The imprecise science of chasing twisters sometimes requires a convenient twist of fate. Such was the case on May 4 this year when a team of tornado chasers, electrical engineering students from UMass Amherst and meteorology students from the University of Oklahoma (OU), was engaged in the very unscientific task of waiting for a tire to get patched at a gas station near Protection, Kansas.

If not for the blowout, they might have been in Oklahoma by the time they instead came face-to-face with their meteorological holy grail: a monstrous thunderhead moving in from the south, looking like it had just escaped from The Wizard of Oz.

Later that night, while UMass Amherst chasers were tracking this storm system from 20 kilometers away with specialized radar built here on campus, the supercell spat out the country’s second EF5 tornado (the largest magnitude) in the last decade. (UMass Amherst tornado chasers and their truck-mounted radar tracked the other EF5 cone to touch down this decade, the brutish tornado of 1999 that hit Oklahoma City.) At about 10 p.m., the killer tornado slammed into Greensburg like a mile-high sledgehammer made from winds powerful enough to drive straw into telephone poles. The collected data from it will help students and researchers in their quest to unlock “tornadogenesis,” or the formation and life cycles of deadly twisters.

“As we monitor the life cycle and environment of a storm, this information can then be analyzed and even ingested into a numerical weather prediction model,” explains Stephen Frasier of the UMass Amherst Electrical and Computer Engineering Department and director of its Microwave Remote Sensing Laboratory (MIRSL). “The observations we’re making will lead to improvements in these models to make them more and more reliable. They may ultimately be used to forecast whether a tornado will develop from a supercell.”

One result of their research would be saving lives by giving earlier and more accurate tornado warnings. In the case of the Greensburg tornado, the National Weather Service (NWS) was able to issue a relatively long tornado alert some 20 minutes before the storm touched down. Such lengthy alerts aren’t always possible with stationary radar systems, which are good at spotting parent storms where they form above 1.5 kilometers, but not as good at monitoring where tornadoes actually fall to earth. That’s one reason why the false-alarm rate for tornado alerts approaches 80 percent. “People can become jaded after hearing too many false alarms,” Frasier says, and may not heed them, which can be a deadly decision.

With support from the National Science Foundation, the chase team, led by Howard Bluestein of the OU School of Meteorology and including students from both UMass Amherst and OU, has been in the field each tornado season since 1993. The team has deployed different types of truck-mounted Doppler radars over the years. It now uses X-band radar, which is less disrupted by rain and uses polarization, similar to that found in polarized sunglasses, to provide more information on the details of the storm.

“With polarization,” says Frasier, “we can identify when a tornado has touched down, where the debris cloud is located, and where the rain is located, to get a better sense of the structure, or radar signature, of the tornado. We’re accumulating a database which enhances our understanding of tornadogenesis.”

Rain or shine, UMass Amherst students expect to be in the field again next tornado season tracking every wind shear, vortex, cyclone, whirlwind, and twister that leaves its distinctive signature scribbled in the beams of their millimeter-wave radar.