Contribution of ALOS PALSAR data to forest characterization and monitoring in Australia

Abstract

Research in Australia has focused on using the combination of moderate (25–30 m) spatial resolution ALOS PALSAR, Landsat-derived persistent green fractional cover and ICESAT GLAS data to generate maps of the structure of forests as an intermediate step to mapping above ground biomass (AGB) and AGB change using combinations of JERS-1 SAR, ALOS PALSAR and ALOS-2 PALSAR-2 data. The method involved a) segmentation of ALOS PALSAR and Landsat-derived persistent green fractional cover to produce objects across the landscape, b) subsequent assignment of objects to classes associated with different structural formations and c) combining cover and height metrics (derived from ICESAT data) to generate a classification of forest structural types. The approach was applied Australia-wide to produce a new forest structural classification. In future work, a series of algorithms are being compared to facilitate retrieval of AGB and detection of AGB change using combinations of these data, including the ALOS-2 PALSAR-2. Validation is being performed using a collation of field data and derived biomass estimates and airborne LiDAR acquired for a wide range of sites across Australia.