Tornadoes - Severe Convective Storms and Tornadoes: Observations and Dynamics - page 329

convective temperature may be reached fairly early in the day, around noon; along
coastal areas, the sea breeze front may initiate convective storms by midday; the
offshore land breeze, on the other hand, may initiate convection early in the
morning. Tornadoes in MCSs, which develop from more isolated convection late
in the afternoon, but mature after dark when the cold pool from the cells has built
up in strength and areal extent, occur during the overnight hours. In addition,
supercell tornadoes have been documented overnight when storm formation is not
tied at all to the diurnal heating cycle, but occurs, for example, when a front
finally reaches a region of enough moisture and CAPE for storm initiation.
6.3 TORNADO RESEARCH
The two main scientific problems related to tornadoes are (1) explaining their
structure, particularly their three-dimensional wind distribution, and (2) why they
form (and why at other times they do not form). The former involves an under-
standing of boundary-layer dynamics and how the tornado vortex in the
boundary layer is coupled to a vortex produced aloft, which in supercells is
usually independent of boundary-layer processes ( just the vortex aloft). The latter
involves storm-scale dynamics and how vorticity is produced and amplified on the
storm scale. The two problems may be related, however, for example, when
boundary-layer behavior results in the amplification/diminution of vorticity inde-
pendent of storm-scale processes or when changes in storm-scale flow affect the
character and behavior of the boundary layer. The first problem involves a vortex
of air rubbing against the ground and forcing from above by positive buoyancy
(or an upward-directed dynamic pressure gradient force) or, in the case of a
laboratory model, an exhaust fan. The second problem has been studied using
idealized numerical simulations, in which the practical limits of spatial resolution
and the realism of subgrid-scale parameterizations are the main issues affecting the
utility of simulations. Single and dual-Doppler radar analyses have been helpful,
as has the analysis of photographs, movies, and videos.
To date, it is not an understatement to say that whether or not a parent
convective storm will go on to produce a tornado cannot be predicted accurately.
Owing to the diculties in collecting detailed observations in the right place for a
suciently long period of time, and the diculty in numerically simulating both a
convective storm and a tornado together (simulation of tornadoes requires
extremely high spatial resolution), tornadoes are arguably the least well under-
stood of all phenomena associated with convective storms. However, there are
some instances when there is an ''outbreak'' of tornadic supercells over a broad
area, and one can therefore be confident that there should be a ''signal'' in the
environment that points to why they occur.
The first studies of tornadoes included finding exact analytic solutions to
highly idealized conditions, which have yielded considerable physical insight; these
studies were absolutely necessary before powerful computers became available.
Early research (some of the following is given in more detail in Chapter 1) on