I hope the title doesn't sound too harsh, BUT with Christmas in the long-range forecast view, model-buzzers are all atwitter. You may have already seen some potential forecasts popping up on your news feed. A pattern change and potential wintry system is just adding to the so-called "model pandemonium." So first things first, we'll be carefully tracking next weekend's forecast, looking for trends, and keeping you informed of any changes during each newscast.
Through my short forecasting career, I think Bill and Tom would agree with this too, there are pluses and minuses to having ALL of this weather model data. It's great to have lots of "options" to choose from; all of these options can a bit overwhelming and all consuming too. A forecaster, myself included at times, may rely too heavily on the model data and less on our own experience and knowledge. We may become more of a "modelogist" and less of a "meteorologist."
Earlier today I saw a great post from the National Weather Service in Kansas City about long-range forecasting. The post uses an example of a pachinko board. All forecasting models start with the same variables - current conditions - temperature, pressure, dew point, thickness, etc. These initial conditions are then ingested into a complex series of equations to predict future conditions. A model has an excellent handle on what is going on right now; the current conditions are like a freebie on an exam. Short-term forecasts (one to three days in the future) can be considered fairly reliable. There are more possibilities with a mid-range forecast, but the model only gives you one solution. And the further away from the initialization time, the further into the future a model goes, the range of possibilities becomes even wider. A tweak in the storm's track or speed early in the model run can make the world of a difference eight to ten days out. A tiny change at day two can mean the difference between rainy or white Christmas, for example.