Anthropogenic Impacts on the Exceptional Precipitation of 2018 in the Mid-Atlantic United States

During the summer of 2018, the Mid-Atlantic States of Pennsylvania, New Jersey, Maryland, Washington D.C., Delaware, and West Virginia experienced remarkable total and extreme (99^th percentile wet days) precipitation that contributed to devastating flooding. We analyzed observations from the Global Historical Climatology Network-Daily (GHCN-D) dataset and climate simulations from version 1 of the Community Earth System Model under the Large Ensemble Project (LENS) to assess the contribution of anthropogenic climate change to the exceptional precipitation of 2018. Specifically, we compared total and extreme precipitation data from the LENS historical (natural and anthropogenic forcings) 40-member ensemble to a counterfactual scenario in which we removed the time-evolving ensemble mean and added the climatological ensemble mean from 1920–1950, creating a set of 40 total and extreme precipitation times series for 1920–2018 without late 20^th century (1951–2018) forcing.

We find that the summer 2018 floods in the Mid-Atlantic are associated with the highest total precipitation from January to September and the fourth highest extreme precipitation from May to September 1920–2018. The late 20^th century anthropogenic forcing increases the likelihood of January to September total precipitation similar to 2018 by 52% (Risk Ratio = 1.52, 95% confidence interval: 1.09, 2.25), and increases in the likelihood of May to September extreme precipitation similar to 2018 by 12% (Risk Ratio = 1.12, 95% confidence interval: 1.05, 1.21). We find no statistically significant influence of late 20^th century anthropogenic forcing on occurrences of combined January to September total precipitation and May to September extreme precipitation similar to 2018, and caveat that the extreme precipitation risk ratio should be interpreted with caution given the limitations of the LENS historical simulations in accurately reproducing extreme precipitation when compared to GHCN-D observations.