From Texas to South Dakota, the great plains of the central U.S. are home to countless tornadoes each year. The worst of these storms can level entire towns, leaving nothing but concrete foundations in their wake. But despite their devastating power, tornadoes are not well understood, says geophysicist Roger Wakimoto from the National Center for Atmospheric Research (NCAR). “It’s embarrassing,” he says, “but we still do not understand what triggers tornado genesis.”

A “Doppler on Wheels” radar dish scans an incoming storm system in eastern Colorado as part of the 2010 VORTEX 2 expediton.

That’s why Wakimoto and a team of experts from around the country organized VORTEX2, one of the largest field studies of tornado weather ever attempted.   With more than 140 scientists and 35 research vehicles, it’s a mobile armada, equipped with doppler radar trucks, car-mounted weather stations, and dozens of deployable instruments that can be placed in a twister’s path.

I recently had the opportunity to follow the VORTEX2 team during the final days of their six-week expedition. When we met up in a tiny crossroads town in eastern Colorado, they had already been on the road for five weeks, rarely sleeping in the same town for more than one night. It was a grueling schedule. But Wakimoto says that staying mobile is necessary in order to see potential tornadoes as they form.

To narrow down the chase, the VORTEX2 team used every data source in its arsenal–from satellite images to weather balloons–to take a broad view of the area. “We have 50 different instrument platforms out here collecting data,” said David Dowell, a severe weather researcher from NCAR. “You need all the pieces to get the big picture of what’s happening out in these storms.”

Dowell manned the nerve center of the entire operation, a white ambulence-like truck with three computer stations inside. Once his team identified a promising storm, he helped coordinate all 50 vehicles as they jumped into action, drove to an area in its path, and surrounded it with weather instruments. If they were lucky, the storm cranked out a tornado–if not, they were left with just wind and rain.

By the end of their expedition in mid-June, the VORTEX2 team had driven more than 16,000 miles, and was able to measure only a handful of tornadoes in all. But Roger Wakimoto says that this still left them with a wealth of information. Although sorting through the data might take years, he thinks it could help shed light on the forces behind tornado formation, and lead to more accurate tornado warnings in the future.

You can see photos from the three days I spent on the road with the VORTEX2 team in the slideshow below, or learn more about tornado science on NOVA’s “Hunt for the Supertwister” website.