UPDATE and BUMPED, 3:07pm Central: The map at the bottom shows what one of the computer models was forecasting on Sunday. We have turned out to be correct in that a major blizzard developed. Take a look at this graphic:
North Kansas City has light rain and 34° (it will change to snow there this evening) while, 50 miles to the north (orange areas), a blizzard will rage. In Iowa, it may be to "historic" proportions."
The latest forecasts, with video, and airport conditions available here: AccuWeather.
Original Post: One of the eternal questions asked of meteorologists is why snow forecasts can be incorrect, occasionally by large margins. The reason is that forecasting the amount of rain and its geographic placement is difficult because water vapor (the raw material that Mother Nature uses to create precipitation) is highly variable in the atmosphere. When dealing with snow, the difference between 0.5 inches of rain and 0.8 inches of rain that would barely matter in summer can be the difference between 5 and 10 inches of snow (in heavier snows, the water to snow ratio is often higher than the oft-quoted 1:10) which does matter a great deal in most areas of the U.S.
Finally, there is the gradient in areas where temperatures vary considerably. Take a look at this computer-generated forecast (thanks Earl Barker) for the upcoming storm.
I'd rather forecast just about anything than a major winter storm -- for me, it is the toughest form of weather to deal with.