A Close-Up Look at Last Week’s Perplexing Colorado Twisters

Early June is the most common time for tornadoes in Colorado, and last week served up a dramatic example. NOAA’s Storm Prediction Center logged 13 preliminary tornado reports in Colorado on June 4 - 6. The tornado-generating thunderstorms were spawned by weak upper-level impulses from the southwest, combined with generous low-level moisture feeding into the state from the Gulf of Mexico. Two widely photographed twisters that seemed to bend the rules of tornado formation, both on June 4, caught the attention of national press and the blogosphere.


Figure 1. This damaging tornado south of Berthoud, CO, on Thursday, June 4, was visible from a deck in Broomfield, about 20 miles to the south. Image credit: Vince Miller.


A westward-moving tornado strikes the Front Range
The strongest Colorado tornado in seven years moved from the adjacent plains toward the foothills between Boulder and Fort Collins on Thursday afternoon. Rated by the NWS Denver/Boulder office as an EF3, the tornado began about 3 miles south of the town of Berthoud, then tracked west-northwest for about 5 miles, dissipating as it approached higher terrain. Several homes were destroyed and about a dozen others damaged. The highly visible tornado was seen by many residents of Longmont and the far north suburbs of Denver.


Figure 2. A home severely damaged in the tornado near Berthoud, CO, on June 4. Image credit: NWS Denver/Boulder.


Although Colorado has recently been averaging about 60 tornadoes per year, they’re uncommon west of Interstate 25 in the populated corridor from Denver to Cheyenne, Wyoming. Moreover, westward-moving tornadoes are quite unusual anywhere in the country. Parent thunderstorms are typically steered by flow through the deep layer of the atmosphere in which the storms reside. Typically, this flow has a west-to-east component ahead of upper-level impulses, where storms are most likely to develop. A supercell thunderstorm will sometimes move to the right of the vertically averaged wind (e.g., toward the east rather than the northeast), which often increases its tornado-producing potential.


Figure 3. High-resolution data from the CSU/CHILL research radar shows the thunderstorm in Larimer County, CO, at 6:34 pm CDT on Thursday, June 5, as it was producing a tornado south of Berthoud. Top image is reflectivity (precipitation); bottom image is velocity (winds toward/away from the radar). The areas marked “reflectivity minimum” (top) and “radial velocity couplet” (bottom) coincide with the approximate location of the tornado. Image credit: Pat Kennedy, CSU/CHILL National Radar Facility. To see the westward development of the storm while the tornado was occurring, check out this loop of imagery from the NWS radar in Cheyenne, WY. The tornado developed near the Boulder/Larimer county line just north of Longmont (LMO on the radar loop). Image credit: NWS/CIRA.


Aside from last week’s tornado near Berthoud, the only other F3/EF3 tornado that moved west of I-25 in Colorado developed around noon on May 22, 2008--and it also had a westward component to its motion. This powerful twister, the most expensive in Colorado history, carved a wide, 39-mile-long path that extended to the north-northwest from near Platteville through the town of Windsor, damaging hundreds of homes there. A team led by Russ Schumacher and Dan Lindsey (Colorado State University) analyzed the event in 2010 in a paper for the journal Weather and Forecasting. Upper-level winds on that day were from the southeast, and instability was unusually strong for the region, which enabled the Windsor storm to develop as a fairly classic supercell, albeit with an unusual track and location. As for last week’s Berthoud storm, which had a much more pronounced westward motion, radar data indicates that it was in a “back-building” stage, growing quickly toward the west while the tornado itself was occurring (see Figure 3). “The westward motion appears to be driven by some combination of storm dynamics and low-level upslope flow,” says Schumacher. Lindsey summarizes the common thread between two events: “It seems to me that virtually the only way to get a real supercellular tornado west of I-25 is to have a storm moving with a component toward the west. And this one certainly qualifies.”

How can a tornado spin anticyclonically?
Social media has been abuzz over the last few days with spectacular images of an anticyclonic tornado near Simla, Colorado, northeast of Colorado Springs. Perhaps the most iconic photo was taken by accomplished photographer and storm chaser Kelly DeLay. Rather than going for a close-up, DeLay captured the entire storm in its full majesty--and the resulting image shows not just one but two twisters.


Figure 4. A supercell thunderstorm near Simla, Colorado, produced two tornadoes on June 4--one anticylonic (lower left) and one cyclonic (lower right). The thunderstorm itself was rotating cyclonically. Image credit: Kelly DeLay.


Anticyclonic tornadoes spin in a clockwise fashion, in contrast to most of their peers. Most supercells in the Northern Hemisphere rotate cyclonically (counterclockwise), as do the tornadoes they produce. This isn’t a direct result of the Coriolis force, a function of Earth’s rotation that causes hurricanes and other large low-pressure centers to spin cyclonically. Thunderstorms and tornadoes are too small to be dictated by the Coroilis force (and so are toilets!). Instead, the juxtaposition of winds at different heights tends to cause a supercell in the Northern Hemisphere to spin cyclonically, especially when it’s moving toward the right of the vertically averaged wind. Sometimes an anticyclonic tornado (usually fairly weak) will develop as a satellite twister on the fringe of a tornado-producing, cyclonically rotating supercell. This appears to be the process that drove the Simla tornado shown in Figure 4 above. There’s another way that an anticyclonic tornado can develop. If a supercell happens to move toward the left of the vertically averaged wind, rather than toward the right, it will spin in the other direction--anticyclonically--and it can thus produce an anticyclonic tornado. Several events like this have been documented in case studies, including a 1998 event in California and one in South Dakota in 2006.

For more background, including perspective from tornado researcher Josh Wurman, see this informative June 8 article by Angela Fritz at Capital Weather Gang.

Tropical Storm Carlos staying offshore of Mexico
Tropical Storm Carlos continues to churn in the Pacific Ocean, about 200 miles south of the Mexican coast. With very warm waters of 30°C (86°F) beneath it and wind shear a moderate 10 - 20 knots, Carlos appears destined to undergo slow intensification through Monday. Satellite loops and radar out of Acapulco shows that the outer spiral bands of Carlos have pushed onshore, but the storm's west-northwest motion, parallel to the coast, should keep the heaviest rains just offshore. Heavy rains of 3 - 6 inches will likely affect portions of the Mexican coast over the weekend, and if Carlos takes a path slightly closer to the coast than expected, dangerous flooding rains will occur. By Wednesday, Carlos may pose a threat to Mexico’s Baja Peninsula, but the cooler waters and more stable air the storm will encounter as it approaches Baja should cause Carlos to weaken below hurricane strength before a potential landfall there. We’ll have a full update on Carlos by Monday morning at the latest.

Have a great weekend, everyone!

Bob Henson and Jeff Masters


Video 1. The late Jim Leonard, a famed hurricane chaser, captured the first known video of a Northern Hemisphere anticyclonic tornado on June 6, 1975, near Freedom, Oklahoma. More details on this and other events can be found in Jim’s online biography. The embedded video above can also be viewed directly at YouTube.


Figure 5. A high-resolution GOES-14 image from around 6:00 pm CDT on Thursday, June 5, as intense supercells were raging across northeast Colorado. The Simla tornado emerged from the southernmost storm, while the Berthoud tornado was in the northermost storm in Colorado. In this animated satellite loop, crosshatch symbols show the approximate locations and occurrence times of the tornadoes. The images in the loop were collected every minute by GOES-14. Image/animation credit: Dan Lindsey, NOAA/CSU/Cooperative Institute for Research in the Atmosphere.