Comparative Evaluation of Different Spatial Resolution Data for Timber Volume Estimation

This study compares the data available from different satellites having different spatial resolutions to estimate the timber volume. The study comprises mainly the stratification of forest from the input of classified satellite data, measurement of timber volume in the field, extraction of the pixel information for the plots, preparation of a continuous volume map and its comparison among different outputs. Regarding this study, the image processing in the study consists of two phases i.e. Pre Processing and Post Processing. The sample size for the strata has been determined analysing the variance of timber volume within strata which was assessed with the Pilot Inventory prior to the final forest measurement. The field measured plot-wise timber volume, extracted pixel information; stratification information and the geographical information have been tabulated and were analysed with regression analysis to find the best-fit regression equations in all sets of data. All PAN sharpened including LISS III MSS subset images have been multiplied by corresponding best-fit linear regression equations to prepare the continuous volume maps. The contribution of remote sensing data for timber volume estimation has been compared with respect to the R-Square values obtained from the regression analysis in other way. The study concludes that there is not only the effect of spatial resolution on the estimation of timber volume but also the spectral resolution of satellite data seems to have contributed more because the distortion produced by PAN sharpening of the image deteriorates the interpretability of the image.