high risk

Friday afternoon could be a bit fun as the strong vort approaches. Convective showers are likely, and with very steep lapse rates, it looks like a few folks could see some lightning.

first thunder of the year in North Laurel!!

We saw this exact scenario 2 weeks ago, also on a Sunday. The wedge doesn't give up easily, and the temperatures are going to bust, except in the western areas. Models are still emphatic that we'll break out in a few hours and get to the upper 50s - maybe we can get that high, but I think that 60's in the metro areas are pretty unlikely.

Lots of talk about the light snow not being able to stick and references to the struggles of last Sunday, but this isn't a Sunday redux. 1) Air and ground temps going in to this storm are much colder 2) it's not necessarily a light snow event - if you look at the NAM3 in particular, it's mostly a 2hr burst of steadier snow that will rapidly cool the low levels Don't get me wrong, it's still a weak, fast-moving system, and DC and points south will struggle to get accumulation no matter what, especially if the GFS is right getting us to the low 40s. But north of DC could get a quick 1-2" if the NAM3 idea of limited heating prior to onset and some short-lived heavier rates is correct.

agreed. nice thump of afternoon snow for areas north of DC with temps in the mid 30s falling to lower 30s.

except the the guidance has repeatedly indicated that the primary warm layer will be around 800 mb

it's really worth noting that the NAM and NAM3 10:1 maps show accumulation overnight for much of the I-95 corridor, even though the p-type is mostly rain. That means that the microphysics inside the model are cranking out some sleet for sure, even though the p-type shows rain due to the p-type being computed outside out the model purely based on the thermodynamic profile. Don't take the accumulation numbers literally, as a 10:1 ratio obviously shouldn't apply to sleet, but the fact that it's non-zero is telling us that the model has at least some frozen falling in a profile that is so close to supporting more frozen, and it supports the idea that it wouldn't take much cooling to be a much bigger deal.

LOL. It was too much and too confusing for me to write "for the southern half of the northern tier."

can we stop saying things like the NAM3 has 5" and crap that like? The TT 10:1 map has 5". The model does not. If you want to know what the actual model is doing, look at the snow depth change or Ferrier maps. They're much lower, indicating significant riming of the snow, consistent with the forecast soundings which show a possible warm layer around 800. It's marginal with that warm nose, so those depth and Ferrier maps may end up too light, but the model does not have 5".

nice event for the northern tier

NAM is a wet outlier, but the NAM3, CMC, and GFS are somewhat in line for 1-2" across parts of DC Metro.

NAM3 brings us back to reality a bit, but it does show the potential for a couple of heavier snow bands.

a reminder to, at the minimum, split the difference between the 10:1 and accumulated snow depth maps, but that is still a LOT of precip. (or at least a lot more than we were expecting to see....)

Remember that the NCEP models tally sleet along with the snow, so when you see those generous snow maps, they're because TT is applying a 10:1 ratio to the liquid equivalent. The positive snow depth change is also plotted there, and it accounts for poor ratios and overall difficulty in accumulating - it's way less snowy for this case (consistent with the sleet).

NAM has some very good 850 mb speed convergence with a modest southerly jet. The previous wetter runs of the GFS had the similar jet structure, but overnight runs have shown a weaker and more veered low level jet.

Disagree with all the people complaining about 1-8" ranges and boom/bust ideas needing to be tossed. Forecasting snow amounts isn't a contest. It's communicating the most likely outcomes along with the range of possibilities to account for model errors, localized bands, and such so that people who may need to travel or decide whether to close a business or school can assess the level of risk and their tolerance of it and make informed choices.

we CAN see the in-between hours for the GFS on the mag.ncep.noaa.gov site which has the full hourly output. You'll see that the changeover is well before 00z, but as has been noted, the GFS is well-known for being way too fast in breaking down inversions and warming the low levels

That's extremely apparent if one looks at the NAM QPF for southern MD.

it was similar with the 18z cycle. very unimpressive thump, with heavier precip coming in after the switch to sleet.

yes, although it usually does really well in these snow --> sleet transitions. I think that what is happening is that the precip total on the NAM3 across the area really isn't impressive until after a warm layer sneaks in (just after 17z). It's probably cranking out a reasonable SLR when "pure" snow is falling; it just doesn't have enough precip to work with. Not saying that this is correct - just pointing out why that map looks meager.

I certainly won't defend the SREF, but it's not a hi-res ensemble (members are 16 km). The HREF is an excellent way to view existing hi-res guidance in terms of means and probabilities, although more winter products are needed. Moving towards an hourly-updated hi-res ensemble around 2021 or so.

There won't be any further upgrades to the NAM, although it's not clear when it can be turned off. As long as the GFS struggles in cold air damming events and with instability in general, the NAMs will stay. The NAM3 sometimes runs a bit cold in damming events and can be too wet, but it's far superior to the GFS in overall handling of this type of event. We saw that very clearly with the November storm.

Feel like we have to hash this out every time, but the NCEP models generate a snow water equivalent that is generated from all frozen hydrometeors (snow + sleet). TT then applies a 10:1 ratio to that, which does not work if you have a good amount of sleet. But TT also displays total positive snow depth change which is what the model actually thinks is accumulating on the ground. It sometimes runs a bit low, but it's WAY, WAY more representative of the model is doing in this mix (or wet snow) events.

for a day or two, until it's confirmed that the implementation "worked", but it won't be available to the public

NAM wants to bring a quick shot of mess through the area Sunday afternoon. NAM nest is not on board.