Session 1A.2 Real-Time High Resolution Numerical Weather Prediction Forecasts for Incident Meteorologists

Monday, 1 June 2009: 9:15 AM
Grand Ballroom East (DoubleTree Hotel & EMC - Downtown, Omaha)
Alexander O. Tardy, NOAA/NWS, San Diego, CA

Presentation PDF (1.2 MB)

During the summer of 2008 the National Weather Service Forecast Office in Salt Lake City, Utah, ran an experimental, nested, high-resolution numerical weather prediction model on a Linux cluster. The goal of the project was to provide real-time 3-km model forecasts for use by Incident Meteorologists at wildfires. A version of the WRF model, running the ARW core with explicit convection, was run using initial and boundary conditions from the NCEP's North American Mesoscale Model.

There were two 3-km nests defined across northern and southern Utah, respectively. Gridded and point output was made available on the Internet in the form of high resolution GEMPAK graphics and numerous point forecasts at selected locations, some collocated with automated observing (i.e., RAWS) sites. Forecast output was available in real-time for evaluating the potential for the development of moist convection within 18 hours of the time of model initialization. The high resolution model output, such as composite reflectivity, proved useful for the short-term temporal and spatial depiction of convection across the complex terrain of southern Utah during the summer monsoon.

Model forecasts of the movement and intensity of convection proved useful as an aid in the forecast and warning process (i.e., prior to and with radar reflectivity). Although the ability to issue warnings based solely on ensemble or deterministic numerical model forecasts is in its early stages, this study offered promise for applying short-term model forecasts in an operational setting.

- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner