high risk

NAM nest at the end of its run looks notably better than the parent.

Mother of Mercy, RGEM!

Fairly good agreement (Euro, Canadian, Both GFS, NAM) of a Philly-NYC jackpot

except that if we fail on the CCB, 6-8" for areas north of DC probably isn't the default option. 3-5" seems more reasonable, and that could legitimately fall to 2-4" (see NAM nest).

Correct. Everyone laughed back in December when the NAM was the first model to show that storm coming much further west and ruining chances for heavy snow along the I-95 corridor. That doesn't at all mean that it's correct now, but to completely dismiss it at this stage comes with risk.

Most of the source of spread in the ensembles (SREF and HREF not included) comes from tweaks to the initial state, and one must go further into the forecast period before those tweaks start to manifest themselves as significant spread in the outcomes. If we forced spread to develop much faster in the forecast, it's likely that the spread in the longer ranges would be unusably large.

Still a bit too far north with its maximum love (for us Howard County folks), but each of the last 4 cycles has see the GFS para shift southward with the location of its max. Just need one or two more....

It's a legit question, but it seems to be more of a function of the GFS PBL scheme, as the model has had these issues for a long time.

Yeah, don't overlook this. NAM nest was notably drier than the parent. PWs seemed quite a bit lower over the source region for the moisture - not sure if it's correct, but I don't recommend completely dismissing it either.

Not really relevant to totals, but the NAM and especially the NAM nest like the idea of a lead band around sunrise Sunday followed by a break before the primary swath of warm advection snow arrives.

It has sufficient resolution; it just sucks at it.

GFS has what appears to be (IMHO) a very poor representation of the cold air damming at 12z Sunday.

Perhaps slightly, but bigger changes in the low-level temperatures are more likely if the background synoptic solution changes. Of course, you're right that as we get closer to the valid time, it's even better to use the 3 km NAM nest.

Look at how they handle the cold air damming over the Mid-Atlantic. The NAM consistently resolves this low-level cold air so much better than the GFS.

Very clear that the 00z NAM has a much stronger wave moving across the midwest than the 12/18z GFS did - looks much closer to the 12z Euro. It's also not as progressive as the GFS, but it's a little faster than the 12z Euro/Canadian.

yes, it accounts for the uncertainty in precip type in the transition zone

sh*t just got real.

If we're going to parse ICON details, the 2m temps show that at some of what is being presented as "rain" would fall into a sub-freezing air mass, with the freezing line barely to DC by 09z Monday.

Some of us had two follow-up snowfalls. That Friday was another synoptic snowstorm, but there was a vort that Tuesday afternoon that put down a band of heavy snow across DC and the east side of town. I got another 6" out of that unexpected event.

All of the standard caveats apply (duh), but I find it very interesting that three different modeling systems (ECMWF, CMC, GFSv16) have now shown the idea of a capture and long-duration event for some part of the mid-Atlantic or northeast in at least one cycle.

As bleak as the radar looks now, the models are pretty insistent that it will fill in again later this evening. I don't think that most of us will have the the thermals for snow by that point, but I'm at 29 right now, so sleet or freezing rain seems likely here in Howard County.

You ain't kidding. Looks similar to the 12z CMC.

Sleet here in southern Howard County. We knew from the forecast soundings that we had little margin for error with the profiles. Let's give this a little time to see whether rates can cool that pesky warm layer aloft just enough.

Very glad to see the dearth of 10:1 maps! Kuchera is a step up, but those will still be a bit generous in an event with such a sizable depth of the lower atmosphere with temps close to 0. It's also worth noting that the 18z NAM and NAM nest both cut down their QPF a bit, relative to 12z... I'm still going with the NAM nest as my model of choice here, and the Ferrier maps will best account for some accumulation being lost to riming with temps so close to 0 at several levels. DC people will be happy that it wants to whiten things there. For the rest of us north of DC, I think it's best to plan on 1-2 and hope that a few places overperform. Anyone expecting widespread 2-4 will likely be disappointed.

They're not, though. They have support with the HRRR and all 3 hi-res windows.