Data-Driven Framework for Enhanced Flash Flood Preparedness and Building Urban Resilience

Building urban resilience is important for disaster preparedness, e.g., for flash floods. Flash floods are especially disruptive as it is a challenge to implement disaster response in densely populated urban regions effectively. With the explosion of data availability, there is existing work on using data-driven methods to predict the event and determine risk perception by the stakeholders. Unlike the former, the latter includes human factors. Given the difference in the data involved and its subsequent analysis, their usage is siloed and ineffective. This work reviews the state-of-the-art machine learning-based models used for flash flood susceptibility mapping (FFSM) to summarize the trends in forecasting. These susceptibility maps, vulnerability maps, and socio-economic impacts are used to determine the flood risk assessment in a given region. At the same time, state-of-the-art disaster preparedness is implemented by and for local stakeholders using their data related to flood management. Revisiting the protocols for building urban resilience is needed to reduce the siloed usage of the two analyses. Since several studies point to the effectiveness of data-driven mechanisms, the novel contribution is a holistic data-driven framework that enhances flash flood preparedness in urban regions. The framework integrates FFSM and the stakeholders’ awareness to improve the overall disaster preparedness. The vision is that the infusion of more heterogeneous data in flash flood preparedness improves urban resilience.