Research Briefs

Scrutinizing the Windsor twister from all angles

Radar image of Windsor tornado.

A Doppler radar image from the National Weather Service shows the tornado-producing supercell thunderstorm near Windsor, Colorado, at 11:35 a.m. on May 22. (Image courtesy NOAA/NWS.)

After a very unusual tornado caused extensive damage along a 34-mile (55-kilometer) swath of northern Colorado in 2008, a team of scientists from NCAR and Colorado State University undertook a multidisciplinary study integrating meteorology, climatology, and social science.

The study, published in Weather and Forecasting, dissects the May 22 tornado near the town of Windsor from a meteorological perspective, places it in a climatological context, and analyzes how severe weather information was communicated to and interpreted by decision makers and then passed on to the public.  

The Windsor tornado, which caused one death and ranks as the costliest tornado in Colorado’s history at $193.5 million in damages, was rated EF3 on the Enhanced Fujita scale. Several characteristics of the tornado were unique for the region. The storm formed in the late morning, in contrast to the usual pattern of afternoon storms. It moved toward the north/northwest (toward population centers) rather than along the more common eastward storm track. It was strong and long-lived for a tornado close to the Front Range, where weaker tornadoes are more common.

The unusual characteristics and rapid development of the tornado created a complex situation for decision makers, whose interpretations of warning information varied widely. For example, since most decision makers knew that tornadoes in the area typically move eastward, some downplayed the possibility that the tornado would move toward them, despite the fact that they were in the tornado’s path. Many decision makers were surprised that such a strong tornado could occur in their area. Although none had been reported in the preceding 50 years, the study pointed out that the area near Windsor had been hit by strong tornadoes several times in the early 20th century.   

The research underscores the growing recognition that societal factors are just as important for the effectiveness of weather warnings as the timeliness and content of those warnings. “To achieve the goal of better tornado warnings, we need to not only improve detection and prediction of tornadoes, but also understand how people receive, understand, and use warnings, so that the information that meteorologists have can be best communicated to decision makers and the public,” says author Russ Schumacher, a professor at Texas A&M who undertook the study while a postdoctoral researcher at NCAR.