Last night’s 1 AM models continue with the major snow storm for Philly. Both models have snow starting earlier tonight (Fri), with snow starting as early as 7-8 PM, with a few inches by midnight.
As is often the case, there are some differences between the NAM and the GFS models, with the NAM QPF up to 2 inches of water and the GFS a bit lower than its previous run at 1.40 inches water.
In past storms, the average QPF between the NAM and GFS has worked. That would still bring us into the 20-25 inches range for snow.
People ask me two things on the blog: why don’t I quote the ECMWF more? ….and why do I talk about the NAM model when no else does for winter storms?
Regarding the ECMWF (European Model), it is subscription only. We’re talking big bucks for their data. They only provide certain data for free and QPF isn’t one of them. If you read the NWS forecast discussions, the ECMWF hasn’t been all that great this season, so I’m not missing much. (It was the ECMWF that gave us that “fizzled blizzard” last winter.)
As far the NAM, for some reason, maybe political, the mesoscale modeling branch of the NWS, which runs the NAM model, isn’t taken too seriously for winter season storms by the global forecast modeling branch of the NWS.
That said, after years of doing these forecasts, I find the NAM can be remarkably good about rain/snow lines and even larger-than-predicted snowfalls. Where the NAM goes wrong is overestimating the QPF several days in advance. But we’re getting into the timeframe where the NAM data is very useful.