Yuhang H.Wutjanam M.Mahidol University2025-10-272025-10-272025-09-01International Journal of Geoinformatics Vol.21 No.9 (2025) , 120-13416866576https://repository.li.mahidol.ac.th/handle/123456789/112768This study develops and applies a reproducible remote-sensing workflow to analyze coastal erosion and accretion dynamics in Nakhon Si Thammarat Province, Thailand, over the period 1990–2025. Leveraging Landsat 5, 8, and 9 archives on the Google Earth Engine (GEE) platform, the method integrates the Automated Water Extraction Index (AWEI) with Otsu thresholding in a semi-automated pipeline for consistent coastline detection across sandy, muddy, and mangrove-dominated coasts. This approach minimizes subjectivity in threshold selection, harmonizes data across sensors, and allows large-scale, long-term monitoring with reproducibility. Results reveal severe erosion in Pak Phanang Bay and Hua Sai (–3 to –5 m/yr, with localized extremes exceeding –30 m/yr), moderate retreat along southern muddy coasts, and notable accretion at Laem Talumphuk (+32.63 m/yr) and river mouths due to sediment deposition. The workflow not only highlights the drivers of change—monsoon storms, sea-level rise, aquaculture expansion, and coastal engineering—but also demonstrates the capacity of GEE-based methods to provide actionable, repeatable insights for coastal management. This methodological innovation offers a scalable template for monitoring vulnerable coastlines globally, supporting the design of adaptive, evidence-based strategies in the face of accelerating sea-level rise and intensifying climate impacts.Earth and Planetary SciencesPhysics and AstronomySocial SciencesArticleSCOPUS10.52939/ijg.v21i9.44512-s2.0-10501925995826730014