John Finley, 19th century tornado researcher

January 25, 2011 5:50 PM

John Park Finley, American meteorologist, wrote the first known book on tornadoes (Tornadoes, 1887). Though some of his "safety" guidelines for surviving a tornado have since been refuted as dangerous (seek shelter on the side of a house facing an oncoming tornado!), the book remains a seminal work in tornado research.

View excerpts or download the entire book, courtesy of the University of Oklahoma. Highlights include plans for a tornado cave (costing under $350), geographic distribution, reports of damage to 300 lb hogs, and two references to what are probably the first known photographs of a tornado.

Through analysis of statistics, damage photos, and anecdotes about past storms, Finley was able to compile a list of "rules" that could signal tornadoes. In 1884, The Signal Corps allowed Finley to issue trial tornado forecasts, but the word "tornado" was banned from use for fear of public panic. Despite Finley reporting a 95.61%–98.65% degree of success, the experimental forecasts would end in 1886 (possibly due in part to opposition recalculating Finley's success rate for predicting tornadoes at only 23% -- Finley's original statistics were inflated by forecasts of non-tornado-spawning storms).*

Internal strife within The Signal Corps would eventually see Finley regulated to the records division, no longer involved in weather forecasting. After retiring from the military at age 64, he would move on to lecture about tornadoes and even provided insurance underwriters with storm risk assessments. In 1932, he opened the National Weather and Aviation School in Ann Arbor, Michigan. Finley would continue to teach meteorology at the school until he he retired for good in 1939. He would die 4 years later at the age of 89.

*One of Finley's supporters, Edward S. Holden, tried to implement a tornado warning system using telegraph poles. But it was overshadowed by a report by Henry A. Hazen, a civilian employee of the corps, who deemed that because tornadoes were "exceedingly rare" and very localized, it was impossible to pinpoint forecasts.

In addition to upholding the "tornado" ban for decades, The Weather Bureau (which assumed jurisdiction from the Signal Corps in 1890) remained skeptical of the value and accuracy of tornado forecasts. It took until 1943 for experimental warning systems to be implemented; a public outcry in 1952 (after a severe outbreak that killed over 200 people) finally helped formalize U.S. tornado research and forecasts.

posted by Wossname (9 comments total)

This is a wonderful post, I can't remember the last time I've clicked through every link in a post. Finley's book is also keyword searchable via Google Books. Thanks for reminding me to take a random walk through the full-text tornado books there.
posted by jessamyn at 5:59 PM on January 25, 2011 [2 favorites]

We had tornado drills once a month when I was growing up in southern Ohio and I remember I never felt safe until I went to a school with a sub-basement. Something about quarried stone walls 15 feet underground was incredibly reassuring on teeth-gritting scary siren day...
posted by toodleydoodley at 6:49 PM on January 25, 2011 [1 favorite]

Thanks a lot, this is amazing
posted by azar at 7:21 PM on January 25, 2011

We test our tornado sirens every Wednesday at noon. I heart meteorology pioneers.
posted by SMPA at 7:29 PM on January 25, 2011

The infrastructure was definitely in place in the 1880s and 1890s for tornado warnings. As mentioned, there were certainly a lot of articles in the scientific literature of that era talking about mobilizing telegraph operators and law enforcement people to keep tabs on dangerous storms and warn towns in the path so they could sound the alarm. The U.S. Weather Bureau, however, dragged its heels and stayed completely out of the warning business (and storm research in general) and by the 1920s and 1930s (when radio was mainstream and breaking news spread across the country, getting past the window of this FPP) fatalities and damage were rising to phenomenally high levels.

It was the Air Force that actually issued the first storm warnings in the 1950s, and Congress actually intervened to force the Weather Bureau to re-assess its priorities. But even this was half-hearted, as the office in Kansas City was relegated to the status of a "NWS unit" until the 1970s. Only by then did we start getting a proper spotter and warning architecture. There's certainly a heck of a story to be found in whatever cronyism, political maneuvering, or whatever took place to keep the Weather Bureau out of storm forecasting for all those years, including during the Reichelderfer era (and this NWS head was definitely competent, which kind of fails to explain why the USWB kept its head in the sand).

The changes made in the late 70s and 1980s were remarkable... the National Weather Service's warning system as it is now is a model of excellence, and now the Doppler radar network in the US is being upgraded to phased array radar -- absolute cutting edge stuff.
posted by crapmatic at 8:03 PM on January 25, 2011

[as the office in Kansas City] = Storm Prediction Center = National Severe Storms Forecast Center = Severe Local Storms unit = issues tornado watches nationwide
posted by crapmatic at 8:05 PM on January 25, 2011

Very timely; we just had a tornado warning this evening!
posted by misha at 9:27 PM on January 25, 2011

It was the Air Force that actually issued the first storm warnings in the 1950s, and Congress actually intervened to force the Weather Bureau to re-assess its priorities.

For a really cool description of early tornado reports, here's the NOAA page on the 1957 Ruskin Heights Torando. This was a huge F5 tornado that devastated southern Kansas City. There was a hodge podge of agencies at the time that reported on such things, so you'll see "Severe Local Storms Unit" instead of the more familiar NWS. As the article notes, they were using older radar even though they had newer WWII leftovers that had just been installed. Still, most of the reports were by airline pilots. Despite having to communicate with teletype and between airports, the accounts are surprisingly real time. This is probably due to the fact there was a large Air Force base in the area (referred to in the article as Grandview AFB).

Surprisingly, there was plenty of warning on the radio and television:

Because of warnings on radio and television, many Ruskin Heights residents were able to take refuge in their basements or with neighbors who had basements. At least fifty people took refuge in one basement, literally lying on top of each other, at East 110th street. The roof was blown off the home, yet no one was injured. Those that did not seek shelter with neighbors gathered family and friends in automobiles and drove out of the storm's path.

Last time I was in the area, there's an annoying train whistle that blows even when not at a crossing. Legend is that it is to let people know that it is a train and not a tornado (Edit: apparently not an urban legend).
posted by geoff. at 9:45 AM on January 26, 2011

What a cool subject; thanks for posting!
posted by LobsterMitten at 12:39 PM on January 26, 2011