Characteristics of synoptic-scale cyclones crossing the West Antarctic Ice Sheet and the impacts of assimilating COSMIC GPS RO in Polar WRF forecasts of these events

Antarctic cyclones still pose the greatest safety threat not only for aircraft operations but also to vessels plying the Southern Ocean. Over land, tourists, research personnel and research activities on can also be significantly impacted by these systems. The cyclones also play a crucial role in the mass balance of the Ice Sheet on the continent. Potential variability in their dynamic characteristics arising from changes in climate therefore calls for greater understanding of their evolution. Such a diagnosis was all but impossible until recently because of the scarcity of data brought about by the extreme climatic conditions. Most Antarctic cyclones remain offshore beyond observations of most of the coastal stations on the continent. But when they make land fall these systems can be very disruptive of activities on land. In this study three cyclones that crossed into the West Antarctic Ice Sheet (WAIS) are identified and simulated using Polar WRF in two sets of experiments. Atmospheric profiles derived from COSMIC Radio Occultation data over the Southern Ocean are assimilated in one set but not the other. The nested configuration used has an outer domain covering much of the Southern Ocean at 60 km in order to maximize the number of usable occultations. But to accurately represent the impact of Transantarctic Mountain range a nested domain covering WAIS at 15 km is used. The simulations are driven by final analysis (FNL) and SSTs from NCEP; sea ice and SSTs are updated every six hours. The strategy adopted is a series of 48 hour integrations with only the last 24 hours used in diagnosis. The first part of this study describes the skill of Polar WRF in forecasting the three cases while the second is aimed at demonstrating the impact of COSMIC profile assimilation. The systems are characterized in the preliminary results by the path of minimum pressure, strength of the winds and spatial extent. Individual time slices and series are used because average statistics do not accurately represent the cyclone evolution. The forecast temporal evolution closely follows that depicted in the recent reanalysis from the European Centre for Medium-Range Weather Forecasts (ERA Interim). However subtle differences appear near the center of these systems. Measurements from the Moderate Resolution Imaging Spectroradiometer (MODIS) on board the Earth Observing System (EOS) Terra and Aqua satellites show large swirling bands of clouds coinciding with the forecast location and extent of the cyclones. With enhancement the images also show cloud bands centered over the Ronne Ice Shelf where the model shows the redevelopment of one of cases. Thus the forecast cyclone track and the spatial extent verify well against MODIS infrared cloud images. Over ice shelf locations on WAIS observed surface pressure decreases while the differences between forecasts and observations increase as the cyclone approaches. The results show differences in surface pressure that are less than 5 hPa for interior stations but are up to 10 hPa at ice shelf locations. Polar WRF forecasts the evolution of the storm realistically but the lowest pressure values are displaced. While the variability of the wind during the life of one of the cyclones is well represented, the forecast wind speeds are slightly greater than observed and the error in wind direction can be as large as 180o.