Same Storm, Two Different Words — A Conversation with the Forecaster Who Couldn't Say "Tornado"

For more than six decades, the United States Weather Bureau prohibited its forecasters from using the word "tornado" in public forecasts. The ban originated in 1886, when the chief signal officer shut down the military's experimental tornado-forecasting program with the stated rationale that "the harm done by such a prediction would eventually be greater than that which results from the tornado itself."¹ Successive editions of Bureau regulations carried the prohibition forward in near-identical language: Forecasts of tornadoes are prohibited.²

In 1938, the Bureau allowed a narrow exception: forecasters could use the word when communicating with emergency management personnel.³ The public ban held. During World War II, the Bureau built networks of volunteer storm spotters around munitions plants and military bases, networks that used "tornado" freely in operational communications, while civilian forecasts continued to substitute the approved euphemism: "severe local storms."⁴

We spoke with Clement "Clem" Fairweather, who joined the Weather Bureau's Kansas City district office in 1934 and spent sixteen years writing forecasts across the Great Plains. He is, we should note, entirely fictional. But the regulations he worked under, the instruments on his desk, the research papers in his filing cabinet, and the storms he couldn't name were all very real. We've reconstructed his professional world from the documented record. His opinions, such as they are, belong to a man who never existed about a policy that killed people for decades.

You joined the Bureau in 1934. When did you first recognize a tornado setup on the map and have to write around it?

Clem: You remember the first time. You're looking at the surface map. Low tracking northeast across Kansas, southerly flow pumping Gulf moisture up into the warm sector, temperature gradient tightening along the front. Finley published his fifteen rules for identifying tornado-favorable conditions back in 1888, and they were right there in the literature if you cared to look.⁵ So you know what you're looking at.

And you sit down at the teletype and type severe local storms.⁴

That's the approved phrase. You know it means nothing to a wheat farmer in Saline County. "Severe local storms" could be hail, could be straight-line winds, could be a heavy thunderstorm. Could be a tornado. The farmer doesn't know which. The whole point is that he can't.

The Bureau partially lifted the ban in 1938. Did that change anything in practice?

Clem: I could pick up the telephone and say "tornado" to the county emergency coordinator.³ That was new. But the district offices had been living under this rule for fifty years by then. Culture doesn't flip because Washington revises a regulation. Very few tornado forecasts went out through the entire 1940s, even after the lift.⁶

The feeling around the office was that the public wouldn't understand what a tornado forecast meant. That was the line. I heard it so often I started wondering whether we were protecting people from panic or protecting ourselves from being wrong. Those two things look identical from the outside, but they feel very different when you're the one not picking up the phone.

During the war you were running spotter networks around munitions plants. Walk me through how that worked.

Clem: That's where it got genuinely strange. By '43, we had volunteer spotters ringing every ammunition factory and military storage depot in the region.⁷ Over two hundred networks by the end of the war.⁸ Systematic operations. We'd brief the spotters, they'd watch the sky, and if they saw rotation or a funnel, they'd call it in and we'd get the plant evacuated or get people to shelter. We used "tornado" freely in those communications. It was expected. The whole point.

But those plants weren't sitting in empty fields. They were near towns. Near farms. Your spotter is standing on the roof of a shell-loading facility watching for the same tornado that's about to cross the highway and hit the houses on the other side. I can warn the plant. I can call the base commander. I can say the word. The family in the farmhouse a mile east gets "severe local storms" on the radio, same as always.

Washington had decided a worker in a munitions plant deserved to know what was coming and the worker's wife at home did not. I understood the military logic. You can't lose an ammunition depot. But understanding the logic didn't make it sit right.

After the war, they disbanded the networks.⁷ Just gone. The spotters, the communication lines, the whole infrastructure. The plants didn't need protecting anymore, so apparently nobody did.

By the mid-1940s you had radiosondes giving you upper-air data, you had Showalter and Fulks' research on tornado environments. How much could you actually see coming?

Clem: More than people realize. More than people want to realize, frankly, because the implications aren't comfortable.

Radiosondes changed everything. Routine upper-air soundings starting in the late thirties, temperature and humidity all the way up through the column.⁹ Before that you were reading the surface map and guessing what was happening at altitude. After the radiosonde, you could see the ingredients stacking up. Warm moist air at the surface, dry air aloft, wind shear through the profile. The instability was right there in the sounding, printed on paper, sitting on your desk.

Then Showalter and Fulks published their report in '43. A hundred and sixty-two pages on the synoptic environments that produce tornadoes, published by the Weather Bureau itself.¹⁰ Our agency published the science and continued to prohibit us from applying it in public. I read that report. I suspect most of us in the district offices did, though I can't say for certain it reached every desk. It confirmed patterns we were already recognizing. The knowledge was accumulating faster than the institution could justify sitting on it. But institutions don't need justification. They have regulations.

Did you ever consider just saying it?

Clem: No. Next question.

April 9, 1947. The Woodward tornado.

Clem: (long pause) The forecast out of Amarillo that day called for cloudy skies, clearing by noon, temperatures near sixty.¹¹ An F5 tore through Glazier and Higgins in the Texas panhandle and then Woodward, Oklahoma. A hundred and eighty-one people killed across the outbreak. Over a hundred in Woodward alone.¹²

The word "tornado" was still prohibited in public forecasts. Nine years after the partial lift.

I'm not going to sit here and perform what I felt. The point is that 181 people died without advance notice from the agency whose entire job was to provide it. The knowledge existed. Finley's rules, the radiosondes, Showalter and Fulks. We knew what those setups looked like. The institution knew. The institution chose not to say.

Less than a year later, two Air Force officers at Tinker made the first successful tornado forecast. Using Bureau research.

Clem: Fawbush and Miller. A tornado hit Tinker Air Force Base on March 20, 1948, and five days later they recognized the same pattern rebuilding.¹³ Issued a tornado forecast for the base. They were right. And the research they'd been studying in those five days? Showalter and Fulks. Our report. Published by our agency five years earlier.¹⁰

They were Air Force. They weren't under our prohibition. They could say the word, and they used our science to say it. I've heard people call it bold, call it a breakthrough. It was honest. That's all. They looked at the data, they saw what it meant, and they said so. They did what we should have been doing, what we had the knowledge to do, what we were told not to do.

In 1954, four years after the Weather Bureau formally lifted its ban, television weatherman Harry Volkman became the first broadcaster to say "tornado" on air, at WKY-TV in Oklahoma City. He hesitated before doing it. He was worried he might be arrested: the Federal Communications Commission still had its own prohibition in effect.¹⁵

Footnotes

1. NOAA Storm Prediction Center, "History of the SPC," spc.noaa.gov/history/early.html ↩

2. NOAA, "History of Tornado Forecasting," noaa.gov/history-of-tornado-forecasting ↩

3. Atlas Obscura, "Until 1950, U.S. Weathermen Were Forbidden from Talking about Tornadoes," atlasobscura.com ↩ ↩²

4. Pacific Legal Foundation, "Tornado Forecasting Was Banned in the U.S. for 60 Years. Why?" pacificlegal.org ↩ ↩²

5. NWS Virtual Lab, "John P. Finley's Rules for Tornado Forecasting," vlab.noaa.gov ↩

6. Galway, J. G. et al., "The Birth and Early Years of the Storm Prediction Center," Weather and Forecasting, vol. 14, no. 4, August 1999 ↩

7. NOAA Storm Prediction Center, "A Brief History of the Storm Prediction Center," spc.noaa.gov; Wikipedia, "Storm Spotting," en.wikipedia.org/wiki/Storm_spotting ↩ ↩²

8. Wikipedia, "Storm Spotting," en.wikipedia.org/wiki/Storm_spotting ↩

9. Wikipedia, "National Weather Service," en.wikipedia.org/wiki/National_Weather_Service ↩

10. Doswell, C. A. et al., "Historical Roots of Modern Tornado Forecasts and Warnings," Weather and Forecasting, vol. 14, no. 4, August 1999 ↩ ↩²

11. NWS Amarillo, "The Panhandles' Only F5: Reconstructing the April 9, 1947 Higgins-Glazier Tornado," weather.gov/ama/1947tornado ↩

12. NWS Oklahoma, "Deadliest Tornadoes in Oklahoma History," weather.gov/oun/tornadodata-ok-deadliest ↩

13. Wikipedia, "1948 Tinker Air Force Base tornadoes," en.wikipedia.org; Maddox, R. A. and Crisp, C. A., "The Tinker AFB Tornadoes of March 1948," Weather and Forecasting, vol. 14, no. 4, 1999 ↩

14. Atlas Obscura, "Until 1950, U.S. Weathermen Were Forbidden from Talking about Tornadoes," atlasobscura.com ↩

15. 3newsnow.com, "WX History Special: The Incredible Story of the First Tornado Warning," March 25, 2024, 3newsnow.com ↩