Identifying tree characteristics to determine the blocking effects of water conveyance for natural flood management in urban rivers

This study aims to present a methodology to improve the performance of hydrodynamic modeling by considering plant resistance with the goal of analyzing flood risk reduction strategies that support ecological conservation enhancement. An unmanned aerial vehicle was used to collect floodplain topography, landcover, and tree characteristic data. A modified maximum likelihood classification scheme was developed by incorporating the tree height information to improve the landcover classification. The water-blocking effect of trees, which represents the reduced flood conveyance area, was investigated by establishing tree obstructions in hydrodynamic models. Two hydrodynamic models were established with different tree obstruction setups and Manning's n values, including blocked obstructions (BOs) and conventional high n-values (Adj-n). Among the model setups, the Manning's n values of the trees remained the same as that of the soil in the BO model, in which trees were modeled as blocking obstructions. The n-values of trees were determined to be 0.022 in the BO model and refined to 0.055 in the Adj-n model after the model calibration and verification processes.

The results indicated that the simulated flood levels of the BO model were very similar to those of the Adj-n model. The NSE values were 0.98 and 0.97 in the simulations of two historical typhoon events, indicating that the BO model could obtain reliable predictions without altering the n-values. The verified BO model was used to evaluate the degrees of influence of tree heights, tree coverage areas, and tree locations on flood levels. Several nature-based solutions were proposed to analyze the tradeoffs between reducing flood prevention and enhancing wetland restoration using the verified BO model, suggesting a win–win strategy for natural river management.