01Public catastrophe modeling
The Florida Public Hurricane Loss Model gives a public account of hurricane wind loss, open to scientific scrutiny where commercial catastrophe models stay proprietary. We treat that loss field as the starting point for financial design.
02Physics-informed machine learning
Environmental risk data are often limited for the extreme events that matter most. Physics-informed machine learning combines observations with physical models and system constraints to produce more reliable and interpretable estimates. The lab couples these methods with stochastic ensembles to propagate uncertainty from hazards through exposure and vulnerability to infrastructure disruption and economic loss, supporting decisions based on the likelihood and range of possible outcomes rather than a single estimate.
03Financial risk-transfer design
A hazard estimate is useful when it becomes a contract. The lab connects loss and revenue shortfalls to catastrophe bonds, parametric insurance, and water futures for utilities, irrigation districts, insurers, and public agencies.