Multi-Source Data Fusion and Hydrodynamics for Urban Waterlogging Risk Identification.

Area of Science:

• Environmental Science
• Hydrology
• Urban Planning

Background:

• Urban waterlogging disasters are complex, influenced by numerous factors, necessitating effective risk identification.
• Existing methods often lack comprehensive data integration and real-time simulation capabilities for urban waterlogging.

Purpose of the Study:

• To propose and validate a novel framework for identifying urban waterlogging risk by combining multi-source data fusion with hydrodynamics (MDF-H).
• To develop a method for real-time simulation and prediction of waterlogging extent and depth in large-scale urban areas.

Main Methods:

• Developed a multi-source data fusion (MDF-H) framework integrating meteorological, geographic, and municipal engineering data.
• Incorporated hydrological analysis to divide sub-catchments for localized runoff calculations, considering 12 input factors for a real-time runoff coefficient.
• Employed hydrodynamic theory for inter-regional surface runoff connectivity and a two-level drainage capacity assessment model.

Main Results:

• The MDF-H framework successfully simulated the transition of rainfall into surface runoff, generating real-time waterlogging distribution maps.
• Validation against a 2019 Shenzhen rainstorm event showed an average accuracy exceeding 95% for waterlogging depth identification.
• The framework demonstrated precise prediction, rapid calculation, and wide applicability for large-scale urban regions.

Conclusions:

• The proposed MDF-H framework offers a robust and accurate approach to urban waterlogging risk assessment and real-time monitoring.
• This method provides valuable insights for urban planning and disaster management, enhancing resilience to waterlogging events.