Effect of Slope on Estimating Tree Biomass using Satellite Data in Belasitsa Mountain Forest, Bulgaria

Abstract

Remote sensing of above ground tree biomass (AGTB) in rugged and mountainous forest sites incurs uncertainties in accuracy due to the influences of various bio-physical and topographic factors. Particularly slope is an important factor affecting the accuracy of remote sensing AGTB. This study investigates how slope affects the estimation of above ground tree biomass using vegetation indices (VIs) calculated from an ASTER image in the Belasitsa Mountain forest in Bulgaria. Field sampled AGTB was put in regressions with several VIs and physical variables like slope, altitude and forest types. No significant relation was observed between VIs and AGTB in simple linear regression. However, multiple regression yielded better results. DVI (Difference Vegetation Index) predicted AGTB better (multiple r2 = 0.35) compared to other VIs. Steeper slope had a significant effect (at p<0.01) on AGTB estimation in multiple regression. Although the coefficients of determinations (r2) of the regressions were low, this study provides an insight on how terrain affects remote sensing AGTB in a mountainous mixed forest site.

Keywords: Above ground tree biomass (AGTB), mountainous forest, slope, vegetation indices (VIs), and regression analysis