Modeling System Description
The model forecast is produced by the NCAR/Penn. State Fifth Generation Mesoscale Model (MM5) version 3.6 whose physical parameterizations have been extensively modified based on boundary layer and free atmosphere validation studies in Greenland and Antarctica. In particular, the planetary boundary layer, surface energy balance, prognostic cloud, and cloud-radiation interaction descriptions have been adapted. There is an explicit treatment for sea ice of variable concentration. The grid spacing is 60 km, with the lowest of the 28 vertical levels being about 20 m above the surface and the highest being around the 100-hPa level. The Antarctic terrain data are provided by a modern digital elevation model constructed by the Radarsat Antarctic Mapping Project (RAMP). The model time step is 180 s. The initial and boundary conditions are obtained from the daily 0000UTC and 1200 UTC run of the Aviation (AVN) model, available from the ftp site maintained by the National Centers for Environmental Prediction. No assimilation of in situ observations is performed. The AVN data set also provides the ocean surface conditions north of Antarctica.
The 0000UTC (1200 UTC) AVN run is available from NCEP around 0100 (1300) Eastern Time each day. The 5-day forecast is completed before 1000 (2200) the same day and immediately posted to the web page. These twice a day forecasts are both posted on the web page with two radio buttons shifting between 0000z and 1200z forecasts so that if one of the model runs crashes, a 12 hour earlier forecast is still available for various communities.
Use of Model Output
The Polar Meteorology Group of the Byrd Polar Research Center at The Ohio State University provides the Antarctic numerical weather prediction as a public service on an experimental basis. While all reasonable attempts are made to provide the best possible forecasts, there is no guarantee made as to their accuracy – let the user beware. Studies will be undertaken to establish objective performance criteria. We believe that the performance of state-of-the-art mesoscale models has advanced to the stage where very useful forecasts can be provided, but these will always be plagued by this area's limited observational database that restrict the fidelity with which the synoptic scale circulation can be resolved.
Because the grid spacing is 60 km, the surface predictions do not provide an accurate result for areas of complex terrain, like around McMurdo Station on Ross Island and near the top of the West Antarctic ice sheet adjacent to Byrd Station.
In relation to specific plots:
Numerical Weather Prediction Group