M. Seefeldt, University of Colorado - Boulder

The Regional Arctic System Model (RASM) is a fully couped regional climate model (RCM) with the Weather Research and Forecasting (WRF) model as the atmospheric component. In this study RASM is used to dynamically downscale a 10-member ensemble from the CESM Decadal Prediction Large Ensemble (CESM-DPLE). The CESM-DPLE dataset provides three-dimensional atmospheric fields for the initial conditions, updates along the lateral boundaries, and nudging for the top half of the WRF model. The RASM ocean, sea ice and land initial conditions are obtained from a RASM hindcast forced with the ERA-Interim reanalysis with the date matching that of the initial date of the RASM-DPLE simulations. The results indicate that the top half of the atmosphere is constrained to and follows the large-scale atmospheric circulation of the CESM-DPLE dataset, as is expected and desired. Meanwhile the lower part of the atmosphere, and most critically the surface state and fluxes, can freely evolve. The result is that the surface state and fluxes have a mean climate state dependent on the WRF physics and surface coupling while reflecting the imprint of atmospheric variability of the driving CESM-DPLE ensemble members. Even though the lower atmospheric mean climatic state is similar across the 10 RASM-DPLE ensemble members, the sea ice states for each ensemble member in the fully coupled RASM exhibit unique states. This is due to the variability in the synoptic-scale weather from the driving data of the individual CESM-DPLE ensemble members. Bias correction, a common requirement for dynamically downscaling Earth system model data was determined to be unnecessary because of the initialization and updating of the surface state from the land, ocean, and sea ice component models.