Lubbock F5 Tornado
May 11, 1970
Jonathan D. Finch
Back to Lubbock Area Tornadoes WebpageBack to Amarillo Area Tornadoes Webpage
Detailed Tornado Cases for Amarillo and Lubbock
The famous F5 Lubbock
tornado occurred in a synoptically quiescient weather pattern. The flow at
mid and upper levels was
fairly light. This tornado occurred on the extreme SE edge of the westerlies--well out ahead of any mid level cooling or forcing
associated with the polar jet. A weak subtropical jet was noted across northern Mexico into the Gulf of Mexico. A strong
shortwave trough was exiting the central Rockies into the northern Rockies and northern plains during the day. A nearly
stationary front was draped from Iowa into central Kansas and then into Colorado. There was no significant shortwave
trough approaching west Texas. 500mb heights were on the rise througout the region from 12 UTC May 11 to 00 UTC May 12.
This was partly due to the exiting shortwave trough as well as to afternoon elevated heating. The 500mb winds at ELP and
ABQ at 00z might lead one to believe that a shortwave trough was approaching. But inspection of the actual upper air data
shows that the winds at 500mb at El Paso were only veered at 500mb and not immediately below and above 500mb.
So it is not clear if the due west 500mb wind at ELP is indicative of a shortwave trough of some kind of measurement error.
Also, thunderstorms were occurring around ABQ near the time of the upper air release and this likely affected the winds.
In fact, the 00 UTC 400mb chart does not show any well defined shortwave trough. The AMA and MAF soundings at 12
UTC and 00 UTC show steepening low level lapse rates due to diurnal heating. The west Texas region was located near the
middle of large 700mb warm plume. Of course this is to be expected since we are dealing with a dryline, with no surface fronts.
The cooler temperature at 700mb at Midland at 00z was due to intense surface heating, vertical mixing and low level convergence
immediately ahead of the dryline. This was despite general 700mb warming across the region. So the 700mb temp. at Midland
held steady in spite of the 700mb warming that was occurring at ABQ, AMA, ELP and DYS. Note that the moist layer on
the MAF sounding extended up to around 700mb. Modern day numerical models often show 700-500mb cooling/moistening in
the vicinity of the dryline during model convective initiation. In fact, sometimes the models don't develop precip but show a narrow
axis of cooling/moistening around 700mb. That superadiabatic layer just above 700mb is the result of the wet bulbing effect and the
data in this layer is in error.
The bottom line
is that if there was an approaching
shortwave trough then it was fairly weak. This dryline
4 pm which
is rather early in the afternoon and possibly indicative of an approaching shortwave trough. But the strong shortwave troughs
traversing the northern plains may have resulted in a further east position of the dryline in Kansas and northern Oklahoma compared to
Texas. This could be why the southern end of the dryline started to retreat before the northern end did. Typical dryline retreats
occur after 5 or 6 pm. I am sure that one could claim the existence of a shortwave trough in the southern Rockie/Plains. Sometimes
it seems that every thunderstorm that pops up is blamed on a shortwave trough. This is of course absurd as mesoscale forcing is
orders of magnitude larger than synoptic scale forcing. Mesoscale and smaller dryline features and terrain features are often
key in west Texas storm initiation in the absence of strong or even weak synoptic scale forcing.
At 22 UTC,
the southwestern end of the dryline was retreating to the northwest.
Surface dewpoints west of the dryline near
Lubbock were in the upper 30s and lower 40sF. Surface dewpoints east of the dryline were in the upper 50s to mid 60sF
depending on elevation. There could have been more than 1 convergence line or dryline across west Texas, but we will never
know since surface observations have been traditionally tied to aviation in this country and have not been positioned according to
meteorological need. Again, the strong shortwave troughs traversing the northern plains may have resulted in a further east position of the
dryline in Kansas and northern Oklahoma compared to Texas.
The dryline continued to retreat after 22 UTC
and was located in the vicinity of Lubbock by 00 UTC.
By 02 UTC
extended from just east of Hobbs to west of Reese AFB to just south of GAG. Due to the lack of surface data and since the
low levels were altered dramatically by ongoing convection in the Lubbock area, it is very difficult to assess the instability and shear.
Nevertheless, the 02 UTC observations from LBB and REE showed plenty of moisture with lower 60s F surface dewpoints.
Instability and shear undoutbedly varied tremendously across the Lubbock region due to small scale convective effects. Thunderstorms
were ongoing in the Lubbock area from 630 pm through the evening. While the storm scale and mesoscale effects are not known,
these effects could have turned an ordinary looking severe weather day into a violent tornado situation. I can make a rough estimate
of surface based instability at 01 and 02 UTC based on the surface data at LBB and surrounding RAOBS. At 01 UTC the T/TD
were 86F/56F at 897mb while at 02 UTC they were 71F/63F. Actually these observations are almost identical in terms of
theta-e. Surface based cape was around 3000 j/kg. Of course, some prefer to use a mean layer CAPE value. But I am not a
huge fan of this. The reason why I am not a fan is simple. How do we know what the vertical distribution of moisture is like
given the wide spacing of radiosonde stations? The answer is that we do not know. If we do not know, then why use it?
For convection that is ingesting low air at the lowest levels(such as supercell storms), I think that surface based cape is a
decent proxy of instability. Also, low level moisture tends to "pool" and be locally deeper in convergent areas near thunderstorm
updrafts. For low top storms, a lower level lifted index or low level CAPE would probably be better.
A timeline of the event can be found here.