Archive of Spotlight Feature Articles

New findings on the origins of tornadoes from VORTEX

By Erik Rasmussen

For a quarter of a century, researchers at NSSL and their colleagues have been working to unravel the mysteries of tornado formation. In 1994 and 1995, NSSL hosted a major field experiment called the Verification of the Origins of Rotation in Tornadoes EXperiment (VORTEX). This field experiment produced a number of high-quality data sets of tornadic and non-tornadic supercell thunderstorms, including airborne Doppler radar data, mobile mesonet data, special soundings, mobile Doppler radar data, and other conventional data sets. Several studies have been published utilizing the airborne Doppler data, but the difficult work of integrating all of these new, unique data sets into coherent pictures of storm structure and dynamics is ongoing. This integrated research approach is the one being utilized by Erik Rasmussen and David Blanchard of NSSL, along with colleagues Jerry Straka and Paul Markowski of the University of Oklahoma.

Thus far, a few important new findings have been made. We have found that boundaries, which are the leading edges of pools of cooler air left behind by thunderstorms, are prime locations for later tornado formation. Evidence suggests that the temperature contrast along these small-scale "fronts" supplies the air with horizontal rotation like a rolling pin. Then, when a mature storm moves across a boundary, the rotation is tilted upward into the storm's updraft so that the spin has the orientation of a top, while at the same time being stretched and intensified. This process imparts strong rotation to the lower levels of the storm updraft, which seems to be a necessary, but not sufficient, condition for tornado formation.

Tornado formation itself seems to be strongly linked to the character and behavior of a downdraft at the back side of the supercell storm, recognized for many years as the "rear-flank downdraft." In tornadic supercells observed in VORTEX, this downdraft straddles two regions of opposite rotation: the developing mesocyclone, with its cyclonic, or counter-clockwise spin, and a region of anticyclonic, or clockwise spin, that spirals around the outside of the downdraft. As this downdraft develops, it carves its way into the main storm updraft in the shape of a hook. In fact, this downdraft often contains enough rain to produce the hook-shaped echo seen on radar reflectivity displays. In effect, the downdraft draws rotation downward from aloft, while at the same time focusing it toward a common center. Once the rotation is focused enough, it becomes strong enough to develop a funnel cloud and raise dirt and debris at the ground, becoming a tornado.

Most of the above process occurs in a very small area, perhaps a couple of miles across. This was not known during VORTEX, when field teams were deployed across a large region of each storm. In subsequent, small focused field efforts, field teams supported by NSSL and the University of Oklahoma through the NSF have been attempting to operate in this small region, mainly in the hook echo and inside it. This region has been called the "bear's cage" by storm chasers for many years-- even experienced storm chasers would rather not be there! But with knowledge gained through VORTEX, and with an extra degree of caution, field teams have gathered data in this region in many more supercells. The goal of this ongoing work is to determine what sorts of rear-flank downdrafts are supportive of tornado formation as opposed to those that actually work to hinder or prevent tornado formation.

VORTEX has produced a number of troubling new findings. For example, it appears that perhaps many fewer supercells and mesocyclones produce tornadoes than scientists originally believed. At one time, researchers felt that tornadoes somehow were caused directly by mesocyclones, and that perhaps one-half of all mesocyclones produced tornadoes. We now know that this is not the case, and that tornado formation is a complicated process that depends perhaps only indirectly on the presence of a mesocyclone. Further, we have learned that the difference between tornadic and non-tornadic mesocyclones can be very, very subtle. We are examining a case in which a storm shows all indications of being tornadic on WSR-88D, and in fact in mobile Doppler radar data it has a vortex with a hook and an "eye" in the hook... indicative of very strong rotation and the centrifuging of raindrops, within about 100 m of the ground. This pattern of reflectivity and velocity is in most respects the same as observed in the tornadic supercells. Yet, no tornado formed.

In the near future, VORTEX follow-on experiments will continue to focus on the subtle differences between tornadic and non-tornadic supercells. Increasing attention will be paid to the degree of buoyancy of the rear-flank downdraft; is it warm, so that it can readily rise when ingested into the tornado, or is it cold so that it spreads away from the storm; and the tornado, if it can form, expends much energy lifting the dense air? To explore this, novel new data-gathering techniques must be developed. Don't be too surprised to see NSSL scientists launching small rockets through these downdrafts to measure their temperature structure in the future!

CLIMATE · OCEANS, GREAT LAKES, and COASTS · WEATHER and AIR QUALITY
ABOUT US · RESEARCH PROGRAMS · EDUCATION · HOME

Contact Us
Privacy Policy