|
Dewey Johnson
In his 2009 book, Why We Make Mistakes, Joseph T. Hallinan brings up Roswell, NM in an exceptional way.
Whereas so many other professions are prone to mistake making, he says that weather forecasters are “remarkably well calibrated.” The reason for this can be traced, of all places, to Roswell, NM and a young weather forecaster who worked there back in the 1920s. Cleve Hallenbeck had bounced around from job to job – railroader, grocery clerk, schoolteacher, etc. Finally, he landed a job with the U.S. Weather Bureau, as it was known then, who sent him to one of the more remote spots a weatherman could hope to find: the Pecos Valley of New Mexico.
Then, as now, the Pecos Valley was alfalfa country. Crops were irrigated from wells, and farmers needed to know when they would have to use their own water, which cost them money, or when they could wait for rain, which cost nothing. They also needed to know when it wouldn’t rain because long dry periods were required for curing the alfalfa hay.
Simply telling farmers that there was a “chance” of rain wasn’t sufficient. They had too much at stake. They needed to know if it was a 50% chance or 75% chance or a 100% chance. So, Hallenbeck began recording and including these probabilities in his forecasts for the Pecos Valley. This was in the 1920s.
It took a while, but these probability statements caught on. Hartford Connecticut began using them in 1954, San Francisco in 1956, and Los Angeles in 1957. In 1965, the National Weather Service initiated a nationwide program in which precipitation probabilities were included in all public weather forecasts, a program that has continued, largely unchanged, until today.
As a result, weather forecasters have established a lengthy track record of predictions. But they also possess the record of the actual results. They know if it actually rained on a certain day. When the predicted results and the actual results are laid out, forecasters come out incredibly well. One study of more than 150,000 forecasts made over a period of two years found that forecasters were nearly perfectly calibrated. For ex., when they predicted a 30% chance of rain, as they did 15,000 times in the study, it rained almost exactly 30% of the time.
So, in addition to Chick, Demi, Nancy, John Denver and John Chisum, Stauback, Joe Bauman, Roy Rogers and Grace Wilkins, Dr. Goddard, the 1956 World Series Little League Team, Tom Brookshier, Robert O., Peter Hurd, Paul Horgan, Mike Smith, Michael Blake, Mine That Bird, and others, I’d hope that there is at least a 95% chance that Cleve Hallenbeck, weather forecaster extraordinaire, will be inducted to the Roswell Walk of Fame! Forecasters say that there is only a 10% chance that such a ceremony will be cancelled because of rain.
|
.png)
