Forest aboveground carbon mapping using multiple source remote sensing data in the Greater Mekong Subregion

The Greater Mekong Subregion (GMS) is rich in forest resources. It is important to estimate forest carbon with high accuracy methods in this region. Remote sensing is an efficient way to estimate forest parameters in large area, especially at regional scale where field data is scarce. LIDAR provides accurate information on the vertical structure of forests. In this study, the forest carbon was estimated at ICESat GLAS footprint level after being trained with field measurements and airborne lidar estimations in GMS. According to different types of ecological zones, a set of categorical regression models was built between ICESat GLAS estimates and ENVISAT MERIS spectral variables together with MODIS VCF product. Then the forest carbon map with continuous value was generated. The estimation agreed well with FAO FRA 2010 report and other published results, and the average difference was about 13.3%.