Hybrid Remote Sensing for Estimating Timber Production and Carbon in Tropical Rainforest

Abstract

Sustainable timber production and global climate change mitigation become important issues in tropical rainforests management around the world, including Indonesia. In this case, the existence of a tropical rainforest is not only directed to stabilize wood supply but also reduce carbon emission in the atmosphere. Estimation of timber production and carbon storage in a tropical rainforest using field inventory requires long-time consuming and high cost. Therefore, an alternative method is proposed to support a more efficient forestry inventory. This study aims to evaluate the potential of remote sensing for facilitating the implementation of forest inventory in a tropical rainforest area. A hybrid approach of remote sensing using two different images resolution, i.e. medium and high was developed to estimate timber production and carbon storage with three predictor variables, namely canopy closure (C), crown diameter (D), and tree density (N). Then, a computational model was constructed from database management systems using a case-based reasoning approach. Results demonstrated that using remote sensing for tropical rainforest inventory provided a good accuracy to estimate timber production and carbon storage with Normalized Root Mean Square Error (NRMSE) around 18%. This study recorded the mean timber production in the study area was 79.91 m3ha -1 with average carbon storage by approximately 14.33 Mg ha -1. Reviewed from these findings, there was an opportunity to use a hybrid approach of remote sensing for supporting forest inventory in the tropical rainforest.