Monitoring agroforestry for REDD+ implementation using remote sensing data and geographic information system: A case study of Repong Damar, Pesisir Barat Lampung

Abstract

Repong Damar is a unique practice of agroforestry found in Lampung that has a canopy structure nearly similar to that of natural forest. Apart from its uniqueness and important roles, effort to map and timely monitor the condition over Repong Damar is very limited. This research aims to analyze the most appropriate and accurate method for detecting Repong Damar using satellite images and to understand the historical change of Repong Damar cover. We tried to relate this effort to REDD+ potential implementation through agroforestry by estimating the forest reference level (FRL) of Repong Damar. Three methods of detection were used i.e. Object Oriented Classification (OOC), Maximum Likelihood Classification (MLC) and Vegetation Indices Classification (VIC). The most accurate method for detecting Repong Damar was the OOC. By using this method, time-series change of Repong Damar from 1990 to 2018 was determined. FRL was established by multiplying carbon stock of Repong Damar and the time-series average coverage area of Repong Damar from 1990 to 2015, i.e. 33,187,752 tC/yr. The emission performance of Repong Damar in 2018 was calculated from the total carbon stock of Repong Damar in 2018 against FRL. Repong Damar has emitted 1,485,378 tons of carbon in 2018.