A study on improving DEM-based forest biomass estimation using PALSAR and forest inventory data

Abstract

The purpose of this study is to estimate forest biomass using ALOS PALSAR imagery and forest inventory data. Above all, it is only appropriate for farmland which is small biomass values in the case of SAR Sigma-nought(σ <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">0</sup> ), and is affected by shadow of bidirectional effects. On the other side, the terrain slope parameters are calculated using an estimation of the gradients in the two directions of the coordinate plane of the Digital Elevation Model (DEM) and so reduces the shadow phenomenon of topographic effects and revises it. To estimate forest biomass, topographic effects should be considered in SAR data processing, because most forests are located on mountainous areas in Korea. This study is for region of Unduryeong forest, and divided directions using DEM. That is, revision of directional effect by incidence angle is needed therefore SAR images are separated in fore-slope(here, west-slope) and far-slope(here, east-slope). As a result, the gamma-nought(γ <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">0</sup> ) values are always lower compared to the sigma-nought values at all polarizations for Unduryeong forest areas of the fore-slope and are always higher in case of far-slope. Backscatter coefficient shows saturations in about 302.56ton/ha, and it is useful in REDD++ mechanism because forest biomass values based on estimated gamma-nought include the information of location coordinate.