If you want to look at a short range forecasting model (goes 84 hrs out) check out the WRF (same as the NAM). Much more detailed mesoscale (small scale) model, does much better than the GFS when it comes to forecasting upslope or lake effect.
GFS is a good synoptic (large scale) longer range model. It is pathetic with precip totals so dont read too much into that, almost never correct. But is one of a few global models that do have a good handle on the general pattern forecasting.
If you want to estimate precip totals for the mnts it is best to watch the weather and see what patterns bring upslope snow or lake effect snow and estimate totals without reading directly what the model says.
good sites for the models:
Penn State Ewall -personal favorite, for the WRF click on the most recent time (0Z, 6Z, 12Z, or 18Z) and "select regional loop." You can get a zoomed in model forecast for our area, much more detailed than the GFS.
NCEP model page (nam, GFS and others) -Personal favorite for the long range GFS. Use the "850mb" forecast or 1000-500mb thickness maps, best for diagnosing snow/rain.
