Uncertainty in Future Summer Precipitation in the Laurentian Great Lakes Basin: Dynamical Downscaling and the Influence of Continental-Scale Processes on Regional Climate Change

Physics-based miniensembles of Weather Research and Forecasting (WRF) Model configurations have been employed to investigate future precipitation changes over the Great Lakes basin of eastern North America. All physics configurations have been employed to downscale multiple distinct Community Earth System Model, version 1 (CESM1), simulations driven by the representative concentration pathway 8.5 (RCP8.5) radiative forcing scenario, spanning a range from moderate (2045–60) to considerable (2085–2100) climate change. Independent of the physics configuration employed, all projected future precipitation changes are characterized by a general increase and a fattening of the tail of the daily rainfall distribution by the end of the century. The fattening of the tail can however be masked by natural variability in the case of the moderate warming expected by midcentury. The heavy-rainfall-derived precipitation increase is projected to be larger than or equal to the Clausius–Clapeyron thermodynamic reference of 7% increase per degree Celsius of surface warming, whereas the increase of average-rainfall-based precipitation becomes limited only for the largest global warming projections. This limitation is dramatically illustrated in one physics configuration at the end of the century. By downscaling the results obtained from the initial-condition ensemble, it is demonstrated that the extreme drying of the Great Lakes basin region characteristic of the most extreme end member of the CESM1 ensemble is significantly modified by downscaling with the version of WRF coupled to the Freshwater Lake model (FLake) of lake processes. This result does, however, depend upon the physics configuration employed in WRF for the parameterization of processes that cannot be explicitly resolved.