Abstract Understanding the dynamic changes in aquaculture is crucial for effective coastal resource management, particularly in rapidly evolving deltaic regions. This study investigates the spatiotemporal dynamics of aquaculture ponds in Thi Nai Lagoon, Vietnam, across 2020, 2022, and 2024. Using multi-temporal high-resolution PlanetScope imagery, annual aquaculture extents were delineated using an object-based image analysis (OBIA) approach integrated with the Random Forest algorithm. This method achieved high accuracy, peaking in 2022 (with an overal accuracy of 90.95% and a Kappa coefficient of 0.80). The analysis revealed a 5.2% reduction total aquaculture pond area, from 776.01 ha in 2020 to 735.71 ha in 2024. Spatially, aquaculture remained concentrated in the southern and western communes, where pond loss occurred in urban-adjacent areas such as Nhon Binh and Phuoc Thuan. GIS-based overlay analysis enabled the precise identification of localized ’gain’ and ’loss’ zones, providing critical insights into patterns and potential drivers of changes. These findings highlight the value of multi-temporal satellite data for monitoring rapid land-use transformations. This study’s methodology and results support adaptive environmental management by informing strategies to mitigate habitat loss and promote sustainable aquaculture in coastal ecosystems.
Dung et al. (Mon,) studied this question.