|Sep17-10, 02:21 PM||#1|
Predictive power of the Farmer's Almanac
The Farmer's Almanac is supposedly predicting a warming trend over the next week. The FA has always mystified me. I wanted to corroborate or refute my understanding of it.
It has always been my understanding that the FA is based on a predictions from empirical past events, without regard to explanatory theories. i.e. Behind the scenes there's some process such as "A warm winter has followed a wet summer in x out of y years, therefore, since it has been wet, this is likely to be a warm winter."
With enough data, it seems to me you might be able to get accurate predictions out of it, but it also seems to me that, without any science behind it making theories, you have no way of determining what causes lead to what effects (maybe the warm winter actually followed a summer with heat waves on the West Coast, or whatever).
It seems to me the FA is premised on the same logic as predicting a coin toss.
"Our data shows that, x times out of y, a tails followed two heads, therefore, since we just saw two heads, we are predicting tails." This, of course, has zero predictive power.
So, is the FA based on more than this rote method of 'what was, should be again'?
|Oct17-10, 12:17 AM||#2|
My understanding is that it is based on the cyclic nature of weather. I don't think they look at the sun's sunspot cycles, but those cycles are a cause for some years to be warmer or cooler, etc.
Up to now those cycles have been followed a regular pattern, but the new sunspot cycle has been slow to start which could cause a significant change in weather patterns.
|Similar Threads for: Predictive power of the Farmer's Almanac|
|A farmer and a house||Calculus & Beyond Homework||3|
|Can I find a free 2010 sky almanac on line?||General Astronomy||4|
|Old Farmer's Almanac: Long range weather forecasts||General Discussion||0|
|PE Farmer||Introductory Physics Homework||2|
|Quantitative counting of predictive power||Beyond the Standard Model||0|