An Assessment of Different Initialization Fields on Short-term Sea Ice Forecast Parameters Using a Coupled Ice-Ocean-Atmosphere Model

We present weather-scale (0-10 day) sea ice forecast validation and skill results from an experimental coupled ice-ocean-atmosphere model during the fall freeze-up periods for 2015 and 2016.  The model is a mesoscale, coupled atmosphere-ice-ocean mixed-layer model, termed RASM-ESRL, that was developed from the larger-scale Regional Arctic System Model (RASM) architecture. The atmospheric component of RASM-ESRL consists of the Weather Research and Forecasting (WRF) model, the sea-ice component is the Los Alamos CICE model, and the ocean model is POP.  Experimental 5-day forecasts were run daily with RASM-ESRL from July through mid-November in 2015 and 2016. Our project focuses on how the modeled sea ice evolution compares to observed physical processes including atmospheric forcing of sea ice movement, melt, and freeze-up through energy fluxes. Model hindcast output is validated against buoy observations, satellite measurements, and concurrent in situ flux observations made from the R/V Sikuliaq in the fall of 2015.  Model skill in predicting atmospheric state variables, wind and boundary layer structures, synoptic features, cloud microphysical and ocean properties will be discussed.  We will show results of using different initializations of ocean sea surface temperature and sea ice extent and the impacts on sea ice edge prediction.