Tornadoes - Severe Convective Storms and Tornadoes: Observations and Dynamics - page 386

2 , which could be in excess of 50m s 1 , an extremely
unlikely occurrence. Robin Tanamachi and collaborators have shown from digital
infrared imagery that the lapse rate of temperature on the surface of a tornado
condensation funnel is, as would be expected, moist adiabatic, but we do not
know what the lapse rate is inside the condensation funnel. Extreme warming has
not been observed at the center of tornadoes, though there have been anecdotal
accounts of fires going on near tornadoes, which more likely might have been a
consequence of the destruction of structures by tornadic winds. Stirling Colgate,
in the early 1980s, attempted unsuccessfully to obtain measurements in tornadoes
using rockets with sensors launched from an aircraft.
Despite in situ measurements near the ground of temperature in the center of
tornadoes being rare (nonexistent aloft), we cannot however totally rule out the
possibility that subsidence warming might sometimes increase the intensity of
tornadoes. In addition to the great amount of work that would be needed to
bring air parcels from high up down to low altitude, lateral mixing (entrainment
of environmental air) would reduce buoyancy and thus reduce the effects of sub-
sidence warming. Moreover, the time it takes a tornado vortex to develop solid
body rotation may be short compared with the time it takes environmental air to
descend substantially, even if it did.
When a non-rotating updraft penetrates above the tropopause, a hydrostatic
''cold dome'' is produced because air parcels become colder than their environ-
ment just above the tropopause. Owing to the cold air aloft, there cannot be a
hydrostatic pressure deficit underneath the updraft at the ground because CIN
above the tropopause should be approximately equal and opposite in sign to
CAPE in the troposphere.
A rotating updraft in a tornado, on the other hand, must be accompanied
by a pressure deficit at the center. Above the tropopause there is divergence, so
that vorticity in the updraft decreases. Because vorticity decreases with height,
there should be a downward-directed perturbation pressure gradient force and
therefore a dynamically driven downdraft, which may appear as a ''crater'' in the
cloud top. So, we are led to the conclusion that a depression in the height of the
anvil region above the updraft in a buoyant cloud that drives a ''deep'' tornado
may belie the hidden updraft below. There have been some observations of collap-
sing tops in thunderstorm anvils near the time of tornadoes, especially by Ted
Fujita. So, while the dynamics of tornadoes depends mostly on what happens near
the surface, a look from above via satellite imagery or aircraft flying above a
storm may have merit.
Doppler radar estimates of the wind speeds in tornadoes frequently exceed the
thermodynamic speed limit (6.57) by a substantial margin, though there are some
significant uncertainties regarding the representativeness of radar data and nearby
thermodynamic data. We would not expect the thermodynamic speed limit to be
realistic, owing to parcel accelerations experienced in the corner region: a tornado
is grossly non-hydrostatic. So why even consider the hydrostatic speed limit? The
main justification is that it can be thought of as a benchmark. We will explain
why shortly.

1 =

have to be ð 2 CAPE Þ