Punxsutawney Phil, the most popular weather-forecasting groundhog, predicted an early spring this year. In the folk myth, if the groundhog casts a shadow, we’re in for six more weeks of winter. If it doesn’t, spring will come earlier.
You may not want to break out the galoshes yet, though. The Stormfax Almanac reported that the rodent’s predictions have been correct 39 percent of the time. No offense to Phil, but that accuracy rate kinda stinks. The National Weather Service tries for a seasonal forecast accuracy rate of at least 60 percent, according to Livescience.
Many variables affect the accuracy of a forecast.
For example, it’s easiest to predict precipitation, in part because you’ll have precipitation or you won’t, according to a story in Gawker that used information from ForecastAdvisor, a company that grades the accuracy of U.S. forecasting organizations like the National Weather Service and AccuWeather.
“Predicting clear skies every day would net you 70 percent accuracy in many parts of the country,” said the Gawker story.
Prediction accuracy also depends on regions of the country. For example, I used to live in eastern Colorado, where one day’s high and low temperatures were sometimes more than 50 degrees apart, even during the winter. That’s common in the Midwest.
The farther out in time meteorologists try to predict the weather, the lower the accuracy rate of their predictions, according to a look at forecast data published in the Huffington Post. This is true even on a week-to-week basis.
Another story in Gawker compared U.S. seasonal forecasts for winter 2013-2014 to the actual national weather for the period. The results were quite different. Seasonal forecasts like Punxsutawney Phil’s just estimate odds that an area will experience a certain type of weather, based on comparing current patterns to historical ones (one article compared it to the way buying baby products at Target suggests that a person may be having a baby). For example, because this has been a strong El Niño year, meteorologists could suggest that major storms were more likely for the southeast and much of California.
Meteorological organizations use different models to predict things like storm tracks, storm speed and precipitation. This can cause problems for people trying to decide how bad a major storm will really be. For example, the January 27, 2015 blizzard that clobbered New England was first considered easier than feared– because it was forecast to slam New York City. It was the sixth-largest storm recorded in Boston.
So, while meteorologists do the best they can – and frankly, their best is far better than it was even 10 years ago, you shouldn’t rely solely on their information for seasonal planning.
Instead, use a dash of early planning. At the beginning of winter, if you live in an area that gets bad snowstorms, make sure you’ve got snow-clearing equipment. Don’t wait until right before a storm, when supplies may be running low. Often, the best time to go shopping for seasonal supplies is at the end of a season, when stores put them on sale.
Make sure you’ve got food, water, and heating supplies.
In short, common sense, uncommon as it may be, is probably the best course in preparing for a season. And not paying attention to a 130-year-old celebration of a rodent.
Who do you trust more with your long-term forecasts, the weatherman or the groundhog?