Capturing Urban Pluvial River Flooding Features Based on the Fusion of Physically Based and Data-Driven Approaches DOI Open Access

Chenlei Ye,

Zongxue Xu, Weihong Liao

et al.

Sustainability, Journal Year: 2025, Volume and Issue: 17(6), P. 2524 - 2524

Published: March 13, 2025

Driven by climate change and rapid urbanization, pluvial flooding is increasingly endangering urban environments, prompting the widespread use of coupled hydrological–hydrodynamic models that enable more accurate flood simulations enhanced forecasting. The simulation method for river caused heavy rainfall has garnered growing attention. However, existing studies primarily concentrate on prediction using hydrodynamic or machine learning models, there remains a dearth comprehensive framework couples both to simulate temporal evolution changes. This research proposes novel simulating integrating physically based with deep approaches. sample set through data augmentation Generative Adversarial Networks, scheduling control signals are incorporated into encoder–decoder architecture results demonstrate strong model performance, provided model’s structural complexity aligned available training data. After incorporating information, simulated water level process exhibits “double-peak” pattern, where first peak noticeably lower than under non-scheduling conditions. current introduces an innovative analyzing large-scale flooding, offering valuable perspectives planning mitigation strategies.

Language: Английский

Capturing Urban Pluvial River Flooding Features Based on the Fusion of Physically Based and Data-Driven Approaches DOI Open Access

Chenlei Ye,

Zongxue Xu, Weihong Liao

et al.

Sustainability, Journal Year: 2025, Volume and Issue: 17(6), P. 2524 - 2524

Published: March 13, 2025

Driven by climate change and rapid urbanization, pluvial flooding is increasingly endangering urban environments, prompting the widespread use of coupled hydrological–hydrodynamic models that enable more accurate flood simulations enhanced forecasting. The simulation method for river caused heavy rainfall has garnered growing attention. However, existing studies primarily concentrate on prediction using hydrodynamic or machine learning models, there remains a dearth comprehensive framework couples both to simulate temporal evolution changes. This research proposes novel simulating integrating physically based with deep approaches. sample set through data augmentation Generative Adversarial Networks, scheduling control signals are incorporated into encoder–decoder architecture results demonstrate strong model performance, provided model’s structural complexity aligned available training data. After incorporating information, simulated water level process exhibits “double-peak” pattern, where first peak noticeably lower than under non-scheduling conditions. current introduces an innovative analyzing large-scale flooding, offering valuable perspectives planning mitigation strategies.

Language: Английский

Citations

0