A Distributed Unit Hydrograph Modeling for Flood Simulation in the Plain River Network Regions

Abstract

This study proposes a distributed unit hydrograph (DUH) method to address the challenge of simulating overland flow concentration in plain river network regions. The DUH framework defines generalized river network polygons (RNPs) to represent flow convergence zones and estimates runoff travel times based on a calibrated confluence velocity parameter, circumventing the need for high-resolution topographic data. The method was applied to the Taihu Basin, where 16 subregions were analyzed under different spatial scales and overland flow velocities. Results show that the DUH method significantly enhances model performance compared to the traditionally used proposed unit hydrograph (PUH) approach. Specifically, DUH reduced the root mean square error (RMSE) of simulated water levels by up to 40%, improved the coefficient of determination (R²) by 0.1–0.2, and reduced the average flood peak lag from 2.1 days to 0.7 days. The model exhibited optimal accuracy at a grid scale of 200 × 200 m, achieving a balance between smooth hydrograph formation and computational efficiency. These findings underscore the DUH method's applicability for flood simulation and decision-making in low-relief, hydraulically complex regions with limited microtopographic data availability.