Investigation of spatial and temporal forest cover change using GEE : a case study of Mabira Forest

Abstract

Mabira Forest, a rainforest in Uganda that is situated between Lugazi and Jinja and spans an area of 300 square kilometers (120 square miles) (30,000 hectares (74,000 acres), has been protected since 1932. People have put a lot of strain on the forest over time, which has been linked to the rise in demand for agricultural land and forest products. This study used Google Earth Engine to look at the spatial and temporal changes in the Mabira forest's forest cover. The NDVI was seen to be changing over time using time series and change detection analyses, with the good NDVI (healthy forest cover) being in the early years of 2002 and 2003 and it (NDVI) is gradually decreasing from 2018 and beyond. The Menon and Bawa model and the NDVI approach were used to calculate the deforestation rates for each year during the last 20 years. With 11430 samples, land cover was classified using the Random Forest classifier utilizing the categories of built up, agricultural land, bare ground, and forested land. This was carried out with a training overall accuracy of 87.4753%, and validation was carried out as well.