Correlation analysis between forest volume and LANDSAT TM data based on sub-compartment objects

Abstract

The correlation of forest volume with LANDSAT TM data and its derivative data was discussed here. The forest resource field inventory data and simultaneous LANDSAT TM data were used here in Shimian county, Sichuan province, P.R.of China. A lot of derivative data were created from LANDSAT TM data by vegetation index calculation, principal component analysis, and tasseled cap transformation. 1204 forest sub-compartments with inner homogeneity were used as samples for correlation analysis. It was found that TM1, TM2, TM3, TM4, TM5 and TM7 were negatively and significantly (99%) related to the pixel forest volume on forest sub-compartment object. The highest is -0.307, which was the correlation coefficient of the TM1 with pixel forest volume. All derivative data except wetness of tasseled cap transformation were negatively and significantly (99%) related to the pixel forest volume, but wetness was positively and significantly (99%) related to the pixel forest volume.