Coastlines estimation and erosion rate assessment in Tuba Island, Langkawi using remotely-sensed digital imageries analysis

Abstract

Malaysia’s coastline is constantly exposed to ocean threats, resulting in coastal erosion and sea-level rise. Pulau Tuba’s is a part of Pulau Langkawi which known as a popular tourist destination in Malaysia. This study focuses on using remote sensing and Geographical Information System (GIS) tools to examine changes along the coastline of Pulau Tuba, Langkawi, during a six-year period. In this research, changes to the coastline were analyzed using Sentinel-2 imagery from 2016 to 2021, as a result of the comparison of coastline extraction methods. Coastline extraction methods used in this study include 3x3-5x5 and 7x7 edge detection filtering techniques. Image classifier of Maximum Likelihood (ML) and Support Vector Machine (SVM) were also used to quantify the estimated coastlines length. Digital Shoreline Analysis System (DSAS) was used to calculate coastline erosion rate from Sentinel-2 images. The comparison between method to extract coastline revealed that the best window size of edge detection image filtering method is 7x7 and SVM outperform MLC classifier. The erosion rate as estimated from DSAS statistical calculation also highlighted that come parts along the Pulau Tuba coastlines experiencing alarming rate of coastal erosion. The authority needs to plan for soft and hard structure to mitigate the impact of coastal erosion in the future.