Tornadoes - Severe Convective Storms and Tornadoes: Observations and Dynamics - page 331

and at the University of Oklahoma/NSSL. Vortex chamber experiments yielded
the first quantitative measurements of vortex characteristics independent of radar
and computer technology. Numerical simulations of convective storms on the
storm scale, with nested grids used to simulate substorm vortices, were also first
carried out then. Airborne Doppler radars were first used to probe severe convec-
tive storms in the early 1990s and especially during VORTEX in 1994 and 1995.
Radars mounted on aircraft allowed storms to be followed and documented, with
300m spatial resolution, for longer time durations, but the time between aircraft
passes was 5min and features near the ground could not be detected very well,
owing to ground clutter contamination. Mobile, ground-based radars mounted in
vans were first used in the late 1980s; while it was dicult to follow storms as well
as in an aircraft, data in tornadoes near the ground could be obtained with even
higher spatial resolution and at much shorter time intervals and much more
frequently than possible at a fixed site radar. Mobile Doppler radar observations
have been vital in obtaining data from real tornadoes, but they are dicult to
obtain, have limited spatial resolution, and suffer from ground clutter contamina-
tion at low levels. High-frequency radars require only modest-size antennas, but
are very susceptible to attenuation; low-frequency radars are much less susceptible
to attenuation, but require larger antennas to yield the same high azimuthal
resolution and are therefore too large for mobile work. Large-eddy simulations
(LES) of tornadoes (vortices interacting with the ground and isolated from their
parent storms) were first carried out in the late 1990s using grid spacing as short
as 1-3m in some places, so that the turbulent aspects of tornadoes could be
better represented. LES models have yielded interesting measurements, but in the
absence of interaction with a parent storm. Studying tornadoes using just an LES
model of the vortex is like studying how a finger works in the absence of the arm,
the brain, etc. On the other hand, studying tornadoes using a model simulating the
entire parent storm is inadequate because boundary-layer processes cannot be
simulated faithfully without higher spatial resolution.
The state of the art for tornado research at the time of this writing in 2011
and 2012, just a year or so after VORTEX2, involves observing tornadoes with
ground-based, rapid-scan (both electronic and mechanical) mobile, Doppler
radars; mobile, rapid-scan (mechanically scanning—not electronically scanning),
polarimetric Doppler radar; mobile, pulsed Doppler lidar, instrumented surface
probes; UAVs; and three-dimensional numerical simulation experiments with grid
spacing down to tens of meters in non-hydrostatic cloud models and 1m LES
models.
6.4 TYPES OF TORNADOES AND TORNADO-LIKE VORTICES
Tornadoes have been classified according to whether they are associated with a
pre-existing, larger-scale circulation that existed before the parent storm harboring
the tornado had formed and the vorticity associated with that circulation then
became the source of vorticity for the tornado, or whether the tornadoes are asso-
ciated with a mesocyclone produced in a supercell. The largest and most intense,