Exploring an efficient sandy barren index for rapid mapping of sandy barren land from Landsat TM/OLI images

Abstract

Sandification is a serious global environmental problem. The spatiotemporal dynamics of sandy barren land is an important indicator in monitoring its ecological restoration. Such monitoring requires fast and accurate identification of sandy barren land. Although several bare indexes have been developed for the mapping of barren land, these have not proven effective in the mapping of sandy barren land. In this paper, the partial normalized sandy barren index (PNSBI), a new barren index based on spectrum reflectance measurement and analysis, is proposed. The efficiency and enhanced accuracy of PNSBI at retrieving sandy barren land was validated using Landsat Thematic Mapper (TM) and Operational Land Imager (OLI) images of the Poyang Lake and Ordos regions of China. These results suggest that the proposed PNSBI achieved higher accuracy than the other bare indexes in retrieving sandy barren land from the other types of bare land. And it is benefit for the study of the spatiotemporal dynamics of sandification, water distribution, soil erosion, etc.