Quantitative projections of changes in heat and cold deaths due to climate change in 209 U.S. Cities

The mortality risks in the United States from extreme temperatures, both heat and cold, have been demonstrated in real life events and quantified in various research efforts. We developed city and month specific mortality relationships from 30 years of historical data for both heat and cold in 209 cities in the contiguous U.S., using a Bayesian approach to increase statistical robustness by clustering cities with similar climates. Due to an observed decrease in the sensitivity of U.S. populations to heat over time we base our final results mainly on the last decade of data. These mortality relationships were applied to two different climate model simulations to develop 21^st century projections by month and city. These quantitative projections can inform both mitigation and adaptation efforts. Implications of the geographic and seasonal patterns of the results will be discussed. Challenges remain in terms of accounting for future adaptation, both planned and autonomous, as well as changes in population due to growth, migration, and changes in proportions of vulnerable populations. These results will be discussed in the context of other studies that also project future mortality in the U.S. The modeling results of the underlying study will be highlighted in the upcoming USGCRP report: Impacts of Climate Change on Human Health in the United States: A Scientific Assessment.