Why is the consoles forecast so inaccurate?
If RUC, ETA, NMC, and other huge computer programs running on supercomputers hose up the forecast pretty routinely, it is probably a bit much to expect a ~$500 home weather station to do better. The forecast should be used for “entertainment value only.” It is interesting to note when it gets it right, though. Davis claims the console forecast is based on latitude/longitude, time of year, wind speed and direction, barometric trend, temperature and current precipitation or lack thereof. In experiments I conducted, using latitude and longitude for Northern Virginia; Palm Springs; Sitka, Alaska, and Michigan, the forecasts were always identical. The forecast also drastically changes quite frequently from hour to hour, which hurts its credibility as well, but again, it is more a feature of interest than something you can count on. Don’t forget the best forecasting tool we have – a window! By the way, the forecast is updated once an hour, on the hour, something not noticed in the manual by
