Tropical Peatland Forest Biomass Estimation by Exploiting Polarimetric Parameters in Synergy with in-Situ Data

Abstract

Polarimetric SAR (PolSAR) data acquired using space-borne sensor perform excellently for forest biophysical parameters estimation. Among the forest biophysical parameters, forest aboveground biomass (AGB) is an important parameter describing all living biomass in the forest. In this research, AGB estimation is carried out by exploiting polarimetric parameters extracted from RADARSAT-2 quad-pol data in synergy with AGB inventory data. Six consecutive RADARSAT-2 quad-pol acquisitions under dry, humid and rainy weather were acquired from Canadian Space Agency (CSA). Polarimetric parameters (backscattering coefficient (σo), eigen decomposition parameters (entropy and α), Yamaguchi decomposition parameters (surface, volume scattering components) and parameters describing canopy structure) were extracted using acquired SAR data. Reference biomass data were extracted from referenced biomass map and field visit in study site. Regression analysis was carried out between the extracted polarimetric parameters and in - situ biomass data collected from reference biomass map. Significant regression models were inverted and biomass estimation was performed. Resultant biomass maps were validated using in - situ data collected through field visit.