Shoreline change detection using DSAS technique: Case of Saint Martin Island, Bangladesh

Abstract

Saint Martin, Bangladesh's only coral island, is a popular tourist destination and a blue economy hub. Saint Martin's shoreline constantly changes due to natural and manmade erosion and accretion. Since no previous work has been done, this research uses geospatial and statistical methods to evaluate spatial and temporal shoreline changes on Saint Martin Island from 1974 to 2021 and anticipated changes between 2032 and 2042. The study used the Normalized Difference Water Index and Tasseled Cap Transformation tools to extract shorelines from the United States Geological Survey (USGS) Landsat satellite images with resolutions of 60 m in 1974 and 30 m in 1988, 2001, 2013, and 2021. Then, the Digital Shoreline Analysis System (DSAS) calculated shoreline change for each year using End Point Rate (EPR), Linear Regression Rate (LRR), and Net Shoreline Movement (NSM) methods. Global Positioning System (GPS) was also used to measure accuracy and shoreline change intensity. Around 253 transects were placed at 50 – m intervals along the island's coast, with approximately 55% of the transects indicating erosion, the remaining indicating accretion, and just 1% in stable condition. The research reveals that Uttar Para experiences the most dramatic shoreline alterations among the five coastal zones, with minimum erosion rates of −2.21 m/yr due to high tidal waves. Aside from that, Dakhin Para has the highest accretion rates, up to a maximum of 4.81 m/yr, due to the coral colonies' protection. Nazrul Para, Purba Para, Majer Para, Deil Para, and Eastern Galachipa erode and accrete moderately. Paschim Para, Uttar Para, and the eastern edge of the Dakhin Para coastline length are exceptionally high-risk areas, according to the Shoreline Length Intensity Index. In contrast, there is a low-risk index on the western side of the Dakhin Para and Deil Para coastlines. The LRR technique yields the best shoreline change estimates, with average erosion at −2.13 m/yr and accretion at 1.4 m/yr across 12.65 km from 1974 to 2021. The Kalman Filter Model (KFM) predicts Saint Martin's coastline will erode at a minimum rate of −2.33 m/yr by 2032 and -2.73 by 2042. Conversely, the coastline could accrete 2.09 m/yr in 2032 and 2.45 m/yr in 2042. Lower NRMSE values of 0.09 provide a highly accurate KFM for shoreline validation methods. Shoreline validation has an NRMSE of 0.16, indicating better agreement between field-derived and satellite-based shoreline positions. Finally, recognizing eroding and accreting hotspots along the Saint Martin Island coast can assist academicians, policymakers, and stakeholders. The research enables coastal researchers to comprehend regional shoreline dynamics.