Coastal dynamics: Assessing erosion and progradation patterns in Campeche coastal region using machine learning techniques for geological insights

Abstract

This study examines shoreline dynamics along the Campeche coastal region from 1974 to 2024, using machine learning techniques to analyse long-term trends in erosion and accretion. Integration of machine learning with geological analysis enables precise evaluation of shoreline changes, enhancing the understanding of coastal dynamics. Spanning 50 years, the research utilizes satellite imagery, historical maps, remote sensing, and GIS to assess shoreline changes and their geological significance. Results show that 93% of the coastline is experiencing accretion, while 7% is undergoing erosion. Accretion rates range from a minimum of 2.36 m (0.05 m/year) to a maximum of 1136.61 m (22.73 m/year), while erosion rates vary from −2.56 m (−0.05 m/year) to −131.41 m (−2.63 m/year). This study identifies key coastal changes, with Zona Centra and Koben regions experiencing significant accretion, while South San Lorenzo faces the highest erosion activity. Understanding these dynamics is crucial for coastal management and mitigation strategies. Campeche's coastline predominantly features carbonate-rich sediment deposits, influenced by ocean currents, sediment supply, and coastal morphology, resulting in widespread accretion. Localized erosion hotspots are linked to human activities and disruptions in sediment transport. Machine learning models identified spatial patterns and quantified the influence of stratigraphy on shoreline changes, highlighting the benefits of combining modern techniques with traditional methods for accurate coastal analysis. Campeche's rich geological diversity and extensive data availability make it an ideal case study for exploring machine learning applications in coastal management and assessing coastal vulnerability. This comprehensive approach offers valuable insights for sustainable development and addressing coastal erosion challenges.