Texas Tech’s Chris Weiss was among a team of researchers collecting data about tornadic thunderstorms.
For the better part of a month, Chris Weiss cast his eyes toward spring southern plains skies percolating with regular foreboding. Day after day, the clouds formed and brought with them an unpredictable array of meteorological threats, including fierce rain, damaging hail and deadly tornadoes.
“I have been part of several projects where we try to understand how tornadoes form out of severe thunderstorms,” he said, “because we know a good chunk of violent tornadoes occur from what we call supercell thunderstorms.”
Weiss, a professor of atmospheric science in Texas Tech University's Department of Geosciences, joined a team of researchers in the field between late May and mid-June collecting data and trying to connect feisty weather dots with two ultimate goals in mind: give people in the path of storms more lead time and ensure greater overall accuracy when warnings are issued.
The Targeted Observation by Radars and Unmanned Aircraft Systems of Supercells (TORUS) project is part of the work done through the National Wind Institute, and it also included representatives from the National Severe Storms Laboratory in Norman, Oklahoma, as well as researchers from the University of Oklahoma, the University of Nebraska and the University of Colorado.
“We are trying to get at clues that will actually tell us what will make a storm produce or fail to produce a tornado,” Weiss said. “Ultimately, that will improve the quality of the warnings, eliminate a lot of the false alarms and give people more time to take action.”
The project, funded by a National Science Foundation grant, began as a two-year initiative in 2019, but was sidelined by the COVID-19 pandemic. Work resumed last year, and the team received an extra year to work on wrapping up loose ends.
“We have found projects like this work really well because everyone has their own pieces they bring to the table,” Weiss said. “We get these nice collaborative data sets as a result. Some of us are recording temperature, humidity, barometric pressure. Some have remote measurements like radar, and we're able to combine all of this to produce a four-dimensional picture of a storm.”
Their time studying storms produced useful information that should inform meteorological practices going forward.
“Anyone who has experienced a thunderstorm, especially in West Texas, usually will feel that cold rush of wind even before the actual rainfall,” he said. “We call that the outflow from the storm, and we think cold air has a lot to do with why tornadoes form. We still have a lot more to do to better understand that connection.
“Our work is aimed toward the warning side in terms of what we can do to help operational meteorologists like those at the National Weather Service anticipate a tornado will happen. Usually, our lead time on a tornado warning is around 15 minutes, but if you could expand that to 30 minutes on average or 45, you could really help people because the earlier folks get the warning, the better able they are to take proper precautions.”
Researchers were in the field beginning May 23 and operated for a dozen days while collecting data on a number of weather-related events. The group's last day was June 15, the day after a tornado struck the Panhandle community of Perryton, and six days before the June 21 tornado devastated the town of Matador, about 90 miles east of Lubbock.
They spent almost all their time in and around the Panhandle, observing storms during an unusually rainy and active stretch.
“It ended up being quite a season for the books,” Weiss said as the field time came to an end. “In the many years I have been doing this work, I cannot recall ever failing to cross Interstate 70 (or even coming close) in a storm season. We spent almost the entirety of the project in or near the Texas Panhandle during a jaw-dropping period of repeated regional rainfall.”
Among the storms they chased were tornadic supercells in eastern New Mexico, on the South Plains and across western Oklahoma. Other storms throughout the area did not spawn tornadoes while several others were observed after a tornado developed.
“Typically, we would cover the whole plains area from the Dakotas down through Texas,” he said, “but because of what happened regionally, we never really got too far north of I-40, so we got a lot of our data closer to Texas. It was very much a southern plains type of year.”
That included the especially active weather day that saw the tornado strike Perryton, killing three people and injuring more than 100.
“We were in the field that day, observing a storm that was just south of the one that produced the Perryton tornado,” Weiss said. “We were unable to get into a safe position for the Perryton storm, though, but we did gather good data on the one that was south.”
The group's work will make a difference, especially in this part of the country where tornadoes are regular occurrences, and significant weather events are expected to increase in frequency.
“If you live in Perryton or Matador, you probably understand why this research is important,” Weiss said. “Each of those storms were a little different. The Matador storm was wrapped in rain, and it just wasn't very visible. The tornado was more difficult to anticipate visually than the Perryton storm, but there were other problems there like the power going out before sirens could be activated.”
It also comes down to time spent observing and learning about those threatening skies.
“We want to help reduce warning fatigue,” he said. “You want warnings people will heed, but the false alarm rate on tornado warnings is pretty high, which is an unavoidable aspect of the process. We typically warn on detection of storm scale rotation because we can't detect tornadoes of typical size and range with our operational radars.
“We have to assume the storm that is spinning significantly will go tornadic, but it's not a clean relationship with the tornado. Expanding lead time and reducing false alarms, those are the two things this research is trying to get at.”
