A Novel Data‐Driven Probabilistic Approach for Consequence Specific Urban Flood Risk Mapping

Urban floods are increasing in frequency, causing significant socio‐economic impacts, especially where risk mitigation is limited. Conventional hydrologic and hydraulic models identify flood‐prone areas but demand extensive data and computation, whereas risk mapping offers a data‐efficient alternative by integrating elements at risk such as population and infrastructure. This study develops a probabilistic framework for consequence‐specific urban flood risk mapping, defining risk as the product of hazard and vulnerability probabilities. Hazard maps are generated from three rainfall events using non‐exceedance probabilities of best‐fit distributions. Similarly, indicator‐specific vulnerabilities are estimated for each consequence and are combined via a linear aggregation. Indicator weights are derived using a novel Empirical‐Improved Analytical Hierarchy Process (E‐IAHP), replacing subjective judgments with a statistical measure (absolute correlation coefficient in this study), thereby ensuring objectivity and computational efficiency. The framework is flexible across temporal scales and can incorporate additional hazard attributes when required. The approach produces tailored risk maps for three urban flood consequences, that is, disruption of emergency service, traffic disruption, and economic losses of buildings, revealing distinct spatial patterns across central, south‐central and northern wards. Event specific crowdsourced flood locations provide indirect validation of the estimated risk patterns. Overall the framework demonstrates its adaptability for real‐time flood assessment and broader disaster management applications.