Improving Polar WRF Forecasts of Antarctic Cyclones using Data Assimilation Techniques

Understanding the dynamics of Antarctic cyclones responsible for transporting heat and moisture to the Antarctic continent is important to forecasting which supports operational activities of the United States Antarctic Program (USAP). The Polar Meteorology Group at the Ohio State University recently evaluated the Polar Weather Research and Forecasting (Polar WRF) model over the Antarctic to assess the impact of model improvements, resolution, large-scale circulation variability, and uncertainty in initial and lateral boundary conditions. They found that Polar WRF has a cold summer and a warm winter bias, stronger than observed surface wind speeds and a large sensitivity to the data used to specify the lateral boundary conditions. The forecast skill was the highest when the ERA-Interim reanalysis was used to drive Polar WRF as the temperature and wind speed biases were minimized.

This study improves on the Polar WRF forecasts with ERA-Interim using simulations with and without assimilation of GPS Radio Occultation profiles. The WRF 3DVAR and the Ensemble Kalman data assimilation techniques are used to improve the forecasts of four representative Antarctic cyclone cases. To make the forecasts, ensembles whose members differ only in the initial conditions are used over the first 24 hours of each 48 hour forecast. It is assumed that spinning-up the model with data assimilation produces the most accurate initial conditions from which the cyclone forecasts can be made.