|
|
|
|
Almanac Accuracy
Wed Oct 01, 2003
Listen in RealAudio 
Hi, I'm Bryan Yeaton for The Weather Notebook. Today, commentator David Laskin has a
beef.
When I first heard that Seattle research meteorologist Nick Bond was taking on the Old Farmers
Almanac, I wasn't sure who was David and who was Goliath. After all, Nick has the scientific
method to back him up - whereas the Almanac has what? A secret formula locked in a tin box
somewhere in New Hampshire? On the other hand, the Almanac is a veritable monster of
commercial success, with millions of fans relying on claims of 80 to 85% accuracy for its
year-long forecasts. Which was exactly what raised Nick Bond's hackles. Not content with
sneering, Nick did a statistical analysis. He assembled 13 years of Almanac winter forecasts
for the Pacific Northwest and compared them with the monthly mean temperature and
precipitation anomalies observed by the National Weather Service.
The results? For temperature, the Almanac forecasts had a correlation of slightly less than
zero with the observed anomalies - meaning that the little tin box has essentially no skill in
predicting temperature. But with precipitation, The Almanac forecasts actually had a
significant negative correlation with the observations - meaning that there were many winter
days when the tin box was not only inaccurate but downright wrong.
So why has the Old Farmers Almanac been going strong since 1792? Clearly there is something
irresistible about its peculiar mix of homespun folksiness and long-range meteorological
nerviness. Thanks to Nick Bond, we now know beyond a shadow of a doubt that these long-range
forecasts are hokum. But, if you're obsessed with weather, how can you resist sneaking a
peek?
The Weather Notebook is funded by The National Science Foundation, and Subaru, Driven By
What's Inside.
Today's Links
The Old Farmers Almanac
http://www.almanac.com/index.php
|
|
