<![CDATA[Comprehensive spatial impact analysis integrating flood detection with land use classification has remained limited despite the occurrence of escalating tidal flooding that threatens coastal communities, aquaculture zones, and agricultural lands in the Sayung District. This study aimed to assess and map the spatial distribution of tidal flooding and analyze its impacts on land use through integrated remote sensing and artificial intelligence methodologies. The change detection and thresholding (CDAT) method utilizing Sentinel-1A imagery was employed for tidal flood detection, whereas the U-net deep learning architecture with stacked Sentinel-1A and Sentinel-2A data was implemented for multiclass land use classification. The CDAT method successfully detected 14.636 km² of inundated area, representing 12.25% of the Sayung District, with an overall accuracy of 0.886 and a kappa coefficient of 0.755. The U-net model classified 11 land use categories with an overall accuracy of 0.81 and a kappa coefficient of 0.78, demonstrating robust performance, particularly for aquaculture ponds and cultivated aquatic, with IoU values exceeding 0.88. The spatial overlay analysis revealed that aquaculture ponds constituted the most extensively affected land use category (6.09 km ², or 41.28 %) followed by cultivated aquatic (30.89 %) and water bodies (18.62 %). The integration of CDAT and U-net provides a comprehensive analytical framework for tidal flood impact analysis, thereby generating essential spatial data for data-driven disaster mitigation strategies and coastal zone management policies in vulnerable coastal environments.]]>
