No wonder we meteorologists make so many mistakes–we’re exhausted.
A National Center for Atmospheric Research (NCAR) survey estimates that United States adults look at approximately 300,000,000,000 (300 billion) forecasts annually. It’s a good thing that we don’t have to keep that number updated on the meteorological equivalent of the golden arches–that’s almost a billion forecasts per day.
Availability
The number is, of course, an estimate based on projections (read the press release if you’d like), not a literal number, and it’s clearly an indication of the availability of forecasts as much as interest in the weather. Getting a weather forecast is as easy as clicking a mouse, and that’s what most of us do——probably more often than our bosses like based on this information.
Automation and Accuracy
Many of these forecasts are, as we’ve discussed before, automated (Forecast Accuracy)–no forecaster need. Automated forecasts of Internet sites often change every time a computer forecast model runs, resulting in quick and sometimes dramatic forecast changes several times per day. When the forecasts change quickly and often, they’re often not perceived to be very accurate–and they often aren’t very accurate.
A changing forecast that might not be trustworthy is going to result in many more people looking for forecasts than a stable forecast that’s usually right. Perhaps inaccuracy and inconsistency is good for business?
That reminds me of a Jerry Seinfeld joke–from his current tour–that a friend recently told me. Seinfeld believes that if a five-day forecast were truly accurate, then we’d only need to check the forecast every five days.
He has a point.
Memphis Weather
I first saw the forecasting survey in a Memphis Weather blog entry in July–about 80,000,000,000 forecasts ago. Thanks, Erik.
–Paul Yeager


Thanks for providing me with another excuse for why I bust a forecast once in a while, Paul! I’m just tired…
Thanks also for the reference. You continue to do a fantastic job and I enjoy reading.
–Erik, MemphisWeather.net
By: Erik on October 7, 2009
at 4:49 pm
Thanks, Erik
Do you do any long-range forecasting?
By: pyeager on October 7, 2009
at 4:52 pm
No, I try not to.
I *am* in the middle of putting together some thoughts on the upcoming winter season for my blog, and have researched your posts and references thoroughly. Thanks for those! But farther out than 7-8 days in a general forecast style, I don’t have much time to devote to.
By: Erik on October 12, 2009
at 3:12 pm