Near Infrared (VNIR) bands of the WV-2 and its high spatial resolution for mapping alterations was determined. Band ratioing, principal component analysis (PCA), data fusion and the boosted regression trees (BRT) were applied to enhance and classify the altered and unaltered marly limestone formation. The alteration zones were identified and mapped by remote sensing analyses. Integrating the WV-2 into the ASTER data improved the spatial accuracy of the BRT classifications. The results showed that the BRT classification of the multiple band imagery (created from ASTER and WV-2) using regions of interest (ROIs) around field data provides the best discrimination between altered and unaltered areas. It is suggested that the WV-2 dataset can provide a potential tool along higher spectral resolution data for mapping alteration minerals related to hydrocarbon seeps in arid and semi-arid areas.