Exploring the implications of methodological choice for exposure and vulnerability integration in impact-based warning provision

Under a changing climate, severe weather events are expected to become increasingly impactful, prompting a shift from hazard-based to impact-based forecasts and warnings (IBFW). While IBFW systems can improve risk perception and response, their effectiveness depends on integrating high-quality exposure and vulnerability data, which remains a global challenge. This study aims to explore how different IBFW methodologies influence warning provision. Three exploratory methodological comparisons have been conducted, comparing warning levels assigned to areas based on (1) hazard intensity using wind gust footprints, (2) exposure thresholds for buildings and population, and (3) building damage states in a risk-based approach. Two case studies from New Zealand (Cyclone Cook) and the UK (Storm Eunice) were used, with automated risk modelling pipelines using RiskScape software. Results show that hazard-only approaches yield the highest proportion of red warnings, while incorporating exposure and vulnerability data reduces the number and severity of warnings, with notable differences between countries and asset types. The urban-rural bias is demonstrated through comparing warning levels assigned to areas with similar hazard severity and varying population levels. The risk-based approach, though more nuanced, is limited by data availability and completeness, particularly for building characteristics and the availability of damage functions. These findings highlight the critical importance of matching risk assessment methodology and data selection to the intended purpose of IBFW systems, as methodological choices influence who receives warnings and at what level. Future work should address data gaps, expand and extend case studies and sensitivity experiments and explore dynamic exposure to improve IBFW system performance and societal outcomes.