01. The use of Landsat 8 imagery for mapping natural forest states in Dak Lak province

Abstract

Assessment of forest’s carbon storage level requires up-to-date information on the area of ​​natural forest states. The purpose of this study is to use the Landsat 8 satellite images to create the map of natural forest states in Dak Lak province, which provides the baseline data for estimating forest carbon storage in the province. The supervised classification algorithm in the Image Analysis of the ArcGIS 10.3 software was applied to classify natural forest states in the province. The classification results showed that the Landsat 8 image was suitable for distinguishing natural forest states. Poor deciduous broadleaf forests occupied the largest proportion (15.5%) of deciduous broadleaf forests, followed by very poor deciduous broadleaf forests (8.09%), and medium deciduous broadleaf forests (5.81%). Rich evergreen broadleaf forests, medium evergreen broadleaf forests and poor evergreen broadleaf forests accounted for 14.40%, 13.74% and 9.48% of the evergreen broadleaf forests, respectively. Rubber forest area occupied a small area (5.08%). In terms of overall accuracy, using Landsat 8 images can produce the map of natural forest states by approximately 90% with Kappa coefficient of 0.8 which is acceptable for estimating carbon storage of natural forest states.