psuhoffman

The signal on the EPS is about as strong as you get for a possible snowstorm at that range...even better than it was for this current threat when it was at that range. If we didn't have a warning level event 24 hours away we would be blowing up this thread right now.

There will be a sharp cutoff once you get north of where the banding close to the northern periphery sets up...but I think it will be like 20 miles further north that CTP does...which isn't that much of a difference. NWS thinks that sharp cutoff will be right near the PA line, I think it will be 20 miles into southern PA. I would include the southern most counties in PA but that's really only a one county difference which is nothing and a noise level error, but of course the people living in that one county don't want to hear that!

My thinking right now... I think the globals are underestimating QPF a little bit in the H7 FGEN driven banding near the northern edge. This happens a lot. It's not 100% but I just have a hunch...1000mb lows that make it into central KY on a NE trajectory and no arctic high on top just don't tend to dry out THAT much. Yes it will hit a wall of confluence but often that even enhances the banding on the norther fringe where the moisture transport hits that wall and you get some nice banding. But I admit I am going out on a limb and riding my gut here. But it doesn't take a huge adjustment from where the GFS and Euro are now, shift the northern edge of the precip 30 miles north and beef up the qpf by about .1-.2 in that northern periphery where the h7 forcing is located (and models often under estimate), add in high ratios and you get this. The other snowfall max is to our south with the h85 driven forcing.

I just got home so I figured I would put out a map now that I have access to a computer. I will probably have to adjust this some tomorrow but this is my thinking right now.

By hour 12 you can already tell where the hrrr is heading. If you pull up the rgem for instance it’s night and day apart by hour 12. That doesn’t mean it’s right but if it fails it’s not because it’s out of range. The divergence happens in the first 12 hours.

I feel pretty good about this...

Fair enough but that means you own that snjokoma poster that shows up to junk up the obs threads every time!

I types that up just before the run hit. The euro did almost exactly what I hoped and now looks really close to what I envisioned. I’d maybe expect another slight beefing up of the QPF along the PA line area but not much. Maybe another .1 qpf. But that’s another 1-2” which gets that area into that 5-8” range I’ve been thinking all along.

The northern MD crew doesn’t need it to trend north. There is a legit path to a full forum win here. If that h7 driven band on the northern fringe beefs up just a bit as it often does at the last minute, and the h85 driven band stays where it is, we could get a full forum EZF to the Pa line warning event.

I didn’t think I’ve been negative.

It’s late and been a long week

But more realistic I want it to hold its 18z run but beef up the qpf in that northern fringe band like .1-.2 qpf.

Omg it’s better than the Hrrr and RAp

Because we needed too much. They usually trend some, 30 miles maybe and more importantly they usually under do the qpf along the northern fringe so between a small shift north and a 25% increase in qpf suddenly a place expecting 3” gets 6” or 8”. But if we need some 75 mile shift the last 24 hours that’s unlikely. I don’t mind sitting around the .3 qpf area going into the final 24 hours. But being on the outside looking in or on the fringe of any precip at all usually isn’t good. Tonight’s euro is big. It’s been bouncing a bit. If it holds or improves on the 18z run I feel good about higher totals all the way to PA. If it looks like the icon or rgem I might start to worry some.

Didn’t Jan 2022 shift north enough to get DC good when it looked like it might fringe them?

No that storm in 2019 was one of the only examples that didn’t trend north and I got less than I expected.

I’m just using the hrrr and rgem here but you can sub the NAMs FV3 and GFS for the hrrr and the uk and somewhat euro (but it’s coming around some) and it’s the same. This is just 18 hours from now… The error is happening right now as the wave amplifies in the plains after ejecting from the Rockies. By 18 hours the camps have diverged and it’s just dominoes falling from there. The Hrrr, FV3, GFS, NAMs are amplifying the wave significantly more in the next 12-18 hours. That’s the difference.

By 18 hours it’s already diverged significantly from the “south” camp like the uk and rgem. Someone’s gonna cave very soon because the error is very early on.

The euro and gfs have been slowly converging and now aren’t really that far apart. My guess is they split what little difference there is in the next run or two. Then I don’t expect any appreciable change in where both locate the two main bands here but I do expect the typical adjustment where the h7 associated northern band over performs due to a combo of models typically under estimating qpf a bit and the high ratios with that band causing that small qpf error to be exacerbated wrt snow accumulations. My guess is the final 24 hours will be about pinning down those details. Exactly how much of any sleet mixes in the southern max, how much qpf beefs up in the northern one, and are there any unfortunate bust zones in between the two bands.

For those near the PA MD line that stuff south of DC associated with the h85 forcing isn’t ours. You can see the euro hinting at the duel banding also. It has the h7 associated band right over us. Exactly where I expect it given the synoptic setup. It’s just weak sauce as of now. But that almost always ends up producing and often ends up the better band for accumulations. Get that band over us and I’ll take my chances. We don’t really need the euro to shift north. We just need it to increase the qpf associated with that band some. Not a ton. Right now we’re around .3 qpf. Get that to like .45 and with the high ratios we get up here and under that type of banding and that’s 5-8”. A very slight adjustment and it’s what I’ve expected given the setup.

About all these “how are the models so far apart” posts. Just a reminder we do this EVERY time we have a wave that isn’t some wound up sub 990 hecs monster with a 250 mile wide jack zone. NWP has become absolutely amazing (imo) at picking up on the general synoptic ideas on a 10,000 foot view at ranges unfathomable 20 years ago. I was on a plane a week ago looking at this same threat and we already had the general idea of a wave into the TN or Ohio Valley with gulf moisture overrunning the cold. Look at this from 180 hours! By 150 hours even the hiccup runs were gone and all 3 major ensemble systems were showing the same general idea. We didn’t even discuss or look at anything past 100 hours when I used to attend some of the guidance discussions at Penn State 25 years ago! Most of the models didn’t even go past 120 or 144 and we didn’t take anything seriously until it was within NGM and ETA range which was 48. Now we’re having serious discussions about stuff 7 days away! But NWP cannot pin down the exact location of meso banding or the location of the edges of these bands to the mile. Guidance even at 24-36 hours will shift these things around 40-50 miles run to run and every storm we say the same things “omg how are they doing this” It’s just the limitations of our current science and it doesn’t matter if it’s a large expansive qpf field and you're in the center but in events with a smaller jack zone or for those near the edges this is still all standard procedure.

Wrt NAM, I often toss is, it’s way to jumpy and prone to wild tangents. But it does sometimes pick up on a mid level warm push that can lead to a north shift on this type of setup and it’s not often this locked in and when it is sometimes it’s onto something. But not always.

I’ll give the storm another “reminder” as I fly through in a few minutes. Whichever camp is going to cave (or more likely a compromise) it has to happen soon because the differences that lead to the south v north results occur in only about 24 hours and the rest is just the dominoes falling after that.