USING SPOT-5 IMAGE AND GIS FOR ESTIMATING AND MONITORING BIOMASS AND CARBON IN EVERGREEN BROADLEAF FORESTS OF THE CENTRAL HIGHLANDS

Abstract

                           This paper describes the potential of applying methods to estimate biomass and carbon using SPOT-5 data and GIS for forests in the Central Highlands. The results have been presented as follows: i) Based on the results of unsupervised image classification to establish the relationship between total above-ground tree biomass (TAGTB) and the code of class represented as class-Id for estimating biomass with the reliability from 72-93%; ii) The regression relationship between TAGTB and digital number (DN) with the reliability of 53%; and iii) The overall

accuracy of 29% has been found with the method of supervised classification accoding to biomass stratification. The above results show that the best estimation can be achieved by using the method of unsupervised classification, and then building the relationship between the biomass and the class-Id. It is also indicated that the combination of image –based estimation of biomass and biometric model for estimating biomass and forest carbon using GIS is the potential to monitor CO2 absorption or emission of forest over time.