A tornado expert explains why last week’s twisters were so devastating

Meteorologists were shocked by the tornadoes that devastated the Midwest and Southern US over the weekend. The twisters, which struck during the evening of December 10th, plowed across multiple states with incredible ferocity much later in the year than most tornadoes usually hit. Entire communities were devastated, and at least 90 people lost their lives. Residents across Kentucky, Arkansas, Missouri, Tennessee, Mississippi, Illinois are still recovering and searching for missing loved ones.

Tornadoes are still somewhat of an enigma to scientists. We know the basic meteorological ingredients needed to cause a tornado to form, as well as where and when they tend to appear. But what’s normal for the US — which sees more tornadoes than any other country on Earth — could be changing, according to some early research. Yet the relationship between climate change and tornadoes is still hazy. Scientists still need a better understanding of what is the perfect storm of conditions for triggering tornadoes so that they can suss out how things could change in a warming world.

The Verge spoke with John T. Allen, an associate professor of meteorology with a focus on tornadoes at Central Michigan University. He tells us what scientists are still trying to learn about tornadoes, and what we might expect in the future.

This interview has been lightly edited for length and clarity.

As someone who’s studied tornadoes for a long time, what was running through your head as you were watching what happened over the weekend?

I’ll be honest, it’s never a pleasant thought. Having seen a few of these high end tornadoes in person and seeing the devastation that they leave behind, there’s a sense of powerlessness that comes from… you sort of know that it’s going to happen, but you can do very little about it.

In this particular case, part of you is just hoping that something changes in the storm: that it stops doing this and doesn’t hit a town or it misses a town to the south, or it misses that populated area. You really don’t want to see this outcome.

What made the tornadoes that struck over the weekend unusual?

What I think has made this event stand out was a couple of particularly long-track tornadoes that stretch from Arkansas into Kentucky. One particular storm produced an extremely long-track tornado. We don’t know exactly how long that track was, but it will probably find itself in the upper echelon of the longest track tornadoes we have in history. Another aspect of that tornado was that it was particularly intense. We saw quite a number of fatalities associated with this tornado, had a number of towns absolutely devastated by it. So certainly, it will be one of those historic events that goes down in people’s memories and will be something that will be felt by those communities for a very long time.

This tornado outbreak was certainly on the larger scale for December tornado outbreaks. We have had tornadoes previously in December, even on Christmas day you can have tornado outbreak if the conditions are right. So at least in terms of getting tornadoes in December, that’s not too surprising. The sheer number of tornadoes wasn’t overly large. We’re talking on the order of somewhere around about 30 tornadoes overall, we won’t know until the survey’s finished in the coming days.

Was there anything else that might have made this tornado outbreak especially dangerous, like its path or the fact that it happened at night when it’s difficult for people to see them coming?

This area is relatively highly populated, and that’s never a good thing when it comes to tornadoes. Out in the plains, the population is relatively sparse. It means that there are fewer places that are likely to be impacted. But when you’ve got a very long track like this, the chance of hitting something goes up. As you move into the southeastern United States, it increases exponentially. [Editor’s note: some research has also found a trend toward more tornadoes in parts of the Midwest and Southeast and fewer in the Great Plains, where they’ve been more common in the past.]

Adding to that, we’ve got a tornado that’s late at night when people aren’t necessarily expecting it. That we’re not in what people typically think of as tornado season means that the overall vulnerability and chance that people may get caught unawares was much higher. Nocturnal tornadoes certainly have a disproportionate share of fatalities.

Questions are already being raised around what can or should be done to prepare for tornadoes in the future. What do we know so far about how climate change might affect tornadoes in the US?

It’s very difficult to say specifically for tornadoes. They’re very small-scale phenomena. And even on a daily basis, it can be quite difficult to predict whether a storm is going to produce a tornado or not. A lot of the research that has been done to date looks at the sort of ingredients, the conditions that could lead to tornadoes: warm, humid air, changing wind speed with height, which allows storm organization. And what we’ve found is that, at least for the future, there is definitely an increasing sort of length to the season. We’re seeing more potential for these tornadoes. The storms that produce tornadoes in the fall and winter and early spring are increasing up to 25 percent per degree Celsius of warming relative to now.

We had an abnormally warm period — I mean, Memphis hit 80 degrees Fahrenheit on Friday, which was unusually warm for the winter. Storms rely on relatively warm, moist air near the surface to increase the available energy to basically produce strong updraft. So in this case, you get much stronger vertical motion in the storms and that tends to be a thing that we see favoring stronger storms which might produce tornadoes.

The other factor that comes in there, and one that doesn’t really change is wind shear. So if you think about clouds moving in different directions at different heights, that’s what wind shear is. What wind shear does is it allows the storms to sort of organize themselves into something that is more sustained. It lasts a lot longer. It’s not your typical everyday thunderstorm. This is a storm that, in this particular case, might have persisted for over seven hours. And those sorts of storms are known to be more frequent producers of tornadoes.

And how does a tornado form?

There are a lot of different mechanisms that we’re aware of that potentially could produce a tornado. Generally speaking, you need a supercell thunderstorm. That supercell thunderstorm begins to rotate at the mid levels and that process produces a downdraft that then causes rotation at the surface to become more intense, and eventually we refer to it as a tornado when it exceeds 50 miles per hour. But the exact mechanisms of tornado formation are still an open topic of scientific research.

In this case, we had multiple storms producing very strong and intense tornadoes. Usually about six or more is defined as an outbreak, although other metrics exist. Tornado outbreaks are associated with a large-scale synoptic system. This particular system ended up producing snowfall in the Dakotas. It produced strong winds through Michigan, about 60 miles an hour. That particular system, as it moved eastward, pulled warm moist air up from the Gulf of Mexico, which was running a little bit warmer than normal. That warm, moist air is juxtaposed with relatively cool air coming behind the system. The interaction of those and an upper-level system that moved in as well produces a large area over which the conditions are favorable. It pulls those ingredients together. And so in this particular case, we had a system that produced an unusually large area of favorable conditions, which is what allowed for long-track tornadoes.

How do meteorologists forecast tornadoes? And how much warning do we typically have?

Warning times on average are in excess of 13 minutes. In this case, we had in excess of 20 minutes for Mayfield, one of the worst impacted locations. So there was certainly good anticipation of this particular event.

In terms of forecasting the likelihood of tornadoes, it really comes back to those ingredient conditions which we talked about earlier.

We have a whole suite of models we operate at our storm prediction center and other groups to basically predict the storms themselves and try and get an idea of will a storm form? How intense might it be? And we use that to formulate outlooks which are issued up to eight days in advance. The final outlooks are down to the day, then are proceeded by tornado watches. Tornado watches say, “the conditions are likely to get a tornado in the next few hours.” And then a tornado warning is, “the tornado was coming to you.”

What are scientists still trying to understand about tornadoes? And what makes them difficult to study?

Really pinning down why is it that a given scenario produces a tornado, I think, is a question we’re still sort of trying to wrap our heads around.

It’s not uncommon to have field projects in which you’ve had extreme trouble actually getting out in the field to observe a tornado. At the same time, we’ve had other situations where we’ve had very good observations of these events. The challenge is that there are a lot of physical processes at very small scales that really matter for tornadoes. That small scale influence means that it’s very hard to actually identify which storm is going to produce a tornado. Which one is it? On the same token, is it going to produce a long-track, higher-scale tornado? Or is it going to produce a relatively weaker tornado? We don’t have great answers for that.

We really do need a greater degree of field observations. We need more modeling studies to understand what’s going on. There’s a lot of science and research that needs to be done and those sort of things need federal support to do so.