Model For Prediction Across Scales (MPAS) applied to regional climate of the Northwestern U.S.

Climate in the Northwestern United States depends sensitively on processes at multiple scales. These include global patterns of variability like the El Niño Southern Oscillation (ENSO), the planetary-synoptic scale that controls the strength and position of the jet, the mesoscale dynamics of fronts and influence of topography, and the scale of individual catchments where fine-scale topography controls snow accumulation. The last requires integration with high-resolution hydrologic models, but much of the rest of the global to regional-scale sources of uncertainty can be assessed in a global climate context. The Model for Prediction Across Scales (MPAS, Skamarock et al., 2012) is particularly suited to this challenge, because higher resolution at the region of interest is possible in a variable resolution modeling framework. This framework also enables seamless analysis of regional climate, global-scale climate variability, and the connection between the two. We present the initial results of global simulations with the finer, 30km resolution, part of the mesh centered over the Northwest. We use non-hydrostatic dynamics and physics parameterizations from the Community Atmosphere Model (CAM5) with prescribed observed sea surface temperature (SST) for the historical control simulations. A 21st century experiment uses SST changes based on Coupled Model Intercomparison Project (CMIP5) model mean projections for the RCP8.5 scenario, with associated greenhouse gas forcing. From an ensemble of simulations started with different initial conditions, we analyze climatic features important for the Northwest and evaluate their projected changes and uncertainty due to internal variability.