A spectral index ratio-based Antarctic land-cover mapping using hyperspatial 8-band WorldView-2 imagery

This study evaluates the potential of 8-band high resolution WorldView-2 (WV-2) panchromatic (PAN) and multispectral image (MSI) data for the extraction of polar geospatial information. We introduce a novel method based on a customized set of normalized difference Spectral Index Ratios (SIRs), incorporating multiple bands, to improve the accuracy of land-cover mapping in the Antarctic. Most recently available WV-2 data are classified into land-cover surfaces such as snow/ice, water bodies, and landmass using the customized normalized difference SIRs. A novel multi-fold methodology is used to evaluate the effect of pan-sharpening algorithms on spectral characteristics of satellite data, and on subsequent land-cover mapping using an array of SIRs. A set of existing pan-sharpening algorithms was implemented in order to fuse PAN with MSI data, followed by estimation of multiple SIRs to extract target land-cover classes. These algorithms were compared on the basis of their effectiveness in extracting target classes using a defined set of SIRs. Our results indicate that the use of 8-band WV-2, customized SIRs, and appropriate pan-sharpening can greatly improve the extraction of land-cover information.