Multi-scale Historical Simulations of Regional Precipitation Extremes

This project builds on work to use the Model for Prediction Across Scales (MPAS) as a part of the atmospheric modeling framework for climate model studies, with physics from the Community Atmosphere Model - version 5 (CAM5). The focus of this work is on the Northwestern United States. Model integrations use a refined-mesh region that spans parts of the atmosphere over the Northeast Pacific and much of the Western United States (at ~30km resolution and the rest of the globe at ~120km). Historical sea surface temperatures are prescribed.

Multi-scale global modeling allows for greater spatial detail to resolve important topographic effects while also capturing the influence of global-scale modes of variability (like the El Niño Southern Oscillation - ENSO) and variability and change in circulation features like the subtropical jet. Here we characterize the climate simulation over historical integrations. To isolate the effects of resolution, we compare with a globally quasi-uniform coarse resolution integration. The model will be evaluated both on its performance in capturing mean state climate features and regional detail compared to observations and other models. The project probes the influence of variability on uncertainty in regional simulations through analysis of three ensemble members,  each forced by the same 25-years of observed sea surface temperatures. Thus, ENSO is prescribed, and for each event we have three instances to study the relative importance of unforced interannual atmospheric variability. Inter-annually, atmospheric rivers contribute to the variation in extreme precipitation. We have sufficient samples of extreme precipitation events to characterize the representation of interannual variability in the model, and whether the possibilities can be shown to span the observations.