New supercomputer in Wyoming aids Antarctic safety

More-detailed weather forecasts to guide takeoffs, landings

WRF forecast of surface temperatures and winds over Antarctica from Erebu

A test of the AMPS forecasting system on the new Erebus supercomputer produced this detailed 120-hour forecast of surface temperatures and winds over Antarctica.  Temperatures in Celsius are indicated using the color scale at right.  Surface wind direction and speed are shown with barbs; a full-length barb equals 10 knots. (Forecast valid 1200 UTC 11 Dec 2012, courtesy Jordan Powers, NCAR.)

December 5, 2012 | When the weather is good enough for U.S. Air Force pilots to land their C-17 cargo jets on one of the ice runways at Antarctica’s McMurdo Station, they know they’ll probably see volcanic Mt. Erebus rising 13,000 feet into crystalline blue sky.

Part of the reason they’ll get a clear view, and in turn a safe landing for passengers and cargo, will soon have to do with another Erebus. That’s the name of the new NCAR supercomputer dedicated to the real-time Antarctic Mesoscale Prediction System (AMPS) weather forecasts.

Based at the recently opened NCAR-Wyoming Supercomputing Center, Erebus is now in testing phase. When fully operational, the system is expected to provide about 15 times greater computing power for AMPS than the current system, a portion of NCAR’s soon-to-be-retired bluefire supercomputer. Erebus is funded by the National Science Foundation’s Office of Polar Programs.

Since AMPS was created in 2000, its twice-daily numerical weather predictions have greatly increased the safety and efficiency of U.S. and international operations in Antarctica. AMPS forecasts have helped cut in half the number of flights aborted because of adverse weather conditions experienced at ground locations, a major time and cost savings.

The U.S. Air Force and New York Air National Guard make about 100 flights every year between Christchurch, New Zealand, and McMurdo Station, the largest research base in Antarctica and the hub of the United States Antarctic Program. The pilots rely on Antarctic forecasts provided by meteorologists at the Space and Naval Warfare Systems Center, who in turn rely heavily on guidance from AMPS.

Each AMPS forecast uses ground-based and satellite observations of such atmospheric conditions as wind speed, humidity, and temperature. The resulting model output produces five-day forecasts over the continent as a whole, in addition to 36-hour forecasts at higher resolution for certain key regions of the continent.

When operational on Erebus as early as the end of this year, the forecasts will be run in less than 30 minutes, or about 10 times faster than the 5 to 5.5 hours on bluefire. In addition, the resolution of the forecasts will increase by one third, providing more detail.

“We’re very excited about the expanded computing capacity with Erebus and generating more detailed forecasts in less time,” says NCAR scientist Jordan Powers, who helped launch AMPS in 2000.

AMPS forecasts had previously ranged in resolution, or level of detail, from a 15-kilometer grid (9.3 miles) over all of Antarctica, to just under 2-km grids around McMurdo Station and other key areas. Erebus will provide the computing power to pinpoint local forecasts, such as critical ground wind conditions, down to 1.1 km, which is less than the length of a McMurdo runway.

Erebus will also enable the AMPS team to better test and integrate new polar physics approaches for describing the Antarctic atmosphere, which will in turn increase the predictive accuracy of the models.

*Media & nonprofit use of images: Except where otherwise indicated, media and nonprofit use permitted with credit as indicated above and compliance with UCAR's terms of use. Find more images in the NCAR|UCAR Multimedia & Image Gallery.

The University Corporation for Atmospheric Research manages the National Center for Atmospheric Research under sponsorship by the National Science Foundation. Any opinions, findings and conclusions, or recommendations expressed in this publication are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.