So maybe some snow after all?

I think it's unhealthy to follow weather events from 168 hours out- too much stress related to model variability, both intra- and inter-model differences! Nonetheless that's what I've been doing the past 72 hrs, ever since the hint of a snowstorm for the I-95/mid-Atlantic corridor started showing up in the models late last week. The upper-level trough position is really good: a digging trough that tilts negatively and pivots right over the southeast coastal plain, forcing a pretty deep surface low to bomb out just to our southeast. Unfortunately - for whatever reason - the well below-normal cold air we've had the past ... hmm ... 6 weeks is taking a vacation to Canada, leaving us with a thermal profile only supportive of rain at the onset of the precipitation. However, the past several model runs have hinted at the formation of a secondary area of precipitation, *behind* the mid-level low (I know... the forcing is strange...) However, my poor man's multi-model ensemble approach (just looking at ECMWF, GFS, NAM, and Canadian) says there's consensus for this 'secondary' precipitation region... and it might actually be accompanied by cold air!

Here's hoping: we still have 48 hrs to watch it, as it's predicted to occur at the end of the event (sometime late Wed night or Thu). Attached are the 48-hr and 54-hr 1000-500 hPa thickness (dashed contours) and 6hr precip (shadings, in inches) 00Z NAM predictions. I know HPC is discounting the NAM as being "too cold" in its thermal profile, but I'm not sure it's so prudent to completely ignore multiple instances of the same prediction across model initiations (that's called 'consensus'... it may be wrong, but it is at least consistent).