Potential Utility of Spectral Angle Mapper and Spectral Information Divergence Methods for mapping lower Vindhyan Rocks and Their Accuracy Assessment with Respect to Conventional Lithological Map in Jharkhand, India

Abstract

In this study, we have evaluated the potential use of spectral mapping algorithms in deriving spectrolithological maps of metasedimentary rocks of Vindhyan group of rocks. In this regard, we have processed visible near infrared (VNIR) and shortwave infrared (SWIR) bands of Advanced Speceborne Thermal Emission and Reflection Radiometer (ASTER) data using similarity based spectral mapping algorithms such as spectral angle mapper (SAM) and spectral information divergence (SID). Laboratory spectra were collected by Fieldspec 3© spectroradiometer for main rock types of study area and the spectra were resampled to ASTER bandwidth to compare laboratory spectra with image spectra of respective rocks. Overall matching of image spectra of rocks with their ASTER resampled laboratory counterparts justified the spectral integrity of these rocks on the image. Therefore, image spectra of rocks were used as end member for deriving spectral maps using SAM and SID method. These maps were compared with the conventional field based lithological map (consequently updated using ASTER false colour image composite and band ratio images). SAM spectral map had over all accuracy of 67.41% and the SID map had overall accuracy of 69.67%. Present study has brought out the fact that spectral mapping algorithms would be useful in deriving moderate accuracy lithological maps even if the sedimentary rocks are of close mineralogy and these rocks have very close reflectance spectra within the spectral bandwidth of ASTER sensor. Spectral maps corroborate well with the discrete geochemical data.