Almanac Weather Claims vs Scientific Forecasting
What the Forecasters Actually Claim
A groundhog emerging from a burrow and a two-century-old book with a yellow cover share one curious trait: both are consulted for weather predictions that neither can scientifically produce. The gap between them and a modern forecast is not a matter of opinion — it is a matter of physics.
The Old Farmer’s Almanac, first published in 1792, and the Farmers’ Almanac, founded in 1818, both claim accuracy rates that would embarrass a national meteorological service. The Old Farmer’s Almanac states on its website that it correctly predicts the weather about 80 percent of the time. [1] The Farmers’ Almanac describes its own success in similar terms, saying many longtime followers maintain forecasts are 80 to 85 percent accurate. [2] These numbers sound authoritative. They are also the point where the trouble begins, because no independent body has ever verified them.
J. Marshall Shepherd, an atmospheric scientist at the University of Georgia, has long discounted almanacs as legitimate sources of weather information. [3] He told USA TODAY that scientific weather forecasting and short-term seasonal prediction have advanced significantly, and that models, observational systems and the knowledge base have made short-term prediction, hurricane tracking, and El Nino forecasts quite reliable. [7] Yet people still default to groundhogs, almanacs and folklore, he said.
The Timeline Between Methods
Consider how a farmer planting in April and a ship captain routing around a storm both depend on knowing what comes next. One consults a book printed two years earlier. The other consults a model updated twice a day. The distance between those two timelines — two years versus twelve hours — is the distance between the almanac method and the operational forecast.
The Farmers’ Almanac says it makes its forecasts two years in advance and does not go back to change them after an edition is printed. [2] Routine weather forecasts, by contrast, are “only really accurate out to about two weeks,” according to Peter Goble, assistant state climatologist at the Colorado Climate Center at Colorado State University. [4] That two-week horizon is not a limitation of effort. It is a limitation of physics.
To make longer-term seasonal outlooks, forecasters use advanced computer models that combine huge volumes of data and help meteorologists assess the probability of temperatures outside the norm. Goble told USA TODAY that these models look at relationships between the slower-moving components of the climate and how they tend to affect weather. [7] This is not fortune-telling. It is probability estimation. A May 2023 post on the National Oceanic and Atmospheric Administration’s climate.gov stated that forecasters will never be able to say a month in advance that there is a 100 percent chance of a wetter-than-average May. [6] The climate is not that predictable, wrote the post’s author, Rebecca Lindsey.
The almanacs operate on a different principle. Their exact methods are kept secret, but they offer limited insight. The Old Farmer’s Almanac says it uses solar activity and sunspots. [1] The Farmers’ Almanac says its formula incorporates sunspot activity, the tidal action of the Moon, the position of the planets, and more. [2] Neither method has been published in a peer-reviewed journal, and neither has been tested against the operational models that produce the two-week forecasts Goble describes
Sources
1. The Old Farmer’s Almanac — Organisation (homepage)
2. Farmers’ Almanac — Organisation (homepage)
3. University of Georgia — Organisation (homepage)
4. Colorado Climate Center — Organisation (homepage)
5. Colorado State University — Organisation (homepage)
6. National Oceanic and Atmospheric Administration — Organisation (homepage)