Making the Best Use of Landsat MSS Images for Land Use/Cover Change Analysis

Current land use and cover (LUC) mapping is often made using high resolution images such as those from the Landsat TM/ETM+, SPOT, IKONOS, QuickBird sensors etc. However, the historical LUC information was often extracted from the earlier Landsat MSS images. Resultant land cover classification accuracy is often unsatisfactory for quantitative change analysis due to the limited number of spectral bands and the low spectral sensitivity. In this study, digital elevation model (DEM) data and the normalized difference vegetation index (NDVI) were integrated with MSS images. And also, a supervised spectral angle mapper (SAM) classification method was applied to improve the LUC classifications from the MSS images. The results showed that the MSS images classification accuracy could be improve to achieve a better LUC map with this method with overall accuracy and Kappa statistics of 81.4% and 0.781 respectively.