Evaluating Flood Extent Using Synthetic Aperture Radar (SAR) and Modified Normalized Difference Water Index (MNDWI) Methods

Due to climate change and human activities, flooding has been significantly increasing in both intensity and spatial coverage. The upper Awash River basin Tefki area is one of such areas which is frequently affected by flooding. The aim of the study is to detect and map the flood extent and its evolution from a series of radar information, Sentinel-1 SAR, and optical Sentinel-2 images, as well as to evaluate how the delineated flood areas fit with hydrodynamic modeling results. The SAR and optical signatures of land and water targets were evaluated prior to the thresholding to identify land and water classes. Backscatter properties of water and non-water features on the SAR image were tested, and an appropriate change detection threshold value was set by the Otsu analysis. Based on the threshold value, changed and unchanged areas were identified for delineating an inundated zone. Change detection algorithms were applied to extract the flood extent from the processed SAR images. On the other hand, the modified normalized difference water index (MNDWI) was employed to map flooded areas from the Sentinel-2 optical image. Optical images with no cloud cover were better for extracting flood due to their nature of rich information from several spectral ranges. Although SAR band-c images are capable of penetrating cloud and rainfall, the results were over-influenced by turbidity and suspended materials that cause diffuse reflection. The findings from our SAR and optical images are promising, with approximately 96% alignment to GPS-enhanced flood marks and a comprehensive representation of all flood-affected areas through hydrodynamic modeling. © The Author(s), under exclusive licence to Springer Nature Switzerland AG 2025.