MONITORING FOREST ABOVE-GROUND BIOMASS OF GUJARAT STATE USING MULTI-TEMPORAL SYNTHETIC APERTURE RADAR DATA

摘要

Synthetic Aperture Radar (SAR) data has shown great potential in retrieval of forest above-ground biomass (AGB) due to the capability of SAR to provide more dynamic range for vegetation growth variables as compared to optical data. Estimations of forest AGB of Gujarat state was carried out for multiple years using C-band Radar Imaging Satellite-1 (RISAT-1) and L-band Advanced Land Observing Satellite Phased Arrayed L-band Synthetic Aperture Radar (ALOS-PALSAR 1/2) data. In the present study, topographically corrected Medium Resolution ScanSAR (MRS) data of Indian RISAT-1 acquired during 2015-16 and global SAR mosaic products in HH/HV polarizations produced from Japanese ALOS-PALSAR 1/2 data for the years 2007-10 and 2015-16 were used to retrieve temporal forest AGB of Gujarat through semi-empirical model based on multi-linear regression coefficients of HH and HV polarization backscatter with field measured forest biomass. Gujarat has four major forest types namely, (1) tropical moist deciduous forest, (2) littoral and swamp forest, (3) tropical dry deciduous forest and (4) northern tropical thorn forest. Different model coefficients were derived for these forest types based on extensive ground measured forest parameters and the biomass maps of Gujarat were generated. High correlations were observed between γ°HV and γ°HH/HV with field measured biomass over different forest vegetation types with biomass densities ranging from 20-120 t/ha. The study has also presented the advantages and limitations of C and L-band SAR data for estimation of forest AGB with varying biomass densities and has demonstrated how selection of suitable observation period of SAR data enhances retrieval of AGB of deciduous forests.