Author(s): Na Zheng; Zhi Li; Junbo Wang
Linked Author(s):
Keywords: Urban flooding High-performance comuting Kokkos Bidirectional coupled model SWMM
Abstract: In recent years, due to urbanization and the frequent occurrence of extreme climate events, the frequency of urban flood disasters has also been increasing. In studying urban flood disasters, using of hydrodynamic simulation models is a commonly adopted method. These models typically include two-dimensional (2D) surface runoff models and one-dimensional (1D) pipeline models for drainage systems. However, when studying large-scale flood events, computational time becomes a notable obstacle. The application of high-performance computing (HPC) has emerged as an effective means to accelerate the simulation of urban floods. Yet, the diverse range of HPC platform hardware currently available poses challenges for the cross-hardware operation of models. To address this challenge, we propose a high-performance bidirectional coupled model, named SERGHEI-SWMM. This model based on the SERGHEI-SWE and SWMM model, realizes the hydrodynamic characterization by solving the shallow water equations and can accurately simulate the process of urban floods. Additionally, the model employs the Kokkos framework, exhibiting good portability and enabling it to run on different types of HPC platforms without modifying the model source code. We test the accuracy and acceleration performance of the model on an ideal case and a real-world case. SERGHEI-SWMM provides an effective tool for effective simulating urban floods and studying the hydrological and hydrodynamic impacts of multiple factors on the dynamic development of urban floods.
Year: 2025