Estimating tree volume in tropical dry deciduous forest from landsat TM data

Abstract

Abstract Mean whole tree volume was estimated from spectral variables derived from Landsat Thematic Mapper (TM) data, in part of Hunsur forest division, Karnataka, India, comprising tropical dry deciduous forest using direct and indirect approaches. Mean crown diameter was considered to be the intermediate variable in the indirect approach. In the direct approach, the multiple regression equation developed to estimate mean whole tree volume consisted of radiance in TM band 7 and indices (R3 ‐ R1/R3 + Rl) and (R5 ‐ R4/R5 + R4), where R represents spectral radiance values. The mean whole tree volume estimated by both approaches were compared using mean absolute deviation (MAD) and mean square deviation (MSD). It was observed that MAD and MSD were low and comparable for the direct approach and indirect approach. The present study indicates that it is possible to estimate the mean whole tree volume in dry deciduous forest using satellite sensor derived spectral variables.