Extended Abstract (56K)JP1.3
Regional scale modeling for the 2002 Olympic Winter Games
Daryl J. Onton, CIRP, Univ. of Utah, Salt Lake City, UT; and A. J. Siffert, L. Cheng, W. J. Steenburgh, and B. Haymore
In February 2002, the Olympic Winter Games will be held in the Salt Lake City Metropolitan area and the nearby Wasatch Mountains. With over 100,000 spectators and athletes attending and competing daily at various venues, accurate weather forecasts are critical for public safety and games logistics. The NOAA-CIRP real-time MM5 mesoscale modeling system, run at 12-km resolution, will be a key component of the forecasting system implemented for the games.
The NOAA-CIRP real-time MM5 mesoscale modeling system is run twice daily on 8 700 MHz AMD nodes of the University of Utah Center for High Performance Computing Beowulf-class cluster. Initial and lateral boundary conditions for the modeling system are provided by the NCEP Eta model. An objective analysis package that uses the extensive surface observational network of the MesoWest Cooperative Mesonets will be implemented to improve model initial conditions. After an 82 min integration time required to complete a 36 h forecast, native-grid, hourly output from the model is ingested by the National Weather Service's AWIPS system and used by meteorologists at the Salt Lake City and Elko forecast offices. Model forecasts are also available to the public via the Internet and will be accessed by the 2002 Winter Olympics Venue Forecast Team using a software package called FX-NET.
Despite the relatively small grid spacing of the modeling system (12-km), many aspects of the complex topography of the intermountain west remain unresolved. As a result, raw model time series feature large temperature and wind biases that reduce their utility for operational forecasting. MOS-type equations have been developed to provide more accurate point-specific forecasts of wind, temperature, dewpoint, and relative humidity at the Olympic Venues and other locations.
Future improvements to the system include an increase in compute power that will allow for the development of a 5-member ensemble that will allow forecasters to better predict the likelihood of major weather events such as lake-effect and orographic snowstorms.
Joint Poster Session 1, Poster Session - Mesoscale Models—with Coffee Break
Monday, 30 July 2001, 2:30 PM-4:00 PM
Previous paper Next paper