|The upgraded DOW with its new extensible weather station in front of NCAR's Foothills Lab. (Photo by Carlye Calvin.)|
Built on a shoestring four years ago, Doppler on Wheels (DOW) hasn't been stingy with its data. This pack of portable weather radars has crisscrossed the country, peering inside tornadoes, hurricanes, and winter storms. This spring the DOW II and DOW III units received a major upgrade at NCAR, preparing them for field work on two continents over the next year and expanded use by university scientists.
"They don't look [much] different, but their brains and hearts are different," says Josh Wurman, an assistant professor at the University of Oklahoma. Wurman maintains the two existing DOW units with support from NSF's Major Research Infrastructure fund. (The original DOW was retired over two years ago.) NSF also paid for the recent improvements with in-kind assistance from NCAR staff, including technicians Jeff Bobka, Al Phinney, Tim Rucker, and Joe Vinson (Atmospheric Technology Division, or ATD).
Each radar is mounted behind a mini-processing room at the back of a flatbed truck. The most visible addition to DOW III is an extensible weather station to be used to verify conditions at the radar site. Mounted on a hydraulic pole between the antenna pedestal and the truck's cab, the station can be lofted 10 meters (30 feet) above ground level. "The idea is that in very high winds, the pole will stay rigid," says Wurman.
On the inside, DOWs II and III have received processor transplants. They now feature the second generation of the PC integrated radar acquisition board (PIRAQ) developed by ATD engineers Mitchell Randall and Eric Loew. PIRAQ made DOW possible by allowing radar data to be processed in real time through a personal computer rather than a costly, room-sized set of equipment. PIRAQ II has twice the computing power of the original, and "there's a pretty long list of other changes," says Randall. "When we did the first PIRAQ, we didn't realize where it would be used." The new version is compatible with coherent lidar and FM-CW radar, and it can transform a conventional weather radar into a Doppler radar.
Both DOWs have also gotten new antenna controller systems, developed by ATD's Jonathan Lutz. These will allow for faster and more precise pointing and scanning with the radar beam. Radar receivers have also been improved. Together, all of the upgrades should allow for winds to be sampled at an ultra-crisp resolution of 12.5 meters (38 feet).
Just as important to Wurman as the new features is improved reliability. "Our goal is to have it work every time," he says. The very success of the DOWs is one reason why a tune-up was needed: "Life on the road is very difficult. The radars get shaken a lot going down a lot of bad roads."
During last year's devastating tornado at Spencer, South Dakota, on 30 May, two DOW units got within 1.7 km of the twister. Then the antenna of one unit refused to spin. The DOW team ended up obtaining single- rather than dual-Doppler data, which prevented a full two-dimensional wind analysis. Still, the team obtained a peak wind of 115 m/s (258 mph) about 50 meters above the surface. According to Wurman, this was the most disastrous tornado strike for which high-resolution measurements were collected--that is, until the central Oklahoma twisters of last month (see sidebar).
These events may also help to confirm the Fujita tornado-damage scale. The scale gives numerical ratings (F0 to F5) based on observed damage. In turn, these damage ratings are connected to wind-speed ranges, but there have been few data to confirm whether the higher wind ranges are in fact accurate. The peak DOW wind from Spencer corresponds to a strong F4 rating, "which is what they found in the damage survey," says Wurman. Similarly, the F5 wind measured in Oklahoma on 3 May was backed up by F5 damage at ground level.
Wurman led the conceptualization and construction of the first DOW unit, which has since been cannibalized for its two newer siblings. To build the original DOW, he and colleagues pulled together expertise and surplused material from NCAR, OU, and the National Severe Storms Laboratory (NSSL). The first DOW made its debut in the southern plains for the 1995 field session of VORTEX (Verification of the Origins of Rotation in Tornadoes Experiment). Twisters remain at the top of the DOWs' priority list. The radars traversed the plains this spring for the second year of the field project ROTATE (Radar Observations of Thunderstorms and Tornadoes Experiment).
Later this summer, one DOW will set its sights on landfalling hurricanes. In 1996's Hurricane Fran and last year's Bonnie and Georges, DOW data verified the presence of small-scale horizontal rolls that transport upper-level winds downward. These may be the cause of localized streaks of intense damage such as those discovered after Hurricane Andrew. DOWs II and III measured gusts over 55 m/s just above the surface while parked at airports in Gulfport and Keesler Air Force Base, Mississippi, during Hurricane Georges last September. "We like airports in hurricanes," says Wurman. "They're treeless and flat, with no debris."
The other DOW will be shipped across the Atlantic in August to join the Mesoscale Alpine Programme in northern Italy and southern Switzerland for three months. (See the Winter 1998 UCAR Quarterly.) Matthias Steiner (Princeton University) will oversee the DOW's deployment in the Toce (Italy) and Tocino (Switzerland) river valleys, close to the NCAR S-Pol dual-polarization radar. Air flow in the valleys will be invisible to S-Pol, so the DOW will gather smaller-scale measurements beneath the larger radar's coverage area. "Just getting [the radar] down some of those mountain roads will be interesting," Wurman adds. "It's not Oklahoma."
Researchers are invited to contact Wurman about using one or more DOWs outside of the radars' springtime commitments for tornado research. "We're really anxious to have the science community use these," he says. Interested scientists may contact Wurman directly (303- 325-0589, email@example.com). See the DOW Web page.
Oklahoma twister yields a world-record wind
Wurman and his DOW team often drive hundreds of miles to snag a tornado, but one of the fiercest on record paid a visit to their own backyard on 3 May. The Oklahoma City area and several other points in Oklahoma and Kansas were raked by tornadoes that evening that killed more than 45 people and caused up to $1 billion in damage.
The strongest tornado of the day was a long-track F5 that plowed through portions of Oklahoma City and its suburbs. The two DOWs were positioned on either side of this twister (one unit within one mile of it) as the tornado took shape near Chickasha, about 65 km southwest of Oklahoma City, around 5:45 p.m. The DOWs tracked the storm over the next hour, measuring winds as high as 142 m/s (318 mph) a few dozen meters above ground. Should these wind speeds hold up in later analysis, they will be the highest ever measured in nature.
Also within a mile of the tornado was a separate portable Doppler radar, deployed by OU professor Howard Bluestein in conjunction with the University of Massachusetts and support from NSF. With a wavelength of 3 millimeters, this UMass radar offers limited range but very high spatial resolution, as fine as 5 x 5 x 15 m at a range of one mile. As they tracked the tornado, close to a point where most of a house was swept from its foundation, Bluestein's team saw wavelike features--possibly multiple vortices--along the edge of the tornado. "These are the highest-resolution images ever obtained of the inside of a powerful tornado," says Bluestein, who is collaborating with Andrew Pazmany from UMass. "We hope to combine our data with Josh [Wurman]'s and those from the mobile mesonet [operated by OU's Jerry Straka] to form a more complete picture of tornado structure."
DOW followed other tornadoes until after midnight, including a mile-wide twister that damaged or destroyed every building in the town of Mulhall. Afterward, the DOW team caught its breath and contemplated the vast task of sifting through observations from the portable Dopplers as well as the UMass radar, stationary radars, and other equipment that documented the tornado. "We're reeling with the volume of data we've collected," said Wurman. See preliminary DOW data and a map of the tornado's initial path with radar locations.