Resilience Assessment of Flood Detention Zones in the 2023 catchment-scale floods in Hai River Basin, China

As a typical flood-prone area, the Hai River Basin (HRB) in the Beijing-Tianjin-Hebei metropolis of China has been struck by devastating floods in history. Since the 1960s, a series of flood-control programmes in the HRB have been launched to reduce the flood risks. As planned, flood detention zones serve as the last line by storing and detaining floodwaters when the water levels exceed the defense limits of reservoirs, levees, and diversion channels. Land use crisis has been a long-lasting problem in China. People are allowed to use the flood detention zones as their residential communities when these zones are not in use. A dual role played by these specific zones requires not only an effective floodwater storage in response to floods, but also an efficient floodwater recession in the aftermath of floods. However, we lack a quantitative assessment of the functionability of flood detention zones. Our study synergizes multi-source SAR images from Sentinel-1 and Gaofen-3 satellites in the framework of deep learning to accurately and efficiently extract inundation paths which evolved for two months encompassing HBR. A joint use of digital elevation model allows us to recover the three-dimensional inundation structures. We also propose the flood detention resilient coefficient based on our derived lifespan of floodwaters. Our results demonstrate that the flood detention zones in HRB can effectively trap the floodwater within to secure lives and properties, but resilience of some flood detention zones can still be improved.