Forest Discrimination Analysis of Combined Landsat TM and ALOS-PALSAR Data

The joint processing of remote sensing d ata acquired from sensors operating at different wavele ngths has the potential to significantly improve the oper ation of global forest mapping and monitoring systems. This p aper presents an analysis of the forest discrimination p roperties of Landsat TM and ALOS-PALSAR data when considered as a combined source of information. This study is c arried out over a test site in north-eastern Tasmania, Aus tralia. Canonical variate analysis, a directed discriminant technique, is used to investigate the separability of a number of training sites, which are subsequently used to d efine spectral classes as input to maximum likelihood classification. An accuracy assessment of the class ification results is provided on the basis of independent gro und validation data, for the Landsat, PALSAR, and combined SAR–optical data. The experimental results demonstra te that: 1) considering the SAR and optical sensors jo intly provides a better forest classification than either used independently, 2) the HV polarisation provides most of the forest/non-forest discrimination in the SAR data, and 3) the respective contribution of each of the Landsat and PALSAR bands to the separation of different types of fores t and nonforest land covers varies significantly.