Rethinking Climate Risk & Insurance: Learning the Spatiotemporal Dynamics of Interconnected Extremes

Financial systems underpin climate resilience: they fund protective infrastructure, shape preparedness through risk transfer, and provide the insurance that enables communities to recover from disasters. Yet these systems are increasingly strained by climate-driven extremes, evident in the National Flood Insurance Program’s multi-billion-dollar debt and the withdrawal of insurers from high-risk states such as California and Florida. A central driver of this vulnerability is a modeling gap: most catastrophe and risk-pricing methods still treat regional hazards as independent, although floods, storms, droughts, and wildfires emerge as correlated, spatiotemporally organized events shaped by climate variability and change. This talk presents new stochastic and machine learning methods to characterize, forecast, and adapt insurance and infrastructure portfolios to interconnected hydroclimate extremes, using flood risk as a case study. By connecting federal policy, multiscale climate dynamics, and financial solvency, this work provides a foundation for rethinking natural hazard insurance design and adaptation strategies with more resilient, equitable, and economically stable climate-risk systems.

Speaker Adam Nayak, Columbia University

Attend in person or online (see weblink)