Estimating Past, Present, and Future Contributions of Climate Change to Economic Losses from Hurricanes in the United States

Hurricanes are one of the costliest natural disasters in the United States, imparting over $50 billion in economic losses per year. Because each individual event can cause significant immediate damage, as well as long-lasting negative impact, the resilience of coastal regions depends crucially on quantifying the present and future risk of hurricane-driven economic losses. The distribution of these losses is likely to change as climate change alters the frequency, intensity, spatial distribution, and wetness of these events. To estimate this change, we construct and apply a physical-econometric model that directly simulates high-resolution storm surge, wind, and rain from >10 million real and synthetic storm events. We merge this with a sale price dataset for all U.S. properties and state- and claim-level insurance data from historical storms to empirically derive a damage function for each of these three hazards. We then project damages from 1980-2100 using seven different climate models, two emissions scenarios, and 100 stochastic realizations of each hurricane season. We use the distribution of realized losses under constant property value conditions to quantify how hurricane risk has changed across the Atlantic and Gulf coastlines over the past century due to climate change. We compare the range of stochastic simulations to the observed historical record and estimate a likelihood ratio of experiencing the observed historical record in the presence and absence of climate change. The stochastic nature of tropical cyclone seasons renders the observed damages similarly likely in both scenarios. However, we then project damages into the future under RCP 4.5 and RCP 8.5. By mid-century, the climate signal is apparent and large under almost all stochastic simulations. Preliminary results suggest that the expected national average annual loss from hurricanes has grown by roughly 100% relative to a 1980s baseline and may further grow another 300% by the end of the century, under RCP 8.5 (see Figure). These results quantify the contribution of changing hurricane damages in the U.S. to the Social Cost of Carbon (SCC) and have informed state governments, asset management firms, and insurance regulators that seek to accurately price the risk of changing hurricane patterns to their assets and/or their population.