期刊名称：ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences
印刷版ISSN：2194-9042
电子版ISSN：2194-9050
出版年度：2002
卷号：XXXIV Part 4
出版社：Copernicus Publications
摘要：Above-ground biomass estimation of successional and mature forests in moist tropical regions is attracting increasing attention. Because of co mplex stand structure and abundant vegetation species, rarely has remote-sensing research been successfully conducted in biomass estimation for moist tropical areas. In this paper, two study areas in the Brazilian Amazon basin—Altamira and Bragantina— with different biophysical characteristics were selected. Atmospherically corrected Thematic Mapper (TM) images and field vegetation inventory data were used in the analysis, and different vegetation indices and texture measures were explored. Multiple regression models were developed through integration of i mage data (including TM bands, different vegetation indices, and texture measures) and vegetation inventory data. These models were used for biomass estimation in both selected study areas. This study concludes that neither TM spectral bands nor vegetation indices alone are sufficient to establish an effective model for biomass estimation, but multiple regressio n models that consist of spectral and textural signatures improve biomass estimatio n performance. The models developed are especially suitable for above-ground biomass estimation of dense vegetation areas
关键词：biomass estimation; successional and mature forests; Amazon basin; ; Thematic Mapper; vegetation index
