Estimation of Soil Surface Characteristics in Peatland Using ALOS/PALSAR

Abstract

For this study, we applied data of ALOSIPALSAR, L-band Synthetic Aperture Radar (SAR), to analysis of a peatland site, Sarobetsu Mire, Hokkaido, where environmental rehabilitation has been attempted. Using data of 10 scenes that were taken during spring to fall in 2006-2007, we analyzed the relationship of the backscatter coefficient to various factors related to soil, hydrology, vegetation, and geography.
Results of regression analyses with single factors clarified significant relationships for factors including dry bulk density, the rate of nitrogen content, the C/N ratio, the vegetation proportion, and NDVI, revealing the characteristics about scattering and incident angle that are considered peculiar to peatlands. Furthermore, the result of multiple linear regression analysis revealed that the dry bulk density of the surface soil was the dominant contributing factor, enabling construction of a multiple linear regression model, which is significant for estimating the spatial distribution that is peculiar to the peatland surface soil using L-band SAR.