High-Resolution Global Simulations Using the Model for Prediction Across Scales (MPAS) for Use in Climate Change Studies

We present high-resolution multi-seasonal simulations conducted in present and future thermodynamic environments using the Model for Prediction Across Scales-Atmoshere (MPAS-A) for the use in studying climate change effects on various Northern Hemispheric phenomena. MPAS-A is a global, atmosphere only model that utilizes variable-resolution grids, which allows for localized areas of high-resolution to gradually expand to lower resolutions. Here, a variable-resolution mesh with 15-km grid spacing over the entire Northern Hemisphere expanding out to 60-km elsewhere is used.

Ten simulation years were chosen with varying strengths of El Niño-Southern Oscillation (ENSO) and are each integrated for 14.5 months. Given the atmosphere-only nature of MPAS, these years are thus defined by analyzed SST patterns. Model output is post-processed to vertically interpolate fields to select pressure levels and horizontally interpolate from the native, unstructured mesh to a 0.15ºx0.15º latitude-longitude grid. The ECMWF Interim Reanalysis (ERA-I) is used for the present-day initial conditions, with sea surface temperature (SST) and sea ice fields updated daily. For the future climate simulations, the ERA-I initial conditions are altered by applying monthly-averaged temperature changes derived from a 21-member ensemble of Coupled Model Intercomparison Project phase 5 (CMIP5) general circulation models (GCMs) to all atmospheric pressure and soil levels. Carbon dioxide (CO₂) concentrations are also adjusted in the future climate simulations to the level projected by the RCP8.5 emissions scenario for 2100.

The SSTs for the future climate simulations are warmed in the same manner as the atmospheric and soil temperatures in the initial conditions. Given the role SST gradients play in midlatitude cyclone development and other regional climate changes, it is important to handle the lower boundary conditions in such a way that preserves their realistic magnitudes. Additionally, pseudo-daily sea ice fields were created from the monthly-averaged CMIP5 ensemble mean sea ice for both historical (1980–1999) and future (2080–2099 under the RCP8.5 emissions scenario) time periods. The analyzed sea ice in the ERA-I data is then replaced with these climatological fields for use in all model simulations. Handling the sea ice in this manner ensures that it is properly represented in the future climate simulations.

These simulations, which have been shown to reasonably represent present-day large-scale mean fields such as summertime tropical precipitation patterns and wintertime midlatitude jet features as well as tropical cyclone activity and seasonality, are currently being used to study climate change effects on tropical cyclone genesis location and seasonality, extratropical transition events, and persistent anomalies. However, there many more aspects to be explored; therefore, it is our intention to make these model output publicly available in hopes that it is useful to the broader research community for studying various other meteorological phenomena as well as participating in model comparison studies.