Synchronization-Enhanced Deep Learning Early Flood Risk Predictions: The Core of Data-Driven City Digital Twins for Climate Resilience Planning

Abstract

Floods have been among the costliest hydrometeorological hazards across the globe for decades, and are expected to become even more frequent and cause larger devastating impacts in cities due to climate change. Digital twin technologies can provide decisionmakers with effective tools to rapidly evaluate city resilience under projected floods. However, the development of city digital twins for flood predictions is challenging due to the time-consuming, uncertain processes of developing, calibrating, and coupling physics-based hydrologic and hydraulic models. In this study, a flood prediction methodology (FPM) that integrates synchronization analysis and deep-learning is developed to directly simulate the complex relationships between rainfall and flood characteristics, bypassing the computationally expensive hydrologic-hydraulic models, with the City of Calgary being used for demonstration. The developed FPM presents the core of data-driven digital twins that, with real-time sensor data, can rapidly provide early warnings before flood realization, as well as information about vulnerable areas—enabling city resilience planning considering different climate change scenarios.