Several long-lived supercell thunderstorms cranked out destructive tornado families across the Midwest from late Tuesday into Wednesday morning. At least 3 deaths had been reported, according to a weather.com summary
. Power was knocked out for tens of thousands of people as the wind-packing storms barreled east toward the Appalachians.
The NOAA/NWS Storm Prediction Center received 24 preliminary tornado reports for the period from 6:00 am CST Tuesday to 6:00 am Wednesday, with 2 additional reports as of late Wednesday morning. It’s possible that the final tally of tornadoes will come down from these numbers. Most of the reports were generated by just two sequences of supercell thunderstorms that spun out twisters for hours on end. One of these storm complexes tracked from far eastern Iowa into northern Illinois, while the other followed a track from far southeast Missouri through southern Illinois into southwest Indiana--a path eerily reminiscent of the nation’s deadliest tornado on record: the Tri-State Tornado of March 18, 1925
(see below).Figure 1.
Pat Harber of Perryville, Missouri, looks through the wreckage of her home after it was destroyed by a tornado on February 28, 2017. At least one person was killed when the tornado crossed interstate 55. (Jon Durr/Getty Images) Figure 2.
This large tornado was photographed at 5:36 pm CST Tuesday, Feb. 28, 2017, by associate meteorology professor Walker Ashley (Northern Illinois University) as it passed just northwest of Washburn, IL. Image credit: (c) Walker Ashley.
The first and more northerly of the day’s two major tornado sequences unfolded with a short line of fast-moving supercell storms from around 4:00 to 6:00 pm CST Tuesday. A large, wedge-shaped tornado struck near Naplate and Ottawa, IL, around 4:45 pm, causing one death and producing extensive damage. As this circulation weakened, it was succeeded by one in a storm just to the south that generated frequent tornado reports from around Washburn toward Long Point (see Figure 1 above and embedded tweet at bottom). Continuing east as a non-tornadic complex, the storms produced hail as large as golf balls across the south part of the Chicago area. Tuesday’s storms across northern Illinois were being surveyed on Wednesday by NWS staff from Chicago and Lincoln, IL; results will be posted on this NWS/Chicago website
A debris signature (often called a “debris ball”) is evident on this radar reflectivity image from the NWS Doppler radar located near Owensville, IN. The image was taken at 10:04 pm CST Tuesday, Feb. 28, 2017. Image credit: NWS/Paducah, KY
The other major tornado producer was a more classic, isolated supercell storm that maintained its identity for more than 100 miles across parts of Missouri, Illinois, and Indiana, spitting out tornadoes along much of its path. The radar signature of the storm was especially strong over far southeastern Illinois, with a clear debris ball evident (see Figure 3 above). CNN reported
that one man was killed and his wife seriously injured near Crossville, IL. The intense circulation passed extremely close to the NEXRAD radar stationed at Owensville, IN, near the NWS office serving the Evansville, IN, area (see Figure 4). Figure 4.
Stu Ostro (The Weather Channel) tweeted
this uniquely amazing radar image late Thursday night: “Effect of beam height on 3-D radar presentation as the #tornado-producing thunderstorm passed right over the Evansville radar site.” The tornadic circulation lies very close to the radar, with the core of the storm passing just to the northwest (into the page). Image credit: Stu Ostro, @StuOstro
Most of Tuesday’s tornadoes (red dots) occurred within the areas highlighted by the NOAA/NWS Storm Prediction Center at 10:30 am CST Tuesday with the top tornado probabilities. The hatched lines indicate an especially high risk of significant tornadoes (EF2 or stronger). Image credit: NOAA/NWS/SPC
.Another Tri-State event, but this time without the horrific death toll
NOAA’s Storm Prediction Center did a superb job with its Tuesday outlook. On Tuesday morning, it successfully identified the small corridor
where the southern tornado supercell tracked as being the area with the highest probabilities of significant tornadoes. Throughout the Midwest, tornado watches were issued well ahead of the day’s worst activity, and the fact that tornadoes were concentrated within long-lived storms allowed for plenty of lead time on many NWS tornado warnings.
The impressively small death and injury toll from Tuesday stands in stark contrast to what occurred with what may be the most analogous storm to the southern supercell: the Tri-State tornado of March 18, 1925. This event was identified as a tornado family in more recent research led by former SPC forecaster Bob Johns, with the longest continuous path length at 174 miles and frequent damage recorded over a stretch of 219 miles. “Witnesses saw a wedge tornado along most of the damage path and a large multivortex tornado in some areas,” noted Johns and colleagues in a 2013 open-access paper
in the E-Journal of Severe Storms Meteorology. The human tragedies left by this single-thunderstorm event remain unparalleled in U.S. history, with an official death toll of 695 people.
As shown below, the path of Tuesday’s supercell was remarkably similar to that of the Tri-State supercell as implied by damage records from 1925. (One can only imagine how that storm might have looked on NEXRAD radar!) Moreover, the broad, strong upper-level trough in place on Tuesday also closely resembled the one believed to have been in place during the Tri-State Tornado, as noted by Embry-Riddle associate professor Shawn Milrad, @shawnmilrad
. (Because upper-air observations were not made in the 1920s, such reanalysis maps are generated by inferring the upper-level conditions that would correspond to surface observations available from the period.) Of course, there are big differences between the 1925 and 2017 events: the Tri-State tornado was rated F5, whereas it appears very unlikely that any of the 2017 tornadoes will be placed in the F5 range.Figure 6.
Comparison of damage locations along the Tri-State tornado track from March 18, 1925, and selected preliminary tornado reports from Feb. 28, 2017. The first set of tornado reports, from far southeast MO into southwest IL, are a few miles north of the Tri-State track. After this supercell merged with another storm, the next batch of tornado reports occurred very close to the Tri-State track, from far southeast IL into far southeast IN. Original image from Johns et al., "The 1925 Tri-State Tornado Damage Path and Associated Storm System,"
E-Journal of Severe Storm Meteorology (2013).Another exceptionally rare February tornado outbreak
February tornadoes are rare in Illinois, and unheard of as far north in Illinois as where the Ottawa tornado touched down. According to the Tornado History Project
, Illinois experienced February tornadoes an average of once every six years between 1950 - 2007. However, their incidence of February tornadoes has increased dramatically over the past decade, with February tornadoes hitting once every other year, on average. Prior to Tuesday’s Ottawa tornado, there were only two other deadly February tornadoes in Illinois since records began in 1950, both of them well south of the Ottawa event: an F4 twister that hit Saline and Gallatin Counties on February 10, 1959, killing eight, and an F4 twister that hit just southeast of St. Louis on February 24, 1956, killing six. As of 10 am EST Wednesday, NOAA’s Storm Prediction Center
put Tuesday’s tornado tally in Illinois at seventeen, which would make it the state’s largest February tornado outbreak on record. Their previous largest February tornado outbreak came on February 20, 2014, when eleven twisters touched down, including two EF2s.
According to the Tornado History Project
, Missouri experienced February tornadoes an average of once every four years between 1950 - 2007, and once every two years between 2008 - 2017. The deadliest February tornado in Missouri hit St. Louis on February 10, 1959, killing 21 people. Including Tuesday’s Perryville tornado, there have now been six deadly February tornadoes in Missouri since 1950.
Tuesday’s outbreak was the second exceptionally rare February tornado outbreak in the U.S. this year. On February 25, a total of four tornadoes hit Pennsylvania
, and Massachusetts
. The EF1 that touched down in Massachusetts, damaging two houses, was their first February tornado on record--and at latitude 42.51°N, the most northerly February tornado recorded in the Northeastern U.S. since reliable records began in 1950.Figure 7.
Debris is seen from above Wednesday, March 1, 2017, where Tuesday's tornado destroyed homes in Perryville, Missouri. (Robert Cohen/St. Louis Post-Dispatch via AP) How did Tuesday’s tornadoes happen so far north?
One reason February tornadoes are practically nonexistent as far north as northern Illinois is that it’s very difficult to get a warm, moist air mass into the region during late winter without a very potent low-level storm system pulling the air northward from the Gulf of Mexico and a very strong upper-level trough passing through. Such a setup, though, can also lead to thunderstorms that are so widespread that they weaken the ability of more scattered rotating supercells to maintain themselves. In Tuesday’s case, however, the February “warm wave” of the last few days had already brought unstable air close to the region, so it only took a moderately strong surface feature (a 998-millibar low in northern Illinois) to get the unstable air into position beneath a powerful upper-level trough and jet stream. There was also just enough of a “cap” above the unstable air to keep a lid on thunderstorm development until daytime heating peaked and the tornado potential was maximized. “To me, today confirms that, as long as the ingredients come together, the calendar is not important--further evidence that big CAPE [a measure of instability] is not necessary on big days.” said NIU’s Walker Ashley. “It really is about that marriage of sufficient instability with solid [wind] shear, and just a peppering of convective initiation, to make for a big day.”
The severe weather threat continued into Wednesday ahead of the eastward-moving upper storm and associated cold front, with an enhanced risk of severe weather from northeast Mississippi all the way to southern New Jersey. Odds are especially high for damaging thunderstorm-related winds from the central Appalachians into the mid-Atlantic, perhaps including the Washington, D.C. area. Tornadoes should be less widespread and intense than on Tuesday, but a few are still possible, especially with any supercells that develop across northern Alabama and Georgia near the tail end of the cold front.
Bob Henson and Jeff Masters