Although official weather records have been recorded for at least the past 150 years, weather prediction for thunderstorm activity is still not that reliable.
The University of Oklahoma’s Center for Analysis and Prediction of Storms (CAPS) has team up with the National Oceanic & Atmospheric Administration (NOAA) using an IBM Cray supercomputer at the Pittsburgh Supercomputing Center in Pittsburgh, Pennsylvania to improve weather forecasting of storms by using supercomputer analyses of the individual cells that are part of tornadoes and severe thunderstorms.
Numerical weather predictions are not precise because of the reliance on storm geographical areas of 10 Km or larger, which provides only a coarse resolution of storm activity. Focusing on individual cells and using a supercomputer for computational analysis of these thousands of cells gives a more precise analysis of the storm’s intensity.
CAPS and NOAA are using the Cray XT supercomputer to analyze 2-Km geographical areas in various areas of the U.S. by running ten models for “ensemble forecasting” to help eliminate errors and ensure more accurate data collection. The data is being collected and archived at the National Weather Center in Norman, Oklahoma.
Read more here.