The next KE Hub online Triage Workshop will be presented by WieldMore:
AI-based correlation model for “Physical Risk Clusters” in UK and Europe
Industrial enterprises rarely know how individual and combined climate hazards are projected to evolve (and therefore, cluster) at their factory locations in the coming years.
WieldMore is working on building a mathematical climate-correlation model for our software platform, Pomelo, that ingests historical and projected climate data (like flooding, temperature stress, wildfires, wind stress, drought) and identifies multi-hazard hotspots automatically.
The model will incorporate an explainable, quantitative method for aggregating and scoring risk combinations, capturing the fact that some hazard correlations are far more destructive in combination than others. For example, a location with 50% wildfire projection + 50% wind speed increase is far more dangerous (due to fire spread) than 50% flood risk + 50% wind speed increase, which is a weaker interaction.
We are therefore seeking mathematical approaches to formally model and parameterise these cross-hazard interaction effects, in a way that remains interpretable and scalable across thousands of industrial sites and regions. Contributions toward defining a robust aggregation or correlation framework for multi-hazard clustering would directly support the next generation of financial-grade climate risk analytics at WieldMore.
