Aboveground biomass mapping of La Trinidad forests in Benguet, Philippines, using Landsat Thematic Mapper data and k-nearest neighbor method

Abstract

This study was conducted to estimate the aboveground biomass (AGB) of the La Trinidad forests in Benguet, Philippines, using the following data: Landsat Thematic Mapper, forest map, and field data, utilizing the k-nearest neighbor method. A total of 35 plots were established in the Pinus kesiya forests of La Trinidad, Benguet. The root mean square error and bias of the different k values ranging from 1 to 20 and horizontal reference area (HRA) ranging from 4 km to 7 km were determined. Results of the evaluation showed that the optimum k value and HRA were 4 and 7 km, respectively. These values were used for AGB estimation, and results showed that the mean AGB of the La Trinidad forests was 240.46 Mg/ ha, which is higher than the previous studies that estimated the AGB of Pinus kesiya forests in other areas in the Philippines. Overall accuracy was also determined using the confusion matrix and results showed that the overall accuracy was 56%. It is suggested that the AGB map for Pinus kesiya could provide baseline information for considering the carbon-sink potential of this forest, which is essential for sustainable forest management of tropical forests. © 2014 © 2014 Korean Forest Society.