NORMAN — In severe weather events, many who live in Oklahoma’s Tornado Alley are used to seeing storms develop live on television via cameras in helicopters or on the dashboards of storm spotter vehicles.

The University of Oklahoma, AT&T and The Weather Channel have taken this idea and supercharged it, installing live cameras on research vehicles and instruments that will, eventually, send radar data back in near-real time.

The collaboration was announced Friday at a press conference inside the National Weather Center in Norman, only hours after a severe storm had passed.

Currently, two Advanced Radar Research Center (ARRC) trucks are equipped with live video cameras that will be streamed via The Weather Channel. The plan is to equip six total.

“We’re going to put it on everything we can,” said Berrien Moore, director of the National Weather Center and dean of the OU College of Atmospheric and Geographic Sciences. “We’re simply exploiting the 21st century. We know this works.”

Moore said AT&T’s broadband capabilities make it possible for vehicles to not only stream live video for public viewing, but also for valuable storm data to be quickly sent back to forecasters for seamless use in weather models. Normally, this data must be brought back to the National Weather Center so it can be absorbed and analyzed, and that makes multi-day ARRC deployments problematic.

“Now, that data will come back, in real time, feeding our models,” Moore told the Norman, Oklahoma Transcript.

The newly equipped trucks went live on The Weather Channel for the first time Friday, though the live video has gone through months of testing. Data streaming will be available in the immediate future as equipment continues to be installed.

“That’s our ultimate goal,” said Addison Alford, a meteorology Ph.D. student who helped install some of the equipment. “We’d like to provide it to the National Weather Service, and AT&T has given us that opportunity.”

Video is a helpful tool for both the public and forecasters. Doppler radar and satellite data is limited in what it can show happening on the ground, and live video can help make up for that.

“Especially here in Oklahoma, (this is vital), not only in terms of modeling but also getting the word out if a tornado’s about to drop,” said Shawn Riley, National Weather Center IT specialist. “Getting the word out about the dangers as soon as possible, that will only help the situation.”

Steven E. Koch, Ph.D., director of the National Severe Storms Laboratory, said the new installments on ARRC vehicles also will help as meteorologists attempt to improve tornado lead time. Koch said the ultimate goal is to provide a warning that is 60 minutes ahead of time.

“We’re not there, and radar has not proven the ability to see 60 minutes ahead,” he said. “We’re going to have to take a different approach than just using radar alone. To me, it’ll be useful to share publicly what clouds are doing, in addition to radar.”

It also will help to have more mobile radar vehicles sending real time data back at lower levels in the storm. Stationary doppler radar struggles to show what is happening in the lowest level of a storm because after the signal travels about 12 miles, the curvature of the earth becomes an issue. Geographical obstacles, such as a mountain range, limit the amount of data that can be gathered.

Mobile radar, like those on ARRC vehicles, can get closer, move with the storm and read lower levels of rotation. It also can pick up tornado debris signatures, signaling that a tornado is on the ground, even if it is not visible at night or if it is wrapped in rain.

Finally, the data obtained at lower levels can feed models and create a more accurate prediction. For example, researchers are interested in the role updrafts play in storms that produce tornadoes, but those are hard to examine with current methods.

Using the newly equipped ARRC fleet to feed that information back in near-real time, forecasters can then use models that provide a more accurate prediction based on the way storms are forming.

“If we can use a new piece of information to say there is a strong updraft, we can feed that into the model and update it every five minutes. Models tend to retain that information and spin out data about how those storms are going to evolve,” Koch said.

Troxtell writes for the Norman, Oklahoma Transcript.