Regional increases in tornadic activity through midcentury due to global warming

The sky over the American Midwest and Southeast has long been a theater for severe weather, where warm, moist air from the Gulf of Mexico collides with cooler, drier air to fuel powerful thunderstorms. Sometimes, these storms spin up into tornadoes, violent funnels of wind that can devastate communities. For decades, scientists have watched the historical record of these events, noticing a puzzling shift: while the total number of days with at least one tornado has decreased, the number of days with many tornadoes at once has increased. This suggests that when conditions are right, the atmosphere is becoming more prone to producing clusters of storms rather than just isolated ones. The big question for the future is whether a warming planet will make these dangerous clusters more frequent, more intense, or shift their location. To answer this, researchers must look beyond simple weather reports and examine the invisible ingredients that make a tornado possible, such as the energy stored in the atmosphere and the way winds change with height.

A team of researchers has tackled this question by building a sophisticated digital laboratory to simulate the future of tornado activity in the United States. Instead of relying on broad, low-resolution climate models that are too coarse to see individual storms, they used a technique called convection-permitting dynamical downscaling. This method takes large-scale climate data and zooms in to a resolution of three kilometers, allowing the computer to explicitly resolve the formation and rotation of thunderstorms. The team ran these high-resolution simulations using data from eight different global climate models, covering two distinct twenty-year periods: the recent past from 1995 to 2014, and a future scenario from 2040 to 2059 under a high-emissions path. In these simulations, the computer does not predict a specific tornado on a specific date; instead, it identifies "proxies," which are specific attributes of the simulated storms—such as rotating updrafts near the ground and strong radar echoes—that indicate a high likelihood of a tornado forming. By counting these proxies, the researchers could map out where and when tornado-like activity is likely to occur in a warmer world.

The results of these simulations paint a clear picture of a changing landscape. The study projects that tornado activity will increase significantly across a broad region stretching from the Gulf of Mexico northward through the Midwest and into the Northeast. In the historical period, the simulations successfully reproduced the known pattern of tornado activity, with the highest frequency in the Southeast and a seasonal peak in the spring. When the team looked at the future, they found that the number of these tornado proxies would rise by nearly 30 percent in the Southeast and over 50 percent in the Northeast compared to the historical baseline. The most dramatic increases are expected during the spring and winter seasons. Furthermore, the timing of the peak tornado season is shifting; the simulations suggest the annual peak will arrive one to two weeks earlier, moving from mid-May to late April. This shift implies that the window of highest risk for communities in these regions will open sooner in the year.

The study also examined the nature of these future storms, specifically looking at "outbreaks," which are days when a large number of tornado proxies occur simultaneously. The simulations indicate that these outbreak days will become more frequent, though the rate of increase per degree of global warming is projected to be relatively less in the future compared to the historical period. Interestingly, while the total area covered by these outbreaks is projected to grow, the density of the tornadoes within that area may not increase. This suggests that future outbreaks might cover a wider geographic footprint but not necessarily be composed of more tightly packed swaths of damage. The researchers also noted that the year-to-year variability in these events will likely increase, meaning that some years could see far more activity than others, adding a layer of unpredictability to future risk assessments.

What drives these changes? The primary engine behind the projected increase is an abundance of Convective Available Potential Energy, or CAPE. In plain terms, CAPE represents the amount of energy available in the atmosphere to fuel rising air currents. The simulations show that as the planet warms, the atmosphere holds more heat and moisture, particularly from the Gulf of Mexico. This creates an environment with higher CAPE, which is a key ingredient for severe thunderstorms. The study links this increase to a shift in the path of large-scale weather systems, known as storm tracks, which are moving eastward and intensifying. This shift helps transport more warm, moist air into the central United States, setting the stage for more frequent and energetic storms. While the study does not predict the intensity of individual tornadoes, the physics suggest that stronger tornadoes are also possible under these conditions.

The findings offer a crucial reference point for emergency managers, infrastructure planners, and communities across the most vulnerable parts of the United States. The research confirms that the trends observed in the past—where tornado activity is clustering and shifting—are likely to continue under high emissions scenarios, with a projected increase in activity across a broad area. By identifying that the risk is expanding northward and that the season is starting earlier, the study provides actionable intelligence for long-term climate adaptation. It underscores that while the total number of tornado days might not rise everywhere, the days when many tornadoes strike at once are becoming more common in a warming world, demanding a rethinking of how we prepare for and respond to these powerful forces of nature.