Flooding has catastrophic effects worldwide. Rapid flood models, such as the height above nearest drainage (HAND) model, have lower complexity and data requirements than traditional models. However, input stage height data are often lacking because most gauged sites only provide estimates of discharge. In addition, performance is difficult to evaluate because flood extent data during extreme periods are often unavailable. Here, we developed a HAND-flood frequency analysis (FFA)-synthetic rating curve (SRC) approach and applied it to a coastal watershed in China. The HAND-FFA-SRC approach demonstrated effective and efficient flood forecasting, and the C values were 1.19, 1.20, and 1.16, respectively for the floods under moderate rainfall, heavy rainfall, and storm. Meanwhile, the accuracy of the model was highly impacted by the topographic characteristics of the watersheds. The C values were improved as the slope increased from 3° to 20° during the floods under different scenarios. Additionally, the effects of floods were evaluated under different return periods which indicated that the cropland is the most affected land use type but the risk for impervious surfaces is increasing. The proposed approach is viable for forecasting flood susceptibility and can improve resilience planning and flood management.
Read full abstract