The Pittsburgh Supercomputing Center (PSC) has embarked on a major new storm forecast experiment that could speed forecast time and save lives during violent storms that escalate into potentially devastating tornados.
This spring the National Oceanic and Atmospheric Administration (NOAA) launched an experiment using PSC's Cray XT3 system, its newest computing system and the most powerful "tightly coupled" system available for U.S. public research. Through this venture, PSC has enabled meteorologists in Oklahoma to model weather at a finer scale than present systems allow, 2 to 4 kilometers of horizontal grid spacing, at which storms occur.
From mid-April until June 1st, PSC is shipping huge amounts of forecast data daily from Pittsburgh to Oklahoma through a dedicated high-bandwidth
network link contributed by Cisco Systems Inc. and coordinated by PSC
networking staff.
"This is the first time ensemble forecasting has been done at the
scale at which storms occur," explains Michael Schneider, PSC science writer. “Ensemble" forecasting runs a forecast model many times to measure the uncertainty inherent in weather forecasting, making it possible to improve the reliability of severe storm forecasting.
"Tornados are notoriously hard to predict, and better warnings--
hours in advance, instead of minutes, with greater reliability in the
prediction, -- would save countless lives," Schneider adds.
Since the mid-90s, PSC has collaborated with weather researchers in
Oklahoma in several spring experiments.During the 2005 season, meteorologists learned that with sufficiently high-resolution it's possible, in some cases, to predict the details of thunderstorms 24-hours in advance, a milestone in storm forecasting, suggesting that weather at this scale is
inherently more predictable than anyone had thought.
Writer: Debra Diamond Smit
Source: Michael Schneider, Pittsburgh Supercomputing Center