Applications of the new Remote Sensing Method to the Forest Biomass Estimation

For accurate measurement of forest biomass in the Akazawa Forest Reserve, this study analyzed texture measures derived from GeoEye-1 satellite data using the individual tree crown (ITC) method. On this basis, canopy area, tree tops and tree species of individual trees were delineated. Canopy area was used to calculate the DBH of trees in canopy layer based on canopy-DBH curve in this stand. In this study, the estimation models, between DBH and height, and between canopy area and DBH were developed by linear regression using forest survey data. Then according to the results of satellite data interpreted the biomass of every tree was calculated by biomass expansion factor (BEF). This method was verified against the survey data from old–growth Chamaecyparis obtusa stand composed of various cover types. For Chamaecyparis obtusa, the accuracy of biomass estimation was higher than 84%. However, the accuracy of Chamaecyparis pisifera was less than 60%, because some Chamaecyparis pisifera trees were misidentified as Chamaecyparis obtusa, and canopy area of Chamaecyparis pisifera was underestimated in the high-density stand. For Thujopsis dolabrata, the accuracy ranged from 22.4 % to 78.9%, and from 63.4% to 84.6% for broad-leaved trees, because many of them were understory. These results indicated that estimation of old-growth forest biomass based on high resolution satellite data, might be validated for estimating biomass at the individual tree level improved by developing and applying forest stratum–specific models with the ITC-survey data as a bridging reference in addition to spectral information. This approach is useful for biomass estimation whether is used to calculate biomass of individual tree or forest.