40 years ago a rare and deadly tornado outbreak struck Pennsylvania. Are more in our future?

The atmosphere has been generous with horrific and haunting lessons in the fact that its ferocity isn’t particularly discriminating in choosing targets.

And on May 31, 1985, that target was Pennsylvania. Twenty tornadoes touched down in the state — including the most powerful twister in the state’s history, with winds exceeding 200 mph — killing 64 people.

The combination of the ferocity and geography made it “one of the rarest outbreaks on record,” Harold Brooks, a NOAA scientist who has devoted a long career to studying tornadoes, said this week.

They are especially elusive objects for study, he says, as destructive as they are idiosyncratic, at times resembling an eccentric acquaintance for whom unusual behavior is the norm.

“It doesn’t mean that Pennsylvania is in dire straits of being hit by big outbreaks all the time,” said Brooks, senior research scientist at NOAA’s National Severe Storms Laboratory, in Norman, Okla., but “it means, at least, that things like that are possible.”

In its post-storm report, the National Oceanic and Atmospheric Administration concluded soberly that “under the proper atmospheric conditions, major tornadoes can occur irrespective of the location or terrain.”

In all, the tornadoes killed 89 people, and 1,000 others suffered injuries. It was the deadliest tornado event in a decade for anywhere in the country, said severe-storm specialist Greg Forbes, who at the time was a Penn State professor.

Pennsylvania was a final stop in the outbreak that rampaged through northern Ohio, southwestern New York, and even Ontario — areas hundreds of miles removed from the traditional “tornado alley,” a loosely defined portion of the central United States.

That region is especially favorable for the clash of air masses from Canada and the Gulf of Mexico that can breed the fronts that spawn tornadoes, particularly in spring.

What was perhaps the signature feature of the 1985 disaster was the shear destructive power.

Only about 5% of all tornadoes qualify as “severe” — that is, EF3s on the Enhanced Fujita scale with winds of 136 mph.

Half of the 20 tornadoes that hit Pennsylvania that day would have qualified, including the EF5.

In the aftermath, Forbes recalled seeing 10-foot-high stacks of ripped-down trees.

New Jersey already has reached its annual tornado quota this year — two verified so far — but climatology suggests the number will increase. The state is more likely to experience its twisters in July and August.

Pennsylvania, on average, experiences 15 tornadoes annually, and its comparative advantage (disadvantage) relative to New Jersey likely is related to its size.

The Pennsylvania numbers are spread more or less evenly from April to September, according to NOAA’s Storm Prediction Center.

An EF3 touched down in Bensalem, Bucks County, in late May of 2021, with winds of up to 140 mph. That was a warning shot. Another EF3, spawned by the remnants of Ida, hit South Jersey. By year’s end more than 25 had been verified in the Mount Holly National Weather Service coverage area, which encompasses Delaware, eastern Pennsylvania, and most of New Jersey.

Not coincidentally, that was a particularly active year for twister-spawning remnants visiting the region, but in the last three years, the local remnant and twister traffic has dropped off.

A warmer world means more water vapor available for storms, but how are rising temperatures affecting tornadoes?

“It’s a really difficult problem. At this point we don’t know what any climate effect would be,” said Brooks.

Aside from the complexity of the usually short-lived storms, one of the biggest obstacles to making sense of trends is the state of data and the challenges of making comparisons with previous decades, he said.

Along with population changes and improvements in observations and radar technology, the proliferation of smartphones and social media have increased detection exponentially.

Consider that before 1950, one in 10 of verified tornadoes were on the lowest end of the scale, with peak winds 40 to 72 mph.

In 2010 they constituted 60% of all tornadoes.

Research has shown that while the numbers of days with tornadoes have decreased, days with severe outbreaks, 16 or more twisters, have increased.

But NOAA concludes that “So far, the majority of research stops short of connecting historical changes in tornado behavior to a warming climate.”

Yes, said Brooks. The data show that the lead time for a tornado warning averaged 16.1 minutes in the 1986-1990 period, before weather services offices began using sophisticated Doppler radar.

In the 2020-24 period the average was 18.5 minutes between a warning and verified sighting.

Even that, with an evanescent storm like a tornado, is not a whole lot of time, and any number of tornadoes form with no warning.

He said one advantage today is that the conditions favorable for severe storms can be sniffed out three to seven days in advance. Chances are that on a Wednesday, people might be aware that “it could be a big day on Friday.

“What I expect people to do, and this includes myself,” Brooks said, “is to pay enough attention to the forecast to know that today is the day that the weather might kill me.”