Google Earth Engine for Landsat Image Processing and Monitoring Land Use/Land Cover Changes in the Johor River Basin, Malaysia

Abstract

Johor River Basin (JRB), located in the state of Johor in southern Peninsular Malaysia is an important ecosystem providing freshwater and supporting people of the rapidly developing Johor and Singapore. Land use and land cover (LULC) of JRB is changing rapidly due to high economic activities and increasing population in the zone. A 30-years LULC analysis of JRB was conducted in this study by comparing two machine learning classification algorithms, i.e., Support Vector Machine (SVM) and Random Forest (RF) using Google Earth Engine (GEE) as the processing platform. While cloud cover is always a significant issue over the tropical region, the big data and cloud processing capabilities of GEE can provide advantages to generate cloudless Landsat image composites from multiple scenes. The classification results showed high accuracy of 87% and Kappa value above 0.84 respectively. The LULC classification results can serve as the base map for further studies, including the river morphology change analysis, the ecosystem services analysis; and support hydrology modelling, land use policy making, and water resources management.