Brief Case Studies of Non mesocyclone Tornadoes in Small SRH and or High LCL Environments        by Jon Davies
(Important ingredients: a well-defined boundary, steep low-level lapse rates, and significant low-level CAPE)

An informal study I did about tornadoes in small SRH and/or high LCL environments showed a significant signal highlighting low-level thermodynamic settings featuring steep low-level lapse rates (e.g., 0-2 km or 0-3 km above ground) and notable moisture depth resulting in the presence of low-level CAPE (e.g., MLCAPE below 3 km AGL km).  In theory, such environments would have potential for enhancement of  low-level stretching through rapid parcel ascent (steep low-level laspe rates) with reduced mixing and entrainment for parcels below cloud base (notable moisture depth with no significant temperature inversion present to retard rising parcels).  As an example of these characteristics, see this RUC sounding profile associated with the first case below. 

An awareness of such characteristics may be useful in the short term (usually only 1-3 hrs) to highlight some environments where tornadoes from mainly nonsupercell/non-mesocyclone processes may be possible.  To increase the odds of such a "mesoscale accident", the combination of low-level lapse rate and low-level CAPE would probably need to be "maximized" along a well-defined windshift boundary (vertical vorticity source; e.g., Wakimoto and Wilson 1989) where thunderstorms were expected to development (see this composite).  Stretching of vorticity by thunderstorm updrafts along such boundaries could then result in non-mesocyclone tornadoes.  If some notable deep shear were present along with at least some SRH, it is possible that some storms might even develop marginal supercell characteristics.

The following cases are brief examples highlighting storms developing along boundaries in environments having steep low-level lapse rates, some with significant low-level CAPE:

Case 1:  27 August 2004 (south central KS)

from 23z RUC analysis:
<Sig Tor Parameter    <0-3 km lapse rate
<0-3 km MLCAPE   <tornado reports
<0033z Base Refl.     <F2 tornado S of Wellington KS

In this case, the F2 tornado had no pre-existing mesocyclone on radar, and developed rapidly from a new updraft at a boundary intersection that was probably a focus of increased vertical vorticity.  It is possible that the low-level thermodynamic environment combining steep low-level lapse rates and low-level CAPE enhanced low-level stretching at this intersection.  It can be seen that the tornado had full condensation all the way to ground from a high cloud base, not the visual appearance of a weak "landspout".  A severe thunderstorm watch was in effect over the area at the time.

Case 2:  18 April 2004 (Minnesota, Iowa, Nebraska, South Dakota)

from 22z SPC mesoanalysis:
<21z surface              <0-1 km SRH
<MLLCL height        <0-3 km lapse rate (LR3)
<0-3 km MLCAPE  <best overlap LR3 and MLCAPE3
<2230z refl. mosaic  <tornado reports

Most of the tornadoes for this event appeared to be nonsupercell/non-mesocyclone in nature along the boundary southwest of the surface low where cloud bases were high, but low-level laspe rates were steep and there was low-level CAPE present just above the high LCL heights.  A severe thunderstorm watch was originally in effect in the area where most of the tornadoes occurred.

Case 3:  19 May 2003 (west central and northwest Texas)   nontornadic case

from 22z SPC mesoanalysis:
<21z surface             <0-3 km lapse rate
<0-3 km MLCAPE  <2146z base reflectivity

This nontornadic case had separate areas of steep low-level lapse rates and low-level CAPE with no overlap, possibly reducing potenital for enhanced low-level stretching along the boundary.  This  may possibly have impacted the lack of tornado reports from storms that developed directly on the surface boundary that was apparently rich in vorticity.

Case 4:  28 August 2003 (west central Kansas)  nontornadic case typical of summertime environments

from 19z SPC mesoanalysis:
<19z surface             <0-3 km lapse rate
<0-3 km MLCAPE  <best overlap LR3 and MLCAPE3
<2015z visible satellite

This final case is an example of summertime in the central U.S... lapse rates were steep over a large area with low-level CAPE present throughout much of the air mass.   This can be relatively common at times in the warm season.  In such cases, careful assessment of the location of well-defined boundaries and whether storms develop directly on the boundaries may help.  On this particular nontornadic day, it is possible that the storm in west central Kansas may not have been properly aligned with the boundaries near the surface low in this weak shear environment.  Also, the fact that low-level lapse rates and low-level CAPE were "mazimixed" more toward the center of the air mass well south and east of the boundaries and surface low, instead of near the surface low, may have been a contributing issue, but that is only speculation.

Some comments about assessing settingw with short-term potential for non-mesocyclone tornadoes:
1) Sharp well-defined boundaries that are near-stationary or slow-moving with little temperature change across them (like a weak cool front or trough) seem to work best for producing non-mesocyclone tornadoes.  The most typical wind shift is a sharp one from south or south-southwest to west or northwest across the boundary.
2)  Storms on or near boundaries where low-level lapse rates are steep and low-level CAPE significant can produce tornadoes from storms with high cloud bases in high LCL environments.
3)  The above ingredients will be of little use over most of the western U.S. where mountainous terrain results in highly variable and inconsistent lapse rate and CAPE fields in low-levels.
4)  Boundaries oriented northeast to southwest seem to be the most common producers of nonsupercell/non-mesocylone tornadoes when steep low-level lapse rates and low-level CAPE overlap those boundaries where thunderstorms develop (see composite).

My hope is that some of the above material will prove useful to forecasters regarding short-term situational awareness of nonsupercell/non-mesocyclone tornadoes, which are clearly difficult to forecast.

Jon Davies -- updated 5/23/05                                        back to Jon Davies main page