Scientists know how tornadoes form, but they are hard to predict
![Rain obscures the view of a tornado on May 28 in Lawrence, Kan. Rain obscures the view of a tornado on May 28 in Lawrence, Kan. (Kyle Rivas/Getty Images)](https://whyy.org/wp-content/uploads/2019/05/gettyimages-1146801681_custom-9e5983483f2890692519ce293b0c99a678258afa-768x459.jpg)
Rain obscures the view of a tornado on May 28 in Lawrence, Kan. (Kyle Rivas/Getty Images)
Deadly tornadoes have been ripping through parts of the Unites States for weeks. Storms have been leaving a trail of destruction from Texas all the way up to Maryland, and on Monday, 52 tornadoes may have touched down across eight states, says the National Oceanic and Atmospheric Administration.
Patrick Marsh, a meteorologist at the NOAA’s Storm Prediction Center, says it’s unusual to have this kind of sustained tornado activity.
“We’ve had long stretches where we’ve had tornadoes over a long period of time, but the difference was we’d have a day or two here or there where we kind of had a reprieve. We’re not seeing the reprieve this time, and that’s what makes this outbreak so unique,” he tells All Things Considered host Ari Shapiro.
Scientists know how the storms are created, but, he says, it’s nearly impossible to predict where a tornado will touch down — and they don’t have enough data to attribute the recent outbreaks to climate change.
“I would love to be able to tell somebody, ‘You know, tomorrow there’s going to be a tornado that’s going to go through downtown Oklahoma City.’ But the atmosphere is inherently chaotic, and I don’t know if we’ll ever be able to get there,” he says.
Interview Highlights
On why climate change’s role in tornadoes is murky
Even though the vast majority of the world’s tornadoes occur in the United States, it’s still somewhere on the order of about 1,200 tornadoes a year. So when you think about how much land there is in the United States, that’s not a lot of tornadoes for us to observe and predict. And so our dataset for tornadoes is actually quite limited. For us to be able to do attribution studies to assess whether or not changes we see are related to climate change or other factors, we need a much bigger tornado dataset or we need better statistics to assess this. Hopefully in the next few years we’ll be able to say something more definitive, but at this time we just don’t have the tools to do so.
On improving tornado predictions
Some of the tools that we’re developing is a process called Warn On Forecast, an idea that we can run high-resolution numerical simulations out several hours in advance and be able to tell people, “There’s a 20% chance of a tornado moving within a few miles of downtown Oklahoma City in the next two hours.” And hopefully this allows things like hospitals to take safety precautions that they might not be able to implement if they only have a few minutes, which is what the current paradigm is.
For example, you could cut down on elective surgeries so you don’t run the risk of somebody being in surgery as a tornado hits. You could move patients out of their rooms and into hallways at a much slower pace rather than the frenetic pace [when there’s] a tornado bearing down on you. And that will also cut down on injuries and the risk for additional injuries.
On advice to those who live in tornado-prone areas
In a nutshell, what you can do for tornadoes is you want to get into a well-built structure, you want to get to the lowest floor — and this does not necessarily mean below ground — and you want to put as many walls between you and the outside: get in, get down and cover up.
Leslie Ovalle and Sarah Handel produced and edited this story for broadcast. Heidi Glenn adapted it for the Web.
![](https://www.google-analytics.com/__utm.gif?utmac=UA-5828686-4&utmdt=Scientists+Know+How+Tornadoes+Form%2C+But+They+Are+Hard+To+Predict&utme=8(APIKey)9(MDAzMzI1ODY3MDEyMzkzOTE3NjIxNDg3MQ001))