Mesoscale convective systems - Severe Convective Storms and Tornadoes: Observations and Dynamics - page 293

type seen in Figure 5.22 (bottom panel) is called, more specifically, an ''elevated
rear-inflow jet''.
When the rate of generation of horizontal vorticity in the convective system is
greater at the rear edge of the surface cold pool than it is above at the rear edge of
the cloud, the rear-inflow jet descends as it passes through the convective system
( Figure 5.22, top panel), owing to the overwhelming effect of horizontal vorticity
associated with the rear, cold-pool edge. In this case, this feature is called a ''de-
scending rear-inflow jet'', and horizontal vorticity just below it, which is opposite
in sign to that of baroclinically generated horizontal vorticity at the leading edge
of the cold pool, is not advected all the way to the leading edge as it is by the
elevated rear-inflow jet. An important consequence of the descent of the rear-
inflow jet is that the rate at which horizontal vorticity is generated baroclinically
at the leading edge of the cold pool is not balanced by the advection of horizontal
vorticity (into it), so that the MCS is not as long lived as it would have been if the
rear-inflow jet were elevated (and the ascending branch of inflow would not have
been less intense).
In summary, the main factor determining whether or not a rear-inflow jet is
elevated or descending (i.e., whether or not the convective system can be long lived
or not) depends on the relative horizontal buoyancy gradients at the rear edge
associated with the warm cloud above and the cold pool below. Morris Weisman
found in numerical simulations that, in general, when CAPE is ''low to medium''
and low-level vertical shear is ''weak to moderate'' the rear-inflow jet descends to
the surface behind the leading edge of the gust front. When CAPE is ''high'' and
vertical shear ''strong'', the rear-inflow jet is elevated. Further refinement of
estimates of how rapidly horizontal vorticity is generated at the leading edge of
the cold pool and especially at the rear edge of the cold pool depends on
cloud microphysics and the consequent melting and release of latent heat of con-
densation and fusion and on water loading, which reduces buoyancy inside clouds.
5.4 THE PRODUCTION OF VORTICES IN MCSs
In the preceding discussions of the dynamics of MCSs, the two-dimensional
aspects of squall line MCSs were emphasized. It has been found observationally
that many MCSs are fully three dimensional (but ''quasi-linear'' in shape) and
several types of vortices can occur at low or mid-levels. For example, Ted Fujita
in the late 1970s identified and named the ''bow echo'' ( Figure 5.23 ), in which a
40-100 km long convective line segment bulges outward and is associated with
damaging straight line winds at the surface. In addition, bow echoes sometimes
produce counter-rotating vortices at either end of the line at 2-3 km AGL: in
the Northern Hemisphere, an anticyclonic (cyclonic) vortex is produced on the
right (left) side of the end of the line with respect to the mean vertical shear
vector. Morris Weisman in the early 1990s named these features ''bookend vor-
tices'' ( Figure 5.24 ) (also referred to as ''line end vortices''). Not only may bow