MULTI-SCALE MAPPING OF FOREST GROWING STOCK VOLUME USING ENVISAT ASAR, ALOS PALSAR, LANDSAT, AND ICESAT GLAS

摘要

Multi-scale approaches for mapping abovegroundbiomass globally are evaluated that exploit the multitemporalarchive of low-resolution (1 km) ENVISATASAR C-band observations and ca. 30 m resolutionALOS PALSAR L-band and Landsat mosaics. TheBIOMASAR algorithm, which was initially developedfor ENVISAT ASAR C-band data and boreal forest [1],is deployed to map growing stock volume, a proxy foraboveground biomass, globally at 1 km resolution. Weexplore different options for improving ASAR basedmaps using high resolution data. Two approaches arepursued: 1) the BIOMASAR algorithm adopted for Lband,2) a simple re-scaling of ASAR derived estimatesof growing stock volume from 1 km pixel posting to 30m using PALSAR and Landsat data [2]. The initialresults for different forest ecosystems suggest that bothapproaches allow for improved estimates, albeit withthe expected limitations in high biomass forests.