The year 2024 was the warmest on record, exceeding 1.5°C above pre-industrial levels. This warming may lead to more intense hurricanes, as seen in the recent Hurricane Beryl, which was fueled by record-high sea surface temperatures. Traditional natural catastrophe (nat cat) models are based on historical data, but climate change is altering the world, making these models outdated. To address this, reask, a nat cat modeller and data provider, has developed a methodology that combines machine learning with advanced stochastic simulations to provide a climate-informed, forward-looking risk assessment. This approach can help quantify uncertainty in hurricane formation and risk. The model can simulate millions of hurricanes and provides a robust database for training machine learning algorithms. By forcing the model with different climates, the industry can assess how climate change affects the entire risk curve, including the tail. This methodology can be adapted for other climate-driven perils, such as wildfires and floods, to facilitate better risk management, pricing, and assessment. By anticipating non-linear shifts in risk distributions, the industry can avoid future surprises and be better prepared for a rapidly changing climate.
In the path of the gust: Adapting to the evolving threat of hurricanes
