Research on Quantitative Evaluation of Forest Biomass Based on MultiSource Remote Sensing Data

Based on ALOS PALSAR Lband dualpolarization FBD microwave remote sensing data and multispectral optical remote sensing data AVNIR2, preprocessing was conducted, then three microwave data factors including HH and HV dualpolarization backscattering coefficients and polarization ratio, and four original bands and two vegetation indices of NDVI, RVI which combined with optical remote sensing data were extracted with the coordinates of Pinus yunnanensis forest sample plots. After that, the three estimation models of Pinus yunnanensis forest biomass were built by using microwave data, optical data, and the multisource remote sensing data combined with microwave and optical data. Research suggests that and the models reached the significant or extremely significant correlation level by variance analysis. PALSAR Lband dualpolarization backscattering coefficient could be reflected the changes of forest biomass, but inversion accuracy needs to be improved. AVNIR2 data model was better than PALSAR Lband dualpolarization data model. The estimation accuracy of multisource data model was similar to optical data model.