Machine learning-based detection of changes in mapping the mangrove forest of the Yangon estuary, Southeast Asia

Mangrove forests are globally acknowledged for stabilizing coastlines, reducing wave energy, and protecting coastal habitats and adjacent land uses from extreme events. However, most regions experience alarming mangrove loss against natural and human disturbances. This study profiles dynamic changes in mangrove cover and shoreline migration along the Yangon estuary using Landsat imagery and machine learning approach from 1988 to 2023. Mangrove cover declined from 1175 ha in 1988 to 531 ha by 2011. It then increased to 5470 ha by 2023, resulting in a net gain of over 4000 ha. Concurrently, shoreline analysis using the mangrove vegetation line, indicates 92 % seaward progradation along the coastline. The western shoreline recorded mean accretion and erosion rates of +35.6 m/yr and -1.7 m/yr, while the eastern side showed more dynamic rates of +79.6 m/yr for accretion and -29.1 m/yr for erosion. Key findings highlight mangroves' ability to keep pace with the relative SLR, aquaculture as the dominant driver of post-2008 mangrove loss, and underscore the roles of sedimentary variation and high sediment availability, extensive tidal flat existence, and coastal sheltering in supporting recent mangrove expansion. While further studies are needed, these insights offer a valuable foundation for future conservation and management efforts.