AIRBORNE LIDAR FOR ESTIMATING ABOVEGROUND BIOMASS IN DIPTEROCARP FORESTS OF MALAYSIA

ABSTRACT: Dipterocarp forest is an inland forest ecosystem that reside on land with elevation <1200 m is rich with numerous plant species and timber trees. Amount of biomass in this forest is typically high as compared to other ecosystems in the tropics. Given their prominent role in global biogeochemistry, and the likelihood that these high productivity areas will be prime areas for carbon sequestration efforts, better characterization of high biomass forests using remotely sensed data is desirable. However, direct estimation of carbon storage in high biomass forests is difficult for common space borne optical and radar sensors. Airborne light detection and ranging (LiDAR) instruments measure the vertical structure of forests and thus provide alternative for remotely sensed quantification of forest biomass. In this study, the relationships between LiDAR information and aboveground biomass (AGB) measured on the field were investigated. The study covered an area of about 12,000 ha of lowland and hill dipterocarp forests in Terengganu, Peninsular Malaysia. Eight variables were produced from the LiDAR point-cloud namely intensities from all returns, first return, second return, third return, last return, canopy density, digital surface model and canopy height model (CHM). These parameters were converted into raster format with pixel resolution of 1 m. The study found that the CHM is the best predictor variable for AGB and was therefore used for the estimation of AGB in the entire study area. The estimated AGB ranged from 55 to 497 Mg ha-1, which amounted about 2.5 million Mg of AGB for the entire study area with an average of about 206 Mg ha-1 and accuracy at about ±31 Mg ha-1