Ceaseless Wind

Looking NNE from Seymour on 3/12/06. From video shot by Matt Gingery.

In terms of shear numbers and intensity, this outbreak was much worse than the May 4th, 2003 outbreak.

16 different tornadoes tracked through the area indicated on the maps below.  This total does not count the rest of Missouri, Arkansas, Oklahoma, Kansas and Illinois.

Tornado Tracks

Tornado Tracks

Radar at 9:31 pm 3/12/06

Doppler Radar: it can sense whether raindrops or hail are moving away or toward the radar site. This principle also works in an environment without storms because radar beams can actually bounce off of large cloud droplets and turbulent pockets in the atmosphere.  Doppler codes winds that are blowing toward the radar with green, wind blowing away as red.

The first image captured at about 6:00 pm shows winds near the surface of the earth coming from the SSW.  Two hours later at 8:00 pm, the winds had turned slightly or backed  to the SSE.

This turning of the wind in the lowest layer of the atmosphere has been shown to have a profound impact on the potential for thunderstorms to produce tornadoes.

This can also be seen in the Conway profiler data.  Note not only the backing winds but also the incredible overall wind speeds which increased rapidly with height.  This kind of wind profile helps to give supercell thunderstorms longevity.

Doppler at 6 pm

Doppler at 8 pm Conway, MO Profiler on 3/12/06

Conway, MO Profiler on 3/12/06

Eerie Similarity: The map below is actually two maps superimposed: one shows the track (heavy red line) of the May 4th, 2003 tornado while the other shows the track (thin red line) of the March 12th, 2006 tornado.  The tracks are exactly on top of each other near Verona and never drift more than a mile from each other until the very end.  The 2003 storm turned left near the end of its track and struck Battlefield while the 2006 storm maintained more of a straight line.  These maps were created by the Springfield National Weather Service Office and can be found in their original form here (2003) and here (2006).

Thick Red: 5/4/03, Thin Red: 3/12/06

Tornado Tracks OK/AR 3/12/06

On this date in 2009,  a tornado intermittently caused damage to homes and businesses along with trees and power lines starting southwest of the interstection of Republic and Scenic extending northeast to the intersection of Fremont and Catalpa. Winds up to 100 mph…or EF1 on the enhanced fujita scale…was indicated by the damage.

Ordinary radar echoes at 10:45 pm and 10:49 pm on 2/10/2009

Storm-Relative Velocity images at 10:45 pm and 10:49 pm on 2/10/2009

Normalized Rotation (NROT) at 10:45 pm and 10:49 pm on 2/10/2009 (standard MDA settings)

A tornado watch was in effect but Greene County was not under a warning of any type when this tornado occurred!  The reason has to do with the type of tornado which occurred.

Small tornadoes can occur if rotation which exists at ground level finds itself under the updraft of a storm. The updraft stretches the column of air and decreases the radius of rotation.  Since angular momentum must be conserved, the speed therefore must increase (cue obligatory ice skater example).

In theory, this can happen under any updraft.  As a practical matter however, meteorologist look for certain areas in thunderstorms that are favored for tornado development and would likely examine such areas more closely to see if there was rotation.

In this case, the rotation was overlooked because the radar signature gave no cause for concern (I call it a pimple on the radar screen!) and the rotation was so broad and short-lived that computer algorithms simply didn’t register it as a threat.  Even if perchance I had seen this rotation, I might have raised an eyebrow but I don’t think I would have pulled the warning trigger if I were the one with the warning responsibility!  Moreover if I had some magical gift of radar interpretation or inside knowledge and decided to cry tornado, would it have done any good?  The twister was on the ground for a total of five minutes!

Super Tuesday Outbreak 2008

This supercell storm pictured below took on bow echo shape as it dropped a tornado right on top of Gassville, Arkansas. Bow echo tornadoes can be powerful and destructive. Unlike tornadoes emerging from the so-called “classic” supercell, bow echo tornadoes are often shorter-lived and their signature on radar can be less obvious.

Damage surveys put this tornado at EF2, on an EF0-EF5 scale.

This storm was moving to the northeast at around 60-65 mph!

Radar image at 5:06 pm on February 5th, 2008, a few mintues before a tornado strikes Gassville, AR

The Gassville storm was on the northwest edge of a widespread outbreak of tornadoes. Another supercell, this one with a classic radar configuration, raced northeast through central and northeastern Arkansas during the same evening. The first tornado report came out of Ola in Yell County, Arkansas. The supercell then continued northeast, spawning the longest tracked tornado in Arkansas history, killing a total of twelve people. See the complete report from the Little Rock, NWS office.

Long-track Tornado!

Please read more about this outbreak in the official National Weather Service assessment below.  Among other things, the report found that many people didn’t think a tornado watch in February needed to be taken seriously and mobile homes are still the riskiest place to be during a tornado. 

Another link details the Storm Prediction Center’s efforts to raise awareness of the threat of nighttime or off-season tornadoes.

NWS Service Assessment of the Outbreak

SPC Nighttime Tornado Outlooks

NWS Snow Total Map for Missouri

Well, the storm from last week is a few days behind us.  Temperatures have warmed, the thaw has begun. 

It was an interesting storm, first tracked about five to six days before it occurred.  For days, it looked consistently like a Thursday evening and overnight snow.  Then, a intense “cold side” portion of the storm began to take shape which was clearly targeting Friday.

As Thursday evening rolled around, the precipitation started surging northward. Arkansas, at least the portion within the KOLR/KSFX viewing area, started getting light freezing rain followed by sleet and eventually snow.  I sat and watched the radar loops and noticed that the precipitation was “stuck”; it just wasn’t advancing northward into Missouri that evening.  A check of was was going on directly overhead in Springfield via a “balloon launch” at 6 pm showed that the precipitation was trying to fall through an extremely dry layer of air.  As a result, it wasn’t reaching the ground.

This caused me to revise my snowfall projections downward (pictured below) between the 6 pm KOLR show and the 9 pm KSFX show.  The real kick in the head and part of the sweet irony which often accompanies weather forecasting decisions is that if I had left the early projection alone, it would have been closer to what actually happened!  This is because the second wave of snow on Friday exceeded expectations.  It would have been a classic case of “right for the wrong reasons!”

That second wave really started to look interesting on Thursday afternoon.

Everyone was expecting a lot more snow in Springfield Friday morning than what actually occurred of course with 1 to 1.5 inches instead of the 4-6 I reasoned we would end up with.  The above mentioned dry air is to blame.  It didn’t just magically appear; I saw it, it was one reason why, if I can hang my hat on this at all really, my snow projections tended to lean on the conservative side.  That and questions about precipiatation type and changeover times.  When it came down to it, the dry air was just a little drier and the “surge” of moisture both couldn’t overcome it as quickly and was being directed a bit more to the southeast of Springfield specifically. (note that by Friday morning, Mtn. Home, Arkansas had indeed picked up about 4 inches, just the start of a winter wonderland for those folks!)

Despite the failure of the Friday morning snow to be as heavy, I think the storm as a whole was forecast very well.  Everyone had at least 5 days to prepare.  Changes and updates to the forecast were handled quickly and accurately.  The second surge was handled very well.

If some people thought that Springfield would wake up Friday to a foot of snow or that it wouldn’t be snowing into Friday evening, I don’t know what to say except to listen carefully to what meteorologists are saying.  I also strongly suggest checking this blog and other social networking sites.  For instance, I knew shortly after the measurement was taken at about 7 pm Thursday evening that the morning totals would have to be scaled back. 

Tuesday 10 pm Snow Projection

Wednesday 10 pm Snow Projection

Thursday 6 pm Snow Projection

Thursday 10 pm Snow Projection

It was one year ago, January 26-28 2009, that a devastating ice struck northern Arkansas, extreme southern Missouri and then moved eastward on into Kentucky.

Links:

http://icestorm2009.uark.edu/

http://www.srh.noaa.gov/lzk/?n=win0109a.htm

http://www.realclearwx.com/ws12009.htm

National Snow/Ice Depth Loop

Satellite Image During the Storm

Ice storm pictures by Jayme Genz, Mtn. Home, AR

300 mb Map 00z 1/21/2010 ( 6 pm 1/20/10)

An outbreak of severe storms and tornadoes occurred Wednesday evening, January 20, 2010, over portions of northeastern Texas and Louisiana including an EF2 tornado in Van Zandt, County, TX:

NWS Dallas-Fort Worth Preliminary Map & Text

NWS Shreveport Text

The map above shows the area of the tornado outbreak was in the left exit region of a powerful jet stream stretching across the southern U.S. This means that the difference in wind speed between the surface and the jet stream level was great and getting larger as the jet poked its way into the region.  Combining this upper level wind pattern with unstable air at the surface is nearly always a recipe for trouble.

With this strong dynamic set-up, the thermodynamics just need not impede development.  Or stated another way, the atmosphere had to be somewhat unstable with nothing to prevent convection from going up.  CAPE, a measure of instability was more than sufficent for thunderstorms with values in outbreak area ranging from 1000 to 2000 with little to no cap.

Surface-Based CAPE 6 pm 1/20/10

Another key ingredient to the production of tornadoes in supercells seems to be high values of something called storm-relative helicity (SRH).  This can be thought of as the tendency of air low to the ground to want to rotate cyclonically, potentially imparting this rotation to the updraft of a supercell thunderstorm.  “Storm-Relative” just means the motion of the storm in the wind field has to be taken into account as well.

Values of SRH in the 0-3 km layer in the outbreak area were between 200 and 250.  While there are no clear-cut threshold for these helicity values, the values reported in this outbreak were above what is considered to be the lower threshold of significance.

SRH 0-3 km 6pm 1/20/10