Overview
Urban flood disaster mitigation is challenged by the complexities inherent in city systems and the high computational demands of existing simulation models. To address these issues, a novel Urban Flood Dynamical System Model (UFDSM) has been developed. This model utilizes a Cellular Automata Urban Flood Model concept, allowing for flexible cell type customization and selection of water motion or distribution rules pertinent to specific urban environments. The objective is to integrate extensive urban system data while maintaining simplified rules to reduce computational complexity. Concurrently, a first-order conservation nonstandard finite difference algorithm was introduced, designed to preserve the UFDSM's evolutionary properties, ensure positive and conserved water solutions, and maintain fixed-point characteristics consistent with the dynamical system.
Research Context
Urban floods represent significant natural disasters. Although numerous flood simulation models exist and have matured, two primary challenges persist: the oversimplification of urban systems within these models and their substantial computational resource requirements. The present research aims to circumvent these limitations by proposing a new modeling approach that balances urban system detail with computational efficiency. The development of the UFDSM is situated within the broader context of improving the accuracy and practicality of urban flood prediction and management tools.
Approach
The development of the Urban Flood Dynamical System Model (UFDSM) commenced with the adoption of the Cellular Automata Urban Flood Model concept. This framework permits the customization of cell types and the selection of water motion or distribution rules based on explicit urban environmental characteristics. The intent behind this design is to facilitate the incorporation of urban system data while allowing for rules that simplify computational demands. A key aspect of the model's design includes a sufficient condition for water motion and distribution rules, ensuring that the dynamical system's solutions align with macroscopic physical conditions governing water movement.
To preserve the evolutionary traits of the UFDSM, a first-order conservation nonstandard finite difference algorithm was proposed. This numerical method is engineered to guarantee positive solutions and conservation of water within the simulation. It also aims to maintain the same fixed-point characteristics as the underlying dynamical system. The validation of this numerical method involved a comparison of its outputs with an analytical solution.
For model applicability verification, an urban flood simulation experiment was conducted. The results from the UFDSM were compared against those from HEC-RAS, a widely recognized hydrological engineering software. This comparison focused on specific distance discrepancies and relative distances to assess the model's performance against an established benchmark.
Findings
- The Urban Flood Dynamical System Model (UFDSM) is based on the Cellular Automata Urban Flood Model concept, permitting flexible customization of cell types and selection of water motion/distribution rules according to urban environments.
- A sufficient condition was provided for the water motion and distribution rules to ensure that solutions of the dynamical system align with macroscopic physical conditions governing water movement.
- A first-order conservation nonstandard finite difference algorithm was proposed, designed to preserve the evolutionary properties of the UFDSM.
- This numerical method ensures positive solutions and conservation of water, while also maintaining the same fixed-point characteristics as the dynamical system.
- Validation of the numerical method was conducted by comparing its output with an analytical solution.
- In an urban flood simulation experiment, a comparison with HEC-RAS showed an approximate 2mm discrepancy in distance dp' and a 0.02mm discrepancy in distance d2'.
- The relative distance Rp was approximately 7.5%, and the relative distance R2 was approximately 0.06% when compared to HEC-RAS.
Potential Applications
The proposed model is constructed to be easily coupled with other hydrological processes and facilitates data assimilation. This capability offers promising practical applications within urban flood management and prediction frameworks, potentially allowing for more integrated and responsive flood forecasting systems.