jwilson

Well, this is the last minute trend we were worried about. Seems we're now running into a consensus and that takes the truly more impressive totals off the table for PIT. Eight inches is not a bad event, but now we're getting to the point where that looks like our absolute max instead of closer to the lower or middle end of averages. It's a fast reversal of the trend that favored us for a couple days, but it also means that trend could continue into tomorrow. Hopefully not. Analysis - 0Z 3K NAM is unequivocally worse for PIT. With the 18Z NAM, we had snow from 18Z Wednesday (1 PM) until 8Z Thursday (3 AM). 1-1.25" QPF, which actually cut back from 12Z (which was mostly 1.25" in the entire county with a small blip of 1.5") and 6Z (similar). At 0Z, half the county gets only .5 while the eastern edge gets .75. Most of Allegheny gets 6" with no high spots. 12Z and 18Z hit a foot in spots (with the former hitting a foot over the entire county). That's off the table for now. Timing is generally the same; onset at 17-18Z and ending at 8Z Thursday (very light). The biggest difference is this run keeps PIT outside of all the major deform bands and under generally lesser snowfall rates. You can see the compacting of the precip shield with this 0Z NAM run. The western side of the system looks much drier than before. It snows for 12 hours or more and we get 6" of snow for it. You can do the math and see that the rates are mostly abysmal. The best snow growth - according to the 3K - happens during the beginning of the storm. We maximize lift only one hour after the onset (2nd hour of snow) which is followed by one more hour of good lift, but then after that, it's largely garbage snow growth. To my novice eyes, it looks like there are much stronger westerly winds above the surface leading to dry air intrusion, thus effecting ratios. There's also considerably less omega and as I understand it, cold air advection (which is sinking, not rising air). These decreases are additionally reflected in the SREF plumes, which dropped substantially. The average on 15Z approached a foot; it is now down to 8.5" or so. Instead of more than a dozen members over 12", only two now. The majority (18) are still clustered between eight and twelve inches, however. Only eight members are under the average. That's obviously a flinch but not a total loss. These totals are usually overdone regardless, IIRC. All the other mesoscale models dropped our totals, as well, except the RGEM which kept them steady. The WRF-HMM is the most progressive, by far, and gives us only 2-3" in Allegheny. The ARW2 went from an absurd 18-20" down to a "measly" 6-10" (which is likely far more accurate, but it still is a considerable difference). The last thing I'd say is that, once again, this is going to come down to mesoscale features of the system. I don't like our odds of hitting 12" anymore, perhaps less than a 2% chance, but our floor hasn't really decreased much, if at all. 4-8" is a safe bet with no shutouts on the table, at least. Hopefully this trend doesn't continue. It does, however, look like the precip shield will ultimately be less impressive and more susceptible to dry air inundation on the western edge.

Yeah, the 18Z GFS wasn't great, pretty much a "worst case" for us (only drops 3-6" in Allegheny, which would be a bust of sorts). However, keep in mind this storm will be heavily dependent on mesoscale features like banding, and that is not in the GFS' wheelhouse. The GFS has one major band and puts Pittsburgh under nothing but subsidence. That's still a possible scenario, no doubt, but I think you have to resolve those features using other higher-resolution models plus forecasting. On a 1 to 10 scale of bust potential, this storm is probably a 7 or 8. But because of that, it could also overperform, so there's good and bad. Unfortunately, there's also some unknown that comes with it. I still think our absolute max potential is 14" with a low probability (say 10%) while 4" is the floor. I'm good with anything over 3" in mid December, and we've already done it once!

Every model gives PIT at least six inches. Mesoscales included. Some are better, some are around that mark. Averaging them, we get about 11" (across six different models and assuming 10:1 ratios through the entirety). Perhaps ironically, the SREF plume average is also right around 11", which is up slightly from the previous run. There are less lower members and two that are now higher (over 17"). That's the good news. The bad to be aware of but isn't confirmed yet: I hope we don't see a big SE correction last minute, but it's not impossible.

Probably mid-afternoon (1-3 PM). There's some flux, but the evening commute will almost certainly be poor.

At this point, 4-6" seems like the floor with the initial onslaught of snow. The question becomes where the CCB sets up and what kind of rates it can produce. NWS recently updated to puff the totals a little. I think that's a fine forecast for now. We have through 12Z tomorrow, give or take, to see what the mesos say.

Yeah, the GFS isn't as good as the mesos for PIT, but it bumped up considerably from 6Z to 12Z. A nasty gradient in the county from West to East.

Wow, the 12Z NAM doesn't back off. That's incredible. The last time I remember the NAM scoring a coup was 2016, when it was north of all of globals, then with 24-36 hours left, all the globals came north to the NAM. I'm not buying just yet, but I've got my wallet out. Even with the Ferrier correction on the 3K, Pittsburgh metro gets a foot. It shifts the jackpot from Harrisburg to Lock Haven, which isn't insignificant (about 70 miles). If PIT gets under the CCB as it forms along the axis, it could happen, but that's relying on a very specific feature.

Basically the "dry conveyor belt" as the low pressure center begins to mature and occlude, forming that traditional "comma" shape. You can look at a comma and see how there's a natural space between the head and the tail.

So I was checking out the 3K NAM at 6Z which is a pretty decent hit (visualized) for PIT. We wet bulb around 1 PM tomorrow and have decent lift in the DGZ for the first few hours. Then we wax and wane, transitioning to a loss of consistent lift there. Until ~05Z (which would be around midnight) when we hit maximum lift in the DGZ due to the CCB. So many acronyms, I'm sorry. Unfortunately, it doesn't last long, and the best rates appear to diminish quickly as the heaviest band moves hurriedly to the east. If you're a snow watcher, that means the best rates will be in the dark, though the onset of precip doesn't look bad. The heaviest stuff barely avoids PIT metro for a while and looks to bury Westmoreland/Armstrong/Indiana counties, for example. I suppose that's not unusual. There's also a time between 7 and 9 PM where PIT gets very close to sleeting due to southeasterlies above the surface, brought on by (I believe) extreme frontogenesis. This also kills the lift in the DGZ for a little so there might be a minor lull in snowfall rates. This is extreme over-analysis of one run, but it gives a bit more detail about one possibility as to what could occur. Glossary - DGZ - Dendritic Growth Zone (upper atmosphere where snowflake growth is achieved most efficiently; see graphic below) CCB - Cold Conveyor Belt (an aspect of "occlusion" that transports cool, moist air westward around the center of low pressure; "comma head")

The GFS was the last one to hang onto SE PA, and with the 6Z it took that away to some degree. Seems Mt. Holly needs to adjust their forecasts badly. I still can't tell if Scranton or Harrisburg will jackpot. Maybe both?

The 0Z NAM isn't really any further NW, and in fact keeps colder thermal profiles for places like Philly and NJ, but it creates a massive CCB that dumps snow over the entire middle of the state (which attempts to reach out here). Kind of a crazy look. I know the clown maps will be overdone, but I'll laugh at them anyway like a mad scientist.

Haha, you're not alone so don't worry about it. I might chase this one, but I'm still hesitant with the eventual northward jog. I agree with others and think this is going to end up a heavy sleet storm, at least for a while, from my normal locale in Southeast Montco. This is the kind of storm I remember from childhood that was also a bit of a loser for my area and much better up in places like the Lehigh Valley. Maybe the confluence can win out, but I'm not super confident in that with the relatively marginal cold. 2009 this is not.

Does the GFS, et al. still have trouble parsing snowfall from sleet? In other words, do those clown maps take into account sleet accumulation, as well, or are they measuring strictly what it perceives as snow per local soundings? I would imagine the mesoscale models do a better job at handling that differentiation, but that's a hard guess on my part.

I've seen the NAM overamped before, but this is taking it to another level. It is on its own so really it can't be trusted, especially with the GFS and Euro being as relatively consistent as they've been. I do think the GFS will jog a bit more northwest before the event. Enough to save a fringe job? I don't know. 3-6" might be a fine "catch all" and if it overperforms, so be it.

It's amazing to see how far apart the NAM and another meso model like the RGEM are in placement. I don't know that I've seen such a difference when considering the time frame. This is more instinct than anything, but I'm getting the sense this is modeled as an overperformer for certain eastern metro/urban areas. It is screaming sleet storm to me with a warm-water driven SE flow. Relatively marginal cold. This isn't like the cold air that preceded 2009 (low to mid 20s). Is the GFS too progressive? Now I don't know that things will shift as far west as the NAM likes, but as usual, a blend might be the best way to go. The Euro is kind of in the middle. It isn't great for our area, though.

Isn't there a secondary in the Ohio Valley? The trend recently has been to kill it off quite quickly, but at least it used to be there. Also, this might of interest to you guys, but I had no idea there was an 18Z CMC? Is it of equivalent value? You can see the shift in the heaviest axis of precip from 12Z to 18Z, with the latter looking a lot more like the GFS. If this off, I guess we'd find out at 0Z. 12Z 18Z

Based on relatively recent history, that area of subsidence seems fairly consistent. I remember watching those bands just to the west of my house (Hatboro) and being dismayed at the near-miss. That was during the last few of these Miller-B hybrid systems. Maybe it's just coincidence. I don't know if there's a climatological or geographical reason for it to happen that way outside of elevation and/or distance from the ocean.

Appreciate the thoughts. I haven't lived here all my life, but in the time that I have, I can't remember Pittsburgh ever scoring significantly in a Miller B setup. For "significant" I consider 8" or more. I could certainly be wrong or misremembering, anyone care to fill in the gaps? Seems Pittsburgh metro, at least, does best in either big overrunning, Miller A events, or convoluted setups that aren't either traditional coastal "nor'easter" types. '93 was obviously unique and likely never equaled, and I don't know 1950 or 1890 well enough to define them.

I think the NAM is still too far out of its comfort range to be instructive. There's the rare occasion where it picks up on something, but usually you want to avoid the NAM until its under 48-60 hours. As you can witness on the NAM, the high pressure is bouncing all around, but I think it should be fairly locked in because of the overhead blocking (-NAO) and a ~50/50 low. The system that will eventually evolve into Wednesday's threat is just now hitting the west coast. The GFS at 12Z remains the most progressive of all the models. That solution creates such strong frontogenesis that you'll have winners under an INTENSE deform band with divergence elsewhere. The 12Z even screws much of Central PA that looked "safe" under other, previous solutions. Basically, you get a very narrow range of top snow depth instead of a widespread event. I'm not inclined to take the GFS at 100% right now. That said, the entire system is clearly vulnerable to disruptions, because it is early in the winter and we don't have a colder airmass established. The warm Atlantic waters are another wild card. It's also not near a true Miller A, which is what we really need to see for Pittsburgh to capitalize on these setups. Widespread overrunning snows. It's more like 75% Miller B and 25% Miller A. Those coastal transfers are tricky to time down. Too late and the mid-Atlantic misses out entirely.

Case in point: the EPS has a 60-80% chance of Pittsburgh metro seeing more than 3" of snow this week and a 30-40% chance of 6" or more with a mean through Friday of 5". Game isn't over.

There's still quite a few ensemble members of the GFS and Euro that swing a good bit of snow back across the entire state. Ridging out west, the digging s/w, or the high pressure can all influence the direction the low(s) move. The Canadian came right back to the same solution it had a couple days ago, however. It will be interesting to see if it remains consistent. It was curious to see the Euro come around to that same solution today while the GFS hints a bit more at it, but keeps the flow too progressive. Maybe the GFS is the new Dr. No. The GFS keeps it active through the period. The Euro looks like it transitions to a real +EPO signal around Christmas (not good). It will be curious to see what ends up winning out if the AO keeps tanking while we have less pacific ridging. Perhaps we get an "artificial" -EPO due to a shifting or splicing of the PV.

The GFS also has Euro support (or is it vice versa?). Should be an interesting week ahead. The 0Z Canadian also had a system in this same timeframe, but the surface low tracked right over Philly. We'll see if the 12Z suite alters any details.

Well, it's been quiet here. Rightfully so, I guess, because there isn't much going on in the interim. Long-term, the signals keep flipping back and forth, but a -NAO appears relatively likely. Perhaps in 48 hours we're back to a NINA look, I don't know, but it has been quite a while since we've seen any kind of true, even transient, -NAO block during the winter months. The last -NAO for December I can find is 2009 (we remember what happened then). From a "winter" perspective, 2010 was it, with negative values quite literally the entire year and into 2011. Columns after the year run from January to December. That was the end of our blocking winters. Since then, the only real blocking has come during the summer or fall months, and all winters have averaged well above positive. "A decade under the influence." This might all be for naught, but I figured since there isn't much going on, I'd try my hand at some analysis for the upcoming period, which looks like it could be surprisingly active. First is the Canadian for next Tuesday. After this weekend, which pretty much all models are visualizing a cutter, we head into next week that really looks more interesting. We see it is holding a rather large, singular piece of energy back into the southwest. This is usually a good sign for a big NE snowstorm. There's ridging along the west coast and, although the flow looks transient ahead, I think this is more a symptom of this s/w digging. This is the GFS at the same time: The GFS is much less organized. It is stringing out some of the same energy all over, while also holding a bigger piece back off the coast. It comes crashing in a few frames later: That's the remaining piece of energy coming into Seattle. Still, it keeps it separate and moves some smaller bits out ahead into the great lakes area. Very northern-stream dominant. At the same time, the Canadian keeps bowling: Now it's a closed-off ULL. Pretty much concentrating all the energy right there near the four corners. The Canadian then digs this all southward into the Gulf, dragging and phasing the northern stream energies into one significant trough. On the GFS, however: The American model keeps the streams separate. It has these two distinctive blocs, but it also has a piece of NS energy holding up in Canada. Basically, instead of how the Canadian crashes all this energy together into a cohesive system, the GFS times them apart and prevents things from ever morphing. It's a strung-out mess. See the final GFS look here: As I just said, the GFS is strung-out and tries to dig some of the energy south, but the northern stream is still keeping other bits of energy off the same conveyer belt. Eventually it does form a Miller A (of sorts) on its own, but because of the separation here, the cold air is also pulled away and the trailing system (seen here digging into TX) has no cold air to work with. What we get is a rain storm. Meanwhile, the Canadian: Puts a nice closed-off bowling ball right over the Jersey shore. Even the Canadian setup is kind of off-kilter - resulting in a double-barrel low of sorts - but it brings a lot of snow into the Mid-Atlantic. It isn't a great look for Pittsburgh because of the confluence, but if shifted just a little to the northwest, Pittsburgh would be in the game for the bigger totals. Meanwhile, this verbatim would jackpot areas around D.C., Philly, and New York, almost 11 years to the day after a very similar system did the same. The remaining question is - which of these models has support? As of the 12Z runs today, it looks like every model is kind of doing its own thing. There isn't a ton of agreement on the evolution of these pieces of energy. To be expected at this range and with this many moving pieces. Clearly it will take some time to lock in the results, but at least for now, we do have a period to watch upcoming. Whether anything results is pure speculation. Given how this winter was expected to develop as a STRONG NINA (now backing off that idea considerably, to the point where the Nina may be collapsing imminently), I think this is about as positive as we could have hoped.

I can't necessarily speak for the Philly area, but I know in the western part of the state we have had two Top-10 coldest Novembers in the last three years ('18 - 8th, and '19 - 10th). For the actual winter months, I'll have to go out further than Top-10. The most recent notable record is February 2015 (second coldest ever), followed by January 2003 (11th) and December 2010 (12th) . February of '07 and '14, December of '00 and '05, and January months of '03, '04, and '09 are all contemporary records. However, all of those months are way down the lists in the teens or even twenties. March is barely represented. 2014 is it on the "top" list, and it is way down there (19th). In the Top-18 for cold March, the most recent is March of 1984. Seems March hasn't been all that cold in a long time. Daily and yearly records are harder to find. 1994 is the most recent cold day. Seems the 80s were kinda cold, overall. If we consider certain elements to be cyclical, maybe things will shift back the other direction eventually. Or maybe climate change will prevent that and things are forever different and unpredictable. That analysis goes above my pay grade. TL;DR: One Top-10 in the last 20+ years: February of 2015 (for actual winter, anyway).

The CMC quickly abandoned this look, waiting until the storm was off Maine's coast to wrap it up. The Euro shows a prototypical Miller A path from the Gulf, through Alabama/Georgia, and then setting it up off the Delmarva. The GFS has a very similar setup, a bit more convoluted, but there's no "snow side" or cold sector. It's all rain and shears off south of our area. Lack of cold air definitely is one concern. Seems this one might be an inland NE special. Really, I think 2010 was the last of this kind of storm. Pittsburgh is still long overdue for an 8"+ event.