Mapping of land use change using Sentinel-2 Multi Spectral Instrument (MSI) imagery and Google Earth Engine mapping platform at Dramaga, Bogor Regency, Indonesia

Abstract

Dramaga District is an urban buffer area in Jabodetabek Metropolitan Megacity and has various types of land use. The land crisis due to population growth in urban areas has implications for the surrounding buffer areas, so land use changes often occur. Mapping of land use change in the period 2017 to 2021 carried out area is a response to the impact of urban activities and the need for fast and accurate. The mapping platform, Google Earth Engine based on cloud computing, was involved in this research. The land use classification process from Sentinel-2 MultiSpectral (MSI) imagery uses the Random Forest (RF) machine learning algorithm classification method and is assisted by several types of indexes such as vegetation index (NDVI, SLAVI, ARVI, SAVI, and EVI), water index (NDWI, MNDWI, and LSWI), and building index (IBI and NDBI). This study shows that there are changes in several types of land use such as forests, agriculture, built-up areas, plantations, and water bodies. The classification of land use accuracy level reaches 86.00% (Overall Accuracy) and 0.82 (Kappa Statistics) so that this spatial information can be used as consideration for policy makers to determine the allocation and utilization of spatial planning.