Investigating Changes in Historical and Future Rainfall Patterns Using Climate Model Simulations

Comparing climate model simulations to ground-based observations is important for better understanding and quantifying uncertainties in climate models when simulating precipitation and its changes in time. Evaluation of individual climate models and their ensemble members can help gain confidence in future projections of climate model simulations. This work compares historical precipitation data from 70 ensemble members of the Community Earth System Model Version 2 large ensemble (CESM2-LE) of the Coupled Model Intercomparison Project Phase 6 (CMIP6) against a ground-based observations from the National Oceanic and Atmospheric Administration Climate Prediction Center Unified CONUS (CPC) gauge-based dataset. This is performed from 1948 to 2022 through temporal and spatial analyses across the Contiguous United States (CONUS). Time series of seasonal (spring, summer, fall and winter) average precipitation show that CESM2 underestimates the ground observations in summer and overestimates them in winter. Spatial patterns of mean annual precipitation across CONUS averaged during the study period indicate that different random ensemble members and the mean-multi ensemble have similar patterns. According to the relative difference maps between climate models and observations, the CESM2 model has a dry bias over the western and mid-south CONUS, and a wet bias in some parts of the Northern Great Plains and Southwest region. As a second step, this work investigated precipitation patterns changes in the past (1932-2022) and in the future (2023-2100) across CONUS using CESM2-LE for a medium to high emission scenario range SSP3-7.0 protocols from a 70-member ensemble. In general, results indicate dry summers and wet winters in the next century with respect to the past. Slightly upward trends are observed in annual, spring, fall, and winter averages in the recent decade (2010-2022), which will continue in the future projections. Eastern regions have higher annual and seasonal averages, as well as 75^th and 95^th percentiles of the probability distribution, whereas western regions are comparatively dry. In summer, the driest regions of CONUS are expected to get wetter than in the past century (with an increase of ~58% for a random ensemble member and ~40% for the multi-ensemble mean), as are some areas in the mid-north CONUS (with relative increases ranging from ~15% to ~45% for both a random ensemble member and the multi-ensemble mean). Northern NGP will become drier in spring, summer and winter seasons and fall precipitation will change very little in the overall CONUS in the future.