Benthic Habitat Mapping Using Sentinel-2A Imagery in the Waters of Pengudang Village, Bintan Regency

Abstract

Pengudang Village has a diverse shallow-water ecosystem that forms the seabed and not much research has been done to map benthic habitats in these waters. Thus, the availability of spatial data regarding benthic habitats in the area is very limited. This study aims to map shallow water benthic habitat using Sentinel-2A imagery and calculate the accuracy of benthic habitat classification in Pengudang Village waters. Benthic habitat classification is carried out in 4 cover classes, namely      seagrass (LM), sand (PS), sand with dead coral (KMP), and living coral with sand (PKH). In this study, we used supervised pixel-based image classification using two algorithms, namely Support Vector Machine (SVM) and Minimum Distanse Classification (MDC) with thematic layer input from field data. The results showed that the SVM classification algorithm model had a better accuracy of 76% while the MDC algorithm model had an accuracy of 66%. Based on the results of the study, the difference between the two classification algorithms is 10%, and the best classification algorithm in the waters of Pengudang Village is SVM.