Vegetation-Water-Built Up Index Combined: Algorithm Indices Combination for Characterization and distribution of Mangrove Forest through Google Earth Engine

Abstract

Mangroves that live in ecotone areas have a fairly significant role in the economy and ecology. This strategic role requires spatial data to facilitate the management and development of mangrove areas. The mangrove mapping process usually uses a manual method, namely through software, and has shortcomings and limitations in image management that require massive data storage. Cloud computing-based Google Earth Engine (GEE) mapping platform can manage images with an extensive scope and spatiotemporal data processing. However, this platform requires index formulas or combinations to help classify and increase accuracy in mapping the earth’s surface. The innovation with the combined VWB-IC (Vegetation-Water-Built-up Index Combined) formula is projected to classify the characteristics of mangrove areas in Jakarta Bay. The combination consists of three types of indices, namely vegetation index (NDVI, GNDVI, ARVI, EVI, SLAVI, and SAVI), water (NDWI, MNDWI, and LSWI), and buildings (IBI and NDBI). This combination is used to translate the classification of mangroves using the Random Forest (RF) machine learning algorithm method with the Sentinel-2 MSI (Multispectral Instrument) satellite image source and through the GEE platform. This platform generates raster data for land use classification (including mangroves), and then the analysis is continued using ArcMap software. The obtained mangrove area is 220.43 ha, located in Jakarta Bay and divided into the Angke Kapuk Nature Tourism Park and the Pantai Indah Kapuk Mangrove Ecotourism Area. The data from this research is expected to provide a recommendation for a combination index formula for mapping mangrove areas in urban areas. The spatial distribution area can be used as an evaluation material in mangrove areas in Jakarta Bay