Dynamic Population Exposure in Multi-Hazard Early Warning Systems: An Activity-Based Approach Conceptual Method

Effective Multi-Hazard Early Warning Systems (MHEWS) rely not only on the accuracy of the weather forecasting models but importantly on their alignment with human dynamics and the natural and built environment. Risk assessment frameworks available at this stage often treat populations as static, fixed in residential locations regardless of the time of day or hazard onset and progression (Chen et al., 2023). This residence-based approach masks the true exposure of population in transit – individuals commuting, attending school or travelling. They may be physically exposed in low-vulnerability zones while possessing high social vulnerability (e.g., lack of local knowledge or support networks), or vice versa. Recent developments in mobility-based exposure assessment demonstrate that human activity patterns significantly alter vulnerability distributions across space and time, with exposure estimates varying by up to 40% between static and dynamic scenarios during peak activity (Rajput et al., 2024).

Our exercise introduces a methodological framework to operationalise social behavioural geography within a quantitative risk assessment model. Using the CLIMADA (CLIMate ADAptation) impact modeling platform (Aznar-Siguan & Bresch, 2019), we compare two distinct exposure scenarios for a compound hazard event in Ireland: (1) a static 'Residential' baseline assuming population distribution based on census residential locations, and (2) a more dynamic 'Activity-Based' scenario that integrates 2022 Irish Census data on Working From Home (WFH) patterns and commuting flows to redistribute social vulnerability based on diurnal activity patterns. This activity-based approach accounts for temporal mobility, capturing where people could be located during different times of day rather than solely where they reside.

By shifting the analytical focus to socially vulnerable populations in motion, this approach reveals "hidden hotspots" of risk, namely areas where physical hazard severity may be moderate, but the temporal convergence creates compounding crisis conditions (e.g. traffic jams or social event scenarios). Our methodological framework demonstrates that incorporating dynamic population distributions alters exposure assessments, with implications for emergency response resource allocation and warning dissemination strategies. We advocate for a paradigm shift in warning issuance protocols: when possible, transitioning from purely geographic alerts to behaviour-responsive, time-sensitive warnings that account for where people are during hazard events, not merely where they live (Haraguchi et al., 2022). This human-centric characterisation of exposure provides actionable insight for emergency managers and enhances the effectiveness of MHEWS for mobile individuals.

References:

Aznar-Siguan, G., & Bresch, D. N. (2019). CLIMADA v1: A global weather and climate risk assessment platform. Geoscientific Model Development, 12(7), 3085-3097. https://doi.org/10.5194/gmd-12-3085-2019

Chen, X., Hu, Y., Chi, G., & Chen, J. (2023). Assessing dynamics of human vulnerability at community level – Using mobility data. International Journal of Disaster Risk Reduction, 95, 103964. https://doi.org/10.1016/j.ijdrr.2023.103964

Haraguchi, M., Nishino, A., Kodaka, A., & Lall, U. (2022). Human mobility data and analysis for urban resilience: A systematic review. Environment and Planning B: Urban Analytics and City Science, 50(1), 7-27. https://doi.org/10.1177/23998083221075634

Rajput, A. A., Liu, C., Liu, Z., Zhao, J., & Mostafavi, A. (2024). Human-centric characterization of life activity flood exposure shifts focus from places to people. Nature Cities, 1, 290-301. https://doi.org/10.1038/s44284-024-00043-7