The estimation of biomass of tropical forests has continually been refined over the past decade, as the methodologies and the technology available for such estimation improved. Consequently, the uncertainty associated with estimates of carbon flux to the atmosphere from changes in land use has declined. In this paper, we review the efforts, primarily within our group, to refine the process of estimating forest biomass across the tropics. Initial estimates of global biomass resulted from the synthesis of ecological studies on small experimental plots. These studies did not sample from the larger population of interest and were conducted in relatively undisturbed locations, so that extrapolations tended to yield overestimates. Later, forest inventories were used to estimate biomass which resulted in better estimates because large sampling units were used. A large effort ensued to maximize the information extraction and reliability from forest inventory data. However, geographic specificity of the estimate was usually limited to national or large subnational unit. This shortcoming led to the use of geographic information systems (GIS) to increase the spatial specificity and to extrapolate inventory information to areas that have not been assessed. We use GIS to assess biomass in Peninsular Malaysia, a country with very good inventories and maps, and in Continental South/Southeast Asia, a region with very sparse inventory information. In the later case, GIS is used to model forest biomass using a suite of map layers that most influence biomass.
Brown, Sandra; Iverson, Louis R. 1992. Biomass estimates for tropical forests. World Resource Review. 4(3): 366-384.