FOREST DRAGON 2: ADVANCES OF CHINESE PARTNERS DURING THE FOURTH PROJECT YEAR

摘要

In the fourth year for Forest Dragon 2 project, Chinese groups engaged in following activities: 1) forest mapping methods improvement, 2) regional forest map generation, 3) field measurements and airborne observations for forest map validation. An objectoriented forest and non-forest classification method was proposed based on the HH to HV polarization intensity ratio and HV images of ASAR data at single acquisition time in winter for temperate forests. An integrated algorithm for regional forest biomass estimation was developed using in-situ, airborne and spaceborne lidar data, and other imageries. For forest maps from ENVISAT ASAR data, the overall accuracy 83.7% and 86.16% in the Northeast and Southwest of China respectively. For forest aboveground biomass in Yunnan province of China, our remote sensing estimation agreed well with inventory based results, and the difference was 7.8%.