jwilson

I measured another 3" since noon, which includes the half inch I got while shoveling. Rates were back up over an inch an hour, but the dryslot looms until the upper level trough passes through later. That's somewhere between 7.5" and 12" total. We might be able to manage a foot if the backside lasts long enough. Haven't hit that mark since December of 2020.

Too bad we couldn't pair a system like this with some kind of blocking upstream. This would have been a true big dog in that instance, but someone on the northern edge would have gotten fringed. Blocking has been the hardest thing to come by in the last ten years, though. Maybe have to wait for an El Nino return, or we could get surprised again this year. We're on the front end of this pattern. There may be more chances in the next few weeks.

Lull coming into the city proper now. Looks like Youngstown area might be the big winners. Persistent banding up there. Outside of NE and central PA locations.

Lots of sleet signatures down in Greene County, WV panhandle. Hopefully that northern progression ceases. Our snow growth has clearly diminished for now. Pixie dust.

Couldn't keep up with the snow shoveling. Inch an hour rates if not more. I measured between 4.5" and 9" so total somewhere between there, lol. Snow is drifting quite a bit because of the powder consistency.

I do think this one has some overperforming potential. It will be an interesting case study when it's all over with, and we can review how the microclimate handles these two intense competing forces. I could see things going both ways, truthfully, so good luck to everyone and may your snow stay powder.

Checking short-term trends, proclivity is more QPF which also comes with higher chance of warm layers. I think I stick with my call from earlier this AM and hope for an overperformance and minimal sleet/ZR. NBM did increase odds of >18" from 4% to 11% at AGC, but shows more pronounced chance of mix during a brief period. I wonder how the NBM accumulation products account for mixes and ratios with inconsistent output across models. Hard to say. Also, if you check NAM runs from yesterday, they didn't have zero snow at Little Rock. The thermal profile was maybe off but it did show like 3-6" of accumulation by noon today on near every run as far as I can tell. No idea what that person is looking at.

NBM is finally working again. Average precip on the 1 UTC run is 0.88" equating to a mean of 11.4" of snow. Spread is 11 to 15 between Saturday and Sunday. 70% chance of >12", 96% chance of >8" (which is somewhat nutty). That's all at PIT; odds are slightly better at AGC, actually, because of the presumably better rates. 74% and 97%, respectively. Non-snow peak odds are later on Sunday at 11% chance of rain compared to 78% chance of snow. I'm guessing that's picked up mostly by the group of mesos that are showing a warm tongue/small dry slot. I still don't like how close the primary comes to our area. Typically, when that does happen, we change over south of the city. The thermal profiles on all the models check out, but I guess it depends how much you trust them to handle the microclimate. Bufkits for NAM and GFS are all snow at AGC. Between 10" and 13" depending on the run. Ratios aren't great, so any improvement there means more accumulation. 0Z NAM actually barely gets above 10:1 the entire duration. Max is 12:1 except for a brief period of 21:1 at the tail end on 1.33" of QPF (high). That's why storm total is "only" twelve inches. If I had to do a first call, I'd probably go a little conservative on the southern regions of Allegheny, Washington, Greene, and Fayette (Mon Valley): 8" to 14" - 6" to 8" at MGW, less points south - 10" to 14" city points north. Any mesoscale banding that takes place has a significant impact on localized totals because of convergence and associated divergence. Could see lollipops of 16" somewhere in the coverage area. It's arguably low for city-south, but perhaps I'm paranoid. Too many years in MGW. I might adjust that upward tomorrow as more mesos come into confidence range. I think closer to the 8" figure if there's mix/sleet, closer to 14" if there's a clean hit. I still don't necessarily trust the models on thermals assuming the northern progression of the primary low is correct. I'll revisit this thought after the storm. Anyway, just my word vomit. It's 3 AM, give me a break!

NAM says 10 PM give or take an hour, and logs an inch by 3 AM. GFS holds back a bit more and says up to an inch at 5 AM. Starts around the same time but then takes a two-hour break. These according to Bufkit data. I'd imagine the best rates are early Sunday and then maybe again overnight into Monday. We'll get a better idea with the higher resolution models, as has been mentioned.

If you look at thermal profiles, you have the EURO/Canadian/NAM in one camp and the GFS in the other. The 6Z Euro was a slight step colder to a bunch of previous runs, as were the 12Z NAM and 12Z RGEM. Most of the models bring the 850s and surface temps above freezing up to Morgantown, while the GFS remains the most progressive with the transfer. Really, that transfer to the coastal low taking over as the primary is the key element regarding our thermal progression. If the primary low inland continues to dominate longer and maintains a northward push, that's how we get the warm tongue. The depth of the antecedent cold air should give us an advantage. The overall blocking is limited, though, which is why we're even talking about this. I guess you'd typically take a blend and cut the middle road, which should still keep PIT and most of the surrounding areas snow, but with the GFS kind of on an island, you'd take a bit less of it. I think Morgantown is a much trickier forecast. They could end up with 6" of pure snow and then a mix to rain. Either way, NWS will have to revise totals down for some of West Virginia, I think. GFS is the only model that keeps the state fully frozen. I don't think I'll be totally comfortable until we get most of the high-res models in range. It's a very delicate microclimate feature for our area. Those of you north of the city proper have nothing to worry about, though. The only other question is snowfall enhancement with the passing of the upper level trough after the main vort is through. Next weekend has potential, too, but it's a bridge too far right now.

Sorry if I repeated you, I must have missed that comment somehow. 13 UTC NBM run did juice things up quite a bit. Mean is up to 12.5" now. Definitely a consistent trend the last couple days to keep increasing totals despite the one run from overnight (1 UTC) where things went back slightly. I'm kind of hoping that settles because eventually, as you see from some of the individual runs, we start inching into "mix" territory, and we know that warm tongue is often underestimated by the models. The chances of non-snow precip increased a lot (relative, 4% to 10%), as well. To see where things could go wrong, you might compare these two shots at 66 hours. Here we see the 12Z GFS with three distinct pieces of energy at 5H. Northern vort over Montana, secondary in Utah, and primary southern vort over Mexico. Hasn't quite phased yet. This is the 12Z NAM at the same time. We can see the distinct pieces of energy are interacting, instead. Southern energy is pulling north and west as a natural condition of the phase with the northern energies. We want the phase just not too early. The 12Z Euro hedges slightly in between these two, but as far as I can tell, neither the 700 nor 850 temps get above freezing as modeled there. Silver lining is the long-range NAM isn't all that good, but it is remains a possibility. Not quite in the "comfortable" range yet.

The issue with Kuchera isn't necessarily the math but rather model output, so the maps themselves are often flawed. Ground truth is typically less, but that doesn't mean they are always wrong, of course. Snow calculation is complex, as is my explanation here (apologies). The actual equation for Kuchera is this: But nearly none of the output products in models take into account the other variables of snow growth. Lift, dendrite growth, and humidity, for example. Snowfall rates are never constant over X time. As a basic tenet, highest snowfall ratios (say 25:1) occur when most of the lift is situated at temps in the atmosphere between -12 and -18 Celcius. Otherwise known as the "snow production zone." There's a method (Cobb) that uses a layered approach to estimate snowfall ratios at a given time. It can be paired with Bufkit data to estimate total snowfall, because the Bufkit software has the highest vertical resolution to pair with soundings. Using this with the 6Z GFS run, for example, gives us 13.2" of snow on 1.04" of precip at AGC. Ratios vary from 5:1 up to 30:1 in hour snapshots. That's a higher-end result when compared with the averaged NBM data (0.79" precip average). Same method is 9.6" of snow on the 0Z GFS run. We can run this on the NAM, too, but it's not yet within range. My main point is that, looking at the 6Z Euro for instance, if people start expecting 17" they are quite likely to end up disappointed, so I had wanted to put out a general caveat emptor for those reading the maps that maybe aren't aware of the limitations.

NBM mean on today's 7 UTC run up to 11" even. Spread tightened a little more to 7" to 13" on average. 1 in 3 odds of >12" total. >8" up slightly to 68% chance. Most guidance held, so I guess we're waiting to see if a last minute de-amplification occurs. (As a side note, I see a lot of Kuchera maps, but just be careful about those as they tend to get overdone on totals.)

For what it's worth, 19 UTC NBM pretty much held serve from 13 UTC. Mean is 10.4", 73% chance of 6" or more, 63% chance of 8" or more, 36% chance of 12" or more. 4% chance of a historic level event (18"+). The spread has narrowed a bit, 6" to 13" (floor is up a couple inches). If the GFS leads us downward, that would be quite the bust considering all guidance right now, but I suppose it remains a possibility. Better plan for it.

Rare instance where the OP GFS will lead the ensembles, but the GFS has been all over the place so it's incredibly hard to trust no matter which way it goes. Frustrating model.