6z gfs agrees even more But very interesting writeup from FWD this morning. I guess they feel the pressure. Global model snow/ice accumulation forecasts... - The temporal resolution (how many hours apart) of global models in the extended period is usually 6 hours. The model generates a precip-type forecast for the first hour of each time-step and keeps it constant for the entire 6 hour block of time. In the case of passing arctic fronts and/or changing temperatures, the precip type in the real-world will be continuously changing. However, the model is unaware of these temperature changes between time-steps and assumes a constant p-type for 6 hours. The model then generates a 6-hour QPF (Quantitative Precip Forecast or how much liquid water will fall) and assumes all of the QPF will fall as the pre-determined p-type for the entire 6 hours. If the meteorological situation involves rapidly changing temperatures, such as behind arctic fronts, this effect creates wildly unreliable and inaccurate snow & ice accumulation forecasts. Next week is one of these situations, so please take any model snow/ice forecast with a huge grain of salt for the next couple days.