Fractional Floodwater-pixel Fusion for Emergency Response using ALOS-2 and Sentinel-1 data

Emergency hazard and risk mapping services are crucial in providing rapid-response information for disaster management and reducing damage of water-driven cascading disasters such as floods. For rapid-response flood mapping, this study introduces an improved flood detection algorithm to fuse two different C-band and L-band Synthetic Aperture Radar (SAR) images acquired over the same area of interest, e.g., mega-floodplain. The main objective of this study is to propose a new algorithm for dynamic flood detection using two different SAR images acquired at different times and under different conditions. As an image fusion technique, we propose an improved floodwater detection algorithm using the pixel-based water fractional fusion and wavelet-based image fusion. This approach allows investigating the optimization of fusion parameters from two different products in the case of a premonsoon flash flood. This preliminary study was conducted to identify and estimate large-scale flood inundation area over a challenging area, where flash flood events often occur along the Maghna River in northen Bangladesh. The resultant fused map of the maximum flood-detect extent suggested the possibility of a rapid and accurate floodwater detection to identify the flood location and extent area with the validation data from ground-truth survey in the representative experiment sites.