Friday, April 26, 2019

More tornadoes and more deaths in Texas and Louisiana, April 24 and 25, 2019

Yet another tornado episode occurred in the South on Wednesday, April 24 and Thursday, April 25 in Texas (TX) and Louisiana (LA), adding 2 deaths to 2019's tornado toll during the early morning hours of April 25 .

The tornado pictured above near College Station, TX at mid-afternoon on Wednesday didn't kill anyone, but was rated EF2 and was widely photographed.  This tornado is interesting because it formed rapidly from a small cell that was not very impressive on radar (see images below).  As a result of being small and relatively distant from radar sites, there was no tornado warning at the time.

Here's the setting at 4:00 pm CDT... A surface low was just west of College Station (CLL), with a stationary east-west outflow boundary (from earlier storms to the north) also located near CLL.  The circled "S" shows the location of the small supercell that produced the tornado as it crossed this boundary going from south to north:
Why did the storm produce a tornado so quickly?  It probably had something to do with the boundary; remember from other posts I've made this year that tornadoes definitely seem to favor boundaries.  This is suggested by the SPC mesoanalysis images below at 4:00 pm CDT:
Notice in the 2nd graphic above (SRH, or storm-relative helicity) how there was a band of enhanced low-level wind shear along and north of the outflow boundary, likely promoting low-level storm rotation in that area.  This same area overlapped low-level CAPE (surface-based instability that can promote rapid vertical stretching) along and south of the boundary, likely making a very favorable environment for a tornado as the cell crossed the boundary in a narrow corridor before it encountered cooler air farther north.

There's not room to show it here, but the RAP model 1-hour forecast sounding at 4:00 pm for CLL had 1200-1300 J/kg of total MLCAPE, over 300 m2/s2 of 0-1 km SRH (a nicely curved low-level wind profile), and over 100 J/kg of 0-3 km MLCAPE in low-levels.  Along with over 60 kt of deep-layer wind shear, and lifting condensation level heights (LCLs, or estimated cloud bases) well below 1000 m to reduce cool outflow, this was a very favorable environment for supporting tornadoes along and just north of the outflow boundary.  Farther north of the boundary, storms became increasingly "elevated" (less surface-based regarding the instability feeding them), which would be less supportive of tornadoes.

Later that evening and night, a long-lived supercell producing multiple tornadoes emerged from the southern segment of a line of storms over far eastern TX and moved to north-central LA (see radar images below), eventually killing 2 people at Ruston, LA just before 2:00 am CDT on April 25.
The EF3 damage from this tornado at Ruston included this motel (photo below), which lost most of it's 2nd floor.  Although no one died at this location, this is a pointed reminder that shelter below ground (or at least in a small interior 1st floor room) is crucial in surviving tornadoes!

The surface map at 1:00 am CDT early morning on April 25 showed an outflow boundary that was now oriented north-northeast to south-southwest over east TX to northwest LA, with the long-lived supercell (circled "S") moving northeastward roughly along this boundary for 2-1/2 hours:
A subtle and weak warm front progressing northward over central and northern LA (thick red dashed line on surface map above), suggested by slightly backed surface winds and larger dew points by 3-4 degrees F, may have also helped in sustaining the tornadic storm.

SPC mesoanalysis images at 1:00 am CDT (below) also showed a favorable environment for supporting supercell tornadoes over west and northwest LA along and just ahead of the outflow boundary and subtle warm front:
MLCAPE (not shown) was around 1200 J/kg, 0-1 km SRH > 300 m2/s2, and low-level MLCAPE > 100 J/kg over a sizable area through which the supercell was moving.  The orientation of the boundaries apparently allowed the northeastward-moving supercell to take advantage of this supportive environment for tornadoes over a long period of time, while utilizing increased shear and convergence along/near the same boundaries.

Here's a quick look at the larger scale upper air pattern via a 500 mb forecast map from the NAM model, valid at 1:00 am CDT:

A large, intense midlevel trough was moving from TX eastward, forcing ascent over LA with the "spreading jet" pattern ahead of it inducing general upward motion and strong wind dynamics across the area where tornadoes occurred overnight.

Tragically, the tornado outbreaks in the southern U.S. on Feb 23, March 3, April 13, and now April 25 have claimed a total of 29 lives so far in 2019.  Other deaths have resulted from thunderstorm winds blowing trees down on houses and vehicles, and flash-flooding.

As severe weather season continues in 2019, know how to stay safe... watch Surviving the Storm: What Chasers Want You to Know, free on YouTube:

- Jon Davies 4/26/19

Wednesday, April 24, 2019

Landspout-type setting produces first tornadoes of 2019 for Oklahoma & Kansas on April 17

It has been a slow start to tornado season in Oklahoma (OK) and Kansas (KS), very much like last year.  Last year's first Kansas tornadoes were on May 1, and Oklahoma's were on May 2.   But this year's first tornadoes in both states (see images above) were a week ago on April 17 near Shattuck, OK (top image) and Wellington, KS (bottom image).  These appeared to be due to mainly "landspout" processes (storm updraft stretching directly over a pre-existing wind shift boundary), rather than last year's supercell tornadoes on May 1 and May 2 when plentiful low-level wind shear was present over large areas.

The surface map setting at 3:00 pm CDT on April 17 (below) showed a stationary front present as a sharp wind shift over northwest OK into the northeast Texas (TX) panhandle:

This sharp boundary could be seen on high resolution visible satellite images at early to mid afternoon:

Radar reflectivity images from Amarillo showed a new storm developing on this boundary at 3:00 pm CDT in the northeast corner of the TX panhandle (1st image below), north of other storms farther to the south:

A radar image less than an hour later (middle image above) showed the same cell exploding with a red core.  At this time (3:52 pm CDT), a rope-like tornado (not pictured) was on the ground northeast of Canadian TX, while new cells began to develop northeastward into OK along the boundary.  The last radar image above at 4:49 pm CDT showed a supercell storm had developed in northwest OK on the boundary near Shattuck, where two tornadoes were occurring simultaneously (see the photo at the top of this post).

Although at least a couple of the storms in the northeast TX panhandle and northwest OK were supercells (storm rotation indicated on radar, but not shown here), my opinion is that the actual tornadoes were due mainly to "non-mesocyclone" or "landspout" processes, or at least that the tornadoes were a "hybrid" event (combining non-supercell and supercell processes).  This is suggested by the SPC mesoanalysis images below at 3:00 pm CDT before the tornadoes:

These graphics clearly show the stationary front wind shift boundary in the surface vorticity field (blue analysis lines and thick black dashed line on 1st image), steep low-level lapse rates from surface heating (axis of red dots, 2nd image), and no low-level wind shear (storm-relative helicity or SRH, 3rd image) over the northwest OK and the TX panhandle  where the tornadoes occurred during the following couple hours.  These are all factors that relate to landspout tornado formation.

It's also interesting that a supercell farther south that moved into west-central OK (visible on the last two panels of the radar images shown earlier) did not produce any tornadoes, although it was later tornado-warned based on radar.  This storm was within the warm sector and away from the stationary front.  But, without the sharp boundary wind shift, it had no source for low-level vorticity ("spin") as a result of the lack of SRH (low-level wind shear) over western OK during the afternoon.  (Low-level wind shear helps generate low-level rotation and mesocyclones that can give birth to supercell tornadoes.)

Here are a couple more images of one of the two tornadoes near Shattuck, OK between 4:30 and 5:00 pm CDT along the boundary (probably the tornado at right in the earlier photo):

Another tornado (a landspout) occurred in the central TX panhandle just after 5:00 pm CDT east of Amarillo, along the same boundary that extended some distance southwestward:

And, essentially the same boundary also extended into south-central KS as a pre-frontal trough, helping to generate the landspout tornado pictured earlier near Wellington around 6:40 pm CDT.

Again, this case seems to be a good example of how landspout processes can produce multiple tornadoes via strong low-level stretching directly over a sharp wind shift boundary with little immediate temperature contrast (see here for another more prolific case).

Jon Davies - 4/24/19

Wednesday, April 17, 2019

April 13-14, 2019 southern states tornado outbreak

The April 13-14 tornado outbreak across the southern states killed 3 people due to tornadoes (2 in Texas, 1 in Mississippi) and, sadly, several others died due to wind and flooding.  Tornado-wise, there were at least 10 EF2 tornadoes and 2 EF3 tornadoes on the 13th.  

The tornadoes pictured above were near Hearne, Texas (EF3) at late morning, and at Vicksburg, Mississippi at late afternoon.  Damage is also shown in east-central Texas between Weches and Alto where 2 people lost their lives in a tornado around noon-time.  Unfortunately, two children also died nearby in a vehicle from a falling tree in severe thunderstorm winds. 

The outbreak was forecast well by NWS meteorologists, the third deadly tornado outbreak across the South in 2019.  I won't do a detailed analysis here, as there was nothing particularly unusual about it.  But I did want to point out how most of the strongest tornadoes occurred with storms near and just south of the warm front that moved from east-central Texas to northern Mississippi during the day.

Severe weather forecasters know very well that the zone along and near warm fronts tends to have increased low-level shear with backed (more easterly) surface winds, so if strong warming and moistening is also occurring in that area, it becomes a favored location for supporting tornadic storms.  That can be seen on the graphics below showing warm frontal locations and the SPC effective-layer significant tornado parameter:

Even though environments well south of the warm front also looked favorable for tornadoes, the area along and just south of the warm front tended to provide the focus for most of the tornadic storms in this outbreak.

The associated upper system plowing across Texas and into the Dixie states was strong, with an intense 500 mb trough (indicated by thick red dashes):

So, along with this strong upper system, the strong surface warm front and unstable/sheared environment ahead of the surface low generated a number of tornadoes.

Even with a strong southern jet stream propelling storm systems across the southern U.S. so far this early spring, periodic intrusions of cold air from Canada from a northern branch jet stream over the Great Lakes have kept tornadoes away from Kansas and Oklahoma so far in 2019 (Kansas is off to it's 4th slowest start for tornadoes since 1990, similar to last year).  We'll see if that changes soon.

- Jon Davies  4/16/19