Oil spill impacts on mangrove forest from satellite remote sensing

Abstract

The mangrove forest has been continuously threatened by oil spills occurring on the sea surfaces. The oil spills pose and cause severe and long-term effect havoc on mangrove forests that sustain them. Previous research has found that satellite remote sensing technologies are one of the most effective techniques to detect oil spills and assess the health of mangrove forests in contaminated areas. This study utilized the Synthetic-Aperture Radar (SAR) images from dualpolarized Sentinel-1 and Multi-Spectral Instrument (MSI) from Sentinel-2 to study the impact of oil spills on Mangrove Forest in Pantai Cermin, Negeri Sembilan. The Random Forest classification was used to detect the oil spill areas, while vegetation indices were used to assess the impact of oil pollution on mangrove forests in the early stages. Analysis from Sentinel1 imagery shows that the oil spill could be accurately detected using the Random Forest classifer with accuracy of 76%. Spectral indices: the normalized difference vegetation index (NDVI) was explored and evaluated to study the health of mangrove forest after the oil spills event. It is found that the oil spills have caused physical suffocation as well as toxicological effects to the mangrove forests.