Identification of mangrove forest change using object-based analysis in Bontang, East Kalimantan

Abstract

In past half century, 35-86% of global mangrove forest already loss due to conversion into ponds, urban development and timber harvest. This is become challenging because in some area we don't have data about distribution and what mangrove species in that forest. We mapping mangrove forest in Bontang city using PlanetScope images acquired at 2017 and 2018. Remote sensing analysis using OBIA method we obtain mangrove extend in Bontang City in 2017 and 2018 are 2,659.2 ha and 3,155.9 ha respectively. With this result, we already loss 491 ha or 13.5% from the initial extend in 1988. Our finding on this research are although have high accuracy, mangrove mapping with 3-meter resolution gave not quite good result. This is because we only obtain mangrove and non-mangrove classification, not until species determination. The factor mainly atmosphere condition which Bontang area have cloud or many aerosols that hamper light transmission. In future need to mapping with species differentiation with this challenge.