Geoscience Reference
In-Depth Information
eddies are tilted (with respect to the radial direction) so as to convert eddy angular
momentum into mean angular momentum, as happens, for example, in some
synoptic-scale waves in baroclinic westerlies. Victor Starr at MIT in 1968 termed
such a process ''negative viscosity'' and Doug Lilly suggested it could be respons-
ible for affecting the intensity of some tornadoes.
A process similar to that of negative visocity might be the conglomeration of
pre-existing, smaller scale vortices into a larger one: the author and his co-workers
found evidence of tornado formation when smaller vortices along a gust front
seemed to interact to produce a larger scale, tornadic vortex; similar behavior has
been noted in some numerical simulations.
6.5.7 Two-celled mesocyclones and shear instabilities
Roger Wakimoto and his student Chinghwang Liu have suggested that some
tornadoes may be initiated when an occlusion/RFD downdraft forms in a
mesocyclone (as a result of vorticity becoming less cyclonic with height or precipi-
tation loading or evaporative cooling), leading to an annulus of strong shear and
resulting barotropically unstable air within it, which then breaks down into multi-
ple vortices, each of which could become a tornado. This process is like the
formation of multiple vortices in a tornado, but on the scale of the mesocyclone.
(It is important to realize that vortex breakdown, to be discussed later, cannot
occur in mesocyclones because a supercritical end-wall vortex does not form.) It is
the descent of air in the middle of the vortex that is responsible for producing the
barotropically unstable shear zone. However, it is not clear how each unstable
vortex can be stretched up to tornado intensity unless each vortex is co-located
with a region of convergence under the updraft.
6.5.8 Cyclic tornadogenesis
Just as mesocyclones sometimes form and decay in a cyclical fashion ( Figure 4.59 ),
so do tornadoes in a process Don Burgess at NSSL has referred to as ''cyclic
tornadogenesis''. David Dowell and the author, using airborne Doppler radar
data from VORTEX in 1995, further refined the conceptual model of cyclic
tornadogenesis proposed by Don Burgess et al. in the early 1980s ( Figure 6.37 ).
An incipient vortex forms along the rear-flank gust front and propagates along
the horseshoe-shaped updraft associated with flanking line towers, until it reaches
the tip of the end of the horseshoe-shaped updraft, near the RFD. When it
reaches this location of strong horizontal gradient in vertical motion, the tornado
is mature. The tornado then propagates away from the updraft entirely and dis-
sipates in the downdraft region. A new tornado may then form along the bulge in
the rear-flank gust front, and the process of tornadogenesis is repeated. In some
rare instances, a tornado may become locked into such a position that it does not
propagate away from the tip of the updraft region and a long-lived (and possibly
very intense) tornado may result. It is not known why in these rare instances the
tornado remains locked into its mature phase, but observational evidence suggests
Search WWH ::




Custom Search