Texture analysis of PALSAR mosaics for forests carbon stock estimation in central Sumatra

Abstract

Retrieving aboveground forest carbon stocks (AFCS) accurately in tropical region remains challenging due to forest structure complexity and species diversity. In order to improve AFCS estimates, the present study evaluates the potential of high-resolution ALOS/PALSAR mosaics data in the tropical forests in central Sumatra. The study region with AFCS range of 1 to 334 Mg ha-1 consists of natural and plantation forests. Field measurements of AFCS were carried out in 87 plots. Various textures of dual polarized mosaics data for the year 2009 and 2010 were computed and assessed their potential for estimating AFCS applying regression modelling. R2, p-value, variable inflation factor, and root mean square errors (RMSE) were examined. Potential models were cross validated by leave- one-out (LOO) method. The result indicates that simple textures analysis in PALSAR mosaics increases potential of AFCS estimation and reduces errors to 30.5 Mg ha-1. The method and models presented in this study can be a low cost wall-to-wall forest carbon mapping with high level of accuracy in tropical forests in Southeast Asia where other methods are still rare.