Forest above-ground biomass estimation from vertical reflectivity profiles at L-Band

Abstract

Forest height is an important parameter for the allometric estimation of above-ground forest biomass (AGB). However, variable forest stand densities limit the performance of the allometric estimation of AGB from height measurements alone. Recently, the use of vertical forest structure information as an indicator for the variation of stand density has been proposed and used to improve the allometric estimation of AGB from height measurements. In this work, the use of vertical radar reflectivity profiles at L-band obtained from SAR tomography, as a proxy for vertical forest structure for the allometric estimation of AGB, is investigated. L-band reflectivity profiles, reconstructed from data at different polarisations (HH and HV) and acquired under “moist” and “dry” weather conditions, are investigated. The proposed allometric AGB estimator increases the correlation factor from 0.60 to 0.81 and reduces the RMSE from 50.25 Mg/ha to 36.30 Mg/ha when compared to the AGB estimation from forest height alone. The effect of polarisation and weather conditions on the AGB estimation performance is discussed.