Increases in Extreme Precipitation over the Northeast United States using High-resolution Climate Model Simulations

The Northeast United States (NEUS) has faced the most rapidly increasing occurrences of extreme precipitation within the US in the past few decades. Understanding the physics leading to long-term trends in regional extreme precipitation is essential to adaptation and mitigation planning. Simulating regional extreme precipitation, however, remains challenging, partially limited by climate models’ horizontal resolution. Our recent work shows that a model with 25 km horizontal resolution facilitates a much more realistic simulation of extreme precipitation than comparable models with 50 or 100 km resolution, including frequency, amplitude, and temporal variability, based on ensembles generated by GFDL (Geophysical Fluid Dynamics Laboratory) SPEAR (Seamless System for Prediction and EArth System Research) models. The 25-km GFDL-SPEAR ensemble also simulates the trend of NEUS extreme precipitation quantitatively consistent with observed trend over recent decades, as the observed trend is within the ensemble spread. We therefore leverage multiple ensembles and various simulations (with historical radiative forcing and projected forcing following the SSP2-4.5 and SSP5-8.5 scenarios) to detect and project the trend of extreme precipitation. The 10-ensemble member GFDL-SPEAR 25-km simulations project unprecedented rainfall events over the NEUS, driven by increasing anthropogenic radiative forcing and distinguishable from natural variability, by the mid-21^st century. Furthermore, very extreme events (99.9^th percentile events) may be six times more likely by 2100 than in the early 21^st century.

 

We further conduct a process-oriented study, assessing the physical factors that have contributed to the increasing extreme precipitation over the NEUS. We categorize September to November extreme precipitation days based on daily cumulative precipitation over the NEUS into weather types, including atmospheric river (AR), tropical cyclone (TC), and others. In observations, the most precipitation days were AR days or/and TC days. The number of extreme precipitation days related to pure AR events (without any TC-related event in the vicinity) had increased slightly from 1959 to 2020. The greater contribution to the increasing extreme precipitation was caused by TC-related events, especially the influences from extratropical transitions. The extreme precipitation days related to extratropical transitions were 2.5 times more frequent for the 1990 to 2020 period compared to the 1959 to 1989 period. We apply the same analysis to the GFDL-SPEAR 25-km simulations. Similar to observations, the increasing extreme precipitation days were mainly caused by TC-related events, with a smaller influence from pure AR events. However, the increasing number of TC-related days was dominated by hurricane and tropical storm events, while the number of extratropical transitions near the NEUS changed very little from 1959 to 2020. These results are different from the observational results. Ongoing work focuses on the discrepancy between observations and SPEAR simulations. For example, we are assessing whether the prominent increasing extratropical transitions since the 1990s in observations were the results of limited sample size or caused by decadal variability.