Flood Inundation Identification in Katingan Regency, Indonesia using SAR Data and Adaptive Dynamic Thresholding

Flooding is the submergence of land by water due to increased water volume. In BNPB's 2021 Kaleidoscope record, out of 3,092 disaster events, 1,298 of them were floods. Floods become one of Indonesia's recurring and frequent disasters. Katingan Regency is one of the regencies in Central Kalimantan with high flood potential. The worst flood occurred in August-September 2021, where the total affected area was 794,100 hectares. Flood inundation monitoring using SAR image remote sensing techniques can be carried out as a flood mitigation effort. Sentinel-1 SAR image is one of the SAR images which can be used in flood inundation mapping due to its acquisition capability for not being limited by time and atmospheric disturbances. One method which can be used for mapping flood inundation areas with Sentinel-1 SAR image data is the Adaptive Dynamic Thresholding method. This method produces better flood inundation mapping than direct or conventional thresholding. This method is based on the image tiling process, HAND (Height Above Nearest Drainage), scene segmentation with Expectation Maximization, and applying the Kittler & Illingworth algorithm. In this research, flood inundation mapping was carried out in four Districts of the Katingan Regency. Sentinel-1 SAR image with acquisition times of July 24, 2021, and September 10, 2021 were used. The results were validated with the BNPB 2021 Pusdalops data as reference data. The results were in line with the reference data.