Performance Analysis of Iteration-Based Dynamic Flood Routing Model for Dendritic Channel Networks

An effective flood forecasting system requires dynamic routing to accurately predict floods under diverse conditions, from slow- to fast-rising events. However, for operational forecasting, dynamic routing is computationally demanding, especially for large channel networks, necessitating efficient solution methods. The integration of a dynamic model into an operational routing environment must ensure reliable predictions and cost-effectiveness. This study compares the performance of two dynamic flood routing models: (1) the iteration-based junction point water stage prediction correction (JPWSPC) model developed in-house; and (2) the hydrologic engineering center–river analysis system (HEC-RAS) model, developed by the US Army Corps of Engineers Hydrologic Engineering Center. Performance is evaluated for three scenarios: (1) a single river reach; (2) a confluence of three channels; (3) a 15-channel network, with varying inflow peaks, bed slopes, and widths under subcritical flow conditions. Results show that for large river systems with backwater effects, JPWSPC is three to nine times faster than HEC-RAS, depending on channel and flow characteristics. Across a wide range of subcritical hydraulic conditions, the JPWSPC model demonstrated comparable accuracy to the HEC-RAS model while significantly improving computational efficiency. This study demonstrates that the modularized, iteration-based JPWSPC model maintains stability under large-scale river network scenarios and may offer computational advantages over the global solution used in HEC-RAS, making it a promising framework for comprehensive forecasting.