Unusual Behavior of 2008 Windsor Tornado Affected Societal Response, New Study Reveals

Note to Reporters: Photos of Windsor tornado damage and a full copy of the report are available with the news release at http://www.news.colostate.edu/ or at http://www2.ucar.edu/staffnotes/research/scrutinizing-windsor-twister-all-angles.

Some Northern Colorado emergency decision makers reacted to the devastating 2008 Windsor tornado by assuming the storm would behave as most severe storms in the area do. That would be a reasonable response in most cases, but one that proved problematic during the 2008 tornado, according to a new study by federal researchers in collaboration with Colorado State University.

Tornadoes that typically touch down on the Eastern Colorado plains move east and seldom threaten the more densely populated Front Range, but the tornado that struck Windsor on May 22, 2008, took an unusual track to the northwest into a heavily populated urban area. This behavior caught some officials in Weld and Larimer counties by surprise, according to the study led by the National Center for Atmospheric Research, or NCAR, and the Cooperative Institute for Research in the Atmosphere, or CIRA, at Colorado State University.

The tornado killed one person south of Windsor and left $193.5 million in damages in its wake, making it the costliest tornado in Colorado history.

In a forthcoming issue of “Weather and Forecasting,” an American Meteorological Society journal, the report examines the meteorological events that took place that day and concludes that, despite major efforts by the National Weather Service to alert the public to the storms, official warning lead times do not tell the full story.

Detection and warnings captivate the meteorological community, but the societal side of the story is equally important: how people receive, interpret, and react to these warnings, the study said.

Russ Schumacher, an NCAR researcher who is now at Texas A&M University, led the study. Co-authors include Daniel Lindsey of the National Oceanic and Atmospheric Administration in Fort Collins, Andrea Schumacher, Jeff Braun and Steve Miller, researchers at Colorado State’s CIRA, and Julie Demuth with NCAR. This research team includes experts in both meteorology and the societal aspects of weather events.

“Jeff (Braun) and I talked shortly after the event – we were marveling at the dramatic weather events that had just unfolded so close to home and the unique challenges that it must have posed to local emergency managers," said Miller, deputy director of CIRA. "Underscoring the state of confusion, we were receiving warnings of the tornadic threat via e-mail long after the actual storm had passed on to the northwest. It got us to thinking about the possible breakdowns and bottlenecks in the system that can render ineffective or obsolete even the best forecaster guidance that’s based on hundreds of millions of dollars in technology and infrastructure.”

“Some Northern Colorado decision makers understood where the tornado was headed, but others relied on their history with tornadoes on the Front Range generally moving to the east,” Lindsey said. “In Oklahoma City, for example, severe storms occur all the time, so the emergency managers there are generally more accustomed to these situations.”

Researchers conducted 15 interviews with anonymous emergency responders, school officials and decision makers in Northern Colorado. The group was not intended to be a representative sample; interviewees were selected to collect diverse detailed information.

“We went back and looked at the tracks of a bunch of different tornadoes over the last 60 or so years,” Lindsey said. “Most tornadoes in this area do move to the east. As a result, decision makers needed to listen closely to the warning which stated that this particular storm was moving to the north-northwest, toward Windsor. Fortunately, many people did get the warning information in time and were able to get out of harm’s way.”

Unusual factors of the Windsor tornado:

• The storm formed in the late morning hours, not in the late afternoon.
• Once the storm formed north of Denver International Airport, it moved north to northwest toward population centers.
• The tornado itself was “surprisingly strong” – rated an EF3. (Significant tornadoes are those rated EF2 or higher on the Enhanced Fujita scale.)

Researchers found that some decision makers in northern Colorado have already recognized those pitfalls and addressed some of them. For example, some groups are ensuring they’re more prepared for cell phone and power outages. Others are reviewing procedures for text message warnings or hosting drills with the lights off to more accurately simulate conditions experienced during a storm.

“Several organizations are upgrading internal communications systems,” Lindsey said. “This tragedy reminded people that they needed to have a plan in place, that this kind of thing can happen this close to the Front Range.”