A summary of the recent presentation by Joshua Wurman (University of Alabama Huntsville), and Karen A. Kosiba, entitled “Radar Determined Intense Tornado Frequencies, Wind Profiles and the BEST Tornado Study,” presented at the 41st International Conference on Radar Meteorology, August 25-29, 2025, Toronto, Ontario, Canada. Click the link above for the abstract and recorded presentation.
Intense tornadoes can cause catastrophic damage and fatalities. But the relationship between tornado winds and damage is poorly known. Tornadoes rarely impact densely built-up areas, so robust damage-based wind estimates, using well-built structures, are uncommon. Tornadoes are only rarely observed by proximate radars, so radar-based wind estimates based on finescale and low-level data are also uncommon. It is very rare that a tornado damages a large number of structures while being observed by a very proximate radar such as a Doppler on Wheels, or DOW. And, it is even rarer that there is high-quality video and radar data documenting wind effects on structures. Until now.
On May 21, 2024, an intense tornado passed through Greenfield, Iowa, destroying several dozen houses and other structures, killing several people. The tornado tracked > 30 km at speeds sometimes exceeding 30 m s–1. As part of the Boundary-layer Evolution and Structure of Tornadoes (BEST) project, a fortuitously placed DOW scanned the tornado every 7 s, from ranges of 800 m to 1.5 km, as it destroyed much of Greenfield, while another DOW observed the same tornado every 7 s as it destroyed several wind turbines. Storm chasers collected well-documented data of several of the wind turbine failures.
This unique dataset provides the first-ever opportunity for an integrated analysis using multiple proximate DOW mobile radars and multi-indicator damage assessments, combined with navigated synchronized videos of an intense, damaging, and deadly tornado.
DOW data revealed Doppler velocities near 121 m s–1 (~270 mph) and tornado ground-relative winds near 136 m s–1 (~304 mph), among the strongest ever measured. Independently conducted damage surveys by the National Weather Service and the DOW team resulted in inferred winds of at least 185–190 mph at some locations. Comparison of wind speed assessments informed by the in-development updated version of the EF scale of the American Society of Civil Engineers were compared with DOW-determined maximum wind speeds.
Tornadoes Cross Greenfield
Peak DOW-determined winds were typically about 20% stronger than those suggested by damage. This is not surprising, due to a number of factors: complications in determining wind speeds from either of these methods (both of which are likely associated with +/– 20% error bars); most damaged structures were single-family homes, which may have been destroyed before being impacted by the strongest winds; the height of radar observations (30 m) was different than the damage measurements (about 3 m); many houses may have been sheltered by surrounding houses and trees; and the duration of the most extreme winds at any house may have been less than 1 s, since the tornado was very small and moving extremely fast.
Synchronized, high-quality aerial and ground-based storm chaser video revealed the precise timing and mechanisms of wind power turbine failure. At least one turbine was destroyed well outside the visually manifested tornado. Others may have been destroyed before peak winds impacted them. These failures were compared to DOW-determined wind fields to examine what wind speeds, and time-histories of wind speeds, were associated with damage to, and the failure and survival of, the wind turbines.
Tornado Topples Wind Turbines
(green) and destroyed (red) turbines are overlaid on a map of peak wind speeds. Video was also used to understand how the history of
wind speeds affected these structures.
We found that all turbines that experienced winds of at least 69 m s–1 (~154 mph) toppled, while all turbines experiencing peak winds less than 65 m s–1 (~145 mph) survived. Some turbines experiencing winds of 65–69 m s–1 survived, while some fell. This provides a well-bounded threshold wind survivability for these turbines. We are in the process of analyzing the wind speeds at the moments of failure, which may have occurred before peak winds occurred.
These detailed comparisons of radar, damage, and video data in an intense tornado increase our understanding of what wind speeds cause different kinds of damage, how to assess tornadoes that may destroy structures before peak winds occur, and how to improve our ability to assess and mitigate tornado hazards.
