psuhoffman

Look at that beautiful snowhole on the gfs

Right where we want it

I agree with your point about making predictions on unrealistic time leads. But I don’t get your NHC point. Their avg error at 5 days is over 150 miles. It’s that about what the error here was? How is this any different?

When this was a week out the N American H5 was pretty darn close to some of our biggest snows. In the last few days some of the finer details trended the wrong way. A kicker showed up to flatten the PNA. The ULL trended more positively tilted and the flow in front more suppressive. The SS wave isn’t amplified enough. But isnt it hindsight now to say “we should have seen those changes coming”. Now it’s already a good idea to bet against a HeCs bc they’re so rare. But it’s not like people were excited at one of those times the models spit out some ridiculous storm despite every h5 indicator being all wrong. We have a retrograding west based block. It liked like a nice pna ridge for a time before the kicker came along. 50/50 retreating and a ULL digging west of us. Most of the ingredients were actually there this time but the details went all wrong. I don’t think this was a case of blindly following models when they made no sense. I thought this window had potential before the models showed a storm bc the pattern was pretty good. We don’t get a -5stdv AO often!

Fits analogs. I said it either had to come north or be squashed because that’s what every comp did. I just thought it would be north.

Some of its bad luck. But if you look at Nina snow anomalies Baltimore is ground zero for the snow minimum wrt climo. Snow wrt climo increases sharply across the bay on the Delmarva and increases more gradually to the NW. And we’ve been in a Nina type base state for a while. Even in non Nina years. The issue with these Nina like cold wave patterns is when the waves amplify at all or phase some they track north of us. When they don’t amplify they get suppressed south. I don’t think that’s totally chance. I think we are in a dead zone for those waves between the more amplified ones and suppressed ones. That’s why I want a return of the pattern from 2000-2016 where we had a lot more amplified coastal storms each winter. We can definitely get snow in that look but it’s really hard to get a big snowstorm in that look. And it’s way too late to save this god awful disaster up here with anything short of a MECS to end the season, the way 2018 was similarly saved and went from a F to a B winter for me purely bc of a 14” snow in March! Similar winter in that we missed a couple storms to the SE and went into March at like 50% of climo! A couple 3” snows won’t change my opinion of this winter. I’m big game hunting now. Only a wall trophy storm saves this mess.

The analogs are going to be right but in the other way. I was looking for Richmond to Delmarva big snows. I found none of them with a similar h5. I wasn’t even considering it would go south and be a Va beach snow. When I pulled those up….there were some matches. You don’t see many in between. With a TPV there it either phases and comes up or doesn’t and gets squashed way south.

No better would be it comes north then keeps going and we rain

We start the comeback tonight

@Terpeast obviously we’ve talked about our degrading snow climo, but there is an oddity within the trends. It’s enso neutral winters that have been killing is. Enso neutral winters used to be pretty snowy. Actually if you go back far enough there wasn’t much difference between neutral and Nino wrt snowfall. Nino snowfall actually hasn’t declined at all, but it has become more variable with bigger highs (2010) and more duds (2024). La Niña snowfall hasn’t declined much either. It was always bad and remains bad. But enso neutral winters have joined Nina’s more and more and now over the last 20 years are just as bad! Losing enso neutral winters as “snowy” is the biggest part of our snowfall loss. Just to clarify enso neutral was never all snowy but it was half and half. Kinda a crap shoot. Lately it’s more 80/20 bad.

If next winter ends up enso neutral our fate is probably hanging on whether the PDO is actually flipping. +pdo neutral winters can be pretty good -pdo ones have been god awful lately!

2004 there was a storm like this that trended worse until there was no storm anywhere. Or maybe some snow in NC. I stopped following it once it got that bad.

Imo it’s not one thing. It’s a bunch of tiny things. Which is why the 30,000 ft view of the pattern looks great. You have to zoom in on a lot of small details to see the flaws. Death by 1000 paper cuts 1) the kicker, I thought this might limit the setup to a MECS level not hecs level event not destroy it and I don’t think it did, runs with a storm had the kicker. But it was one of the many paper cuts. 2) the TPV didn’t detach fully. There was still a bit too much play between the main TPV and the detached lobe which kept the flow more W-E under it rather than the detached lobe acting like a true ULL pinwheel in the flow. It made the ULL act more like just an extension of the PV. This alone could have been overcome with a strong enough SS wave… 3) not a strong enough SS wave. Weak sauce that needed more of all those other minor flaws not to work against it. If we had some el Nino STJ on roids wave coming across it would have bullied through all these minor flaws and throws a ton of moisture up over this cold air. That’s why we get so many more big snows in a Nino! 4) a duel wave structure with destructive interference between the waves. The weak STJ wave starts to amplify at the exact time the NS wave over top of it was trying. We actually wanted that. Sorry I’m never rooting for some low off thr southeast coast to phase capture and bomb due north. That’s worked out like almost never. We’ve had maybe 4 MECS storms that way. Ever. That ain’t it. We want a healthy wave up into the TN valley that transfers to eastern NC or VA capes. That northern wave gets starved of moisture inflow by the wave down by New Orleans starting to amplify under it. In the end they split the energy and neither becomes the wave we need. 5) the trough takes a little too long to go negative. 6) Jim Cantore booked tickets to DC. I’m not saying he is bad luck everywhere. But not here. Never again. We should assign someone to watch every airport and tackle his ass and say “no” not on our watch.

Agree with all this. But on a side note, I (and the rest of northern MD) really needs this predominant cycle of the last 8 years where almost all our snow comes from these progressive w-e boundary waves to end. It makes sense that some of our lowlands are doing way better wrt climo than here. It takes a lot less luck in those patterns to get to climo when you avg 17” than 39” A couple lucky hits and you’re there. Up here I’m not lucking my way to a 40” that way! I need the amplified coastal to come out of retirement!

I can’t believe this is gonna produce nothing for us. This is one of our worst wasted opportunities in a long long time. This must be how Matt Ryan felt.

JB threw in the towel

@WEATHER53 you have every right to feel the way you do. Really. I’m not gonna try to change your mind. You do you! But I have the opposite perspective. As complex and chaotic as the atmosphere is with so many variables and our limited ability to sample it at every level at every location…given those realities, I’m amazed we are able to model it as well as we do! It’s so complex. I don’t feel let down by the models this week. I feel let down by me. I started getting excited about this window before the models did. When they showed up to the party it just increased my confidence more. But it’s my own excitement and desire for a major event that let me get carried away. I knew a month ago my analogs and pattern recognition weren’t 100%. When I listed the years I pointed out a couple didn’t really produce much snow. Most did. And when the models started showing big storms I started to dream this was one of those years. No hecs I never bought that the pattern didn’t say that but I was thinking secs-MECS. But that was risky. Even good pattern fail a lot. And I should have known better. I let my emotions get the better of me. That’s on me not the models.

Who, who said that? Either someone said that or they didn’t and you’re just making this up. If someone said that who.

Who was saying it can’t snow at all anymore? Who? You say “those” Who is “those”. It’s been snowing. It snowed last year. It snowed in 2023, granted like 1” lol. We got snow in 2022. It’s been snowing every freaking year. Just less. Who is this mythical “those people” saying it can’t ever snow they you are discrediting. This sounds a lot like a scarecrow argument.

I don’t know. I’m very curious though. It’s possible this method could do better at reducing errors due to sampling. Possible. I don’t know enough.

But in the case of the gfs/nam/hrrr the supplier is US! They’re all funded with public dollars. There is no profit. We aren’t defending some corporate overlords. We aren’t taking the same hostility towards these sources because we don’t share your opinion that the shortcomings of the models are some nefarious intentional design. We think the people working on them are doing the best they can and our science just isn’t as advanced as we’d like yet.

someone posted that JB comment yesterday “the better h5 gets the worse the surface gets” it got me thinking last night. And it brought be back to what you said about the upper low being “dead”. You were right. Getting a TPV stuck under a block and having a lobe tear off into the upper Midwest isn’t that unusual. It’s how several of our big snows came about. But it remaining totally disconnected from the flow and acting like a kicker instead of flow pinwheel and amplifying force is what is unusual. The detached lobe is acting more like the actual TPV than a typical ULL after detaching from a TPV. We wanted it further away not closer in that case! But I do wonder if there is a tipping point at which if it does get far enough south it has to phase v act like a kicker? That might be our only hope since we can’t get a storm with a PV sitting that close to us and the ULL is acting more like a weak PV than an ULL. This might also be why I didn’t see any in between solutions. Either snows way way south or north. This just might be a rare enough scenario though that I’m not finding any examples because the first such instance since we started keeping reliable upper level data is about to happen.

I’m not going to be sucked into defending the most extreme crazy statements “on my side” or a debate. Maybe some took it way too far. I’ve not seen a lot of that. We’ve had really good discussions and by running regression studies it looks like we’ve lost about 20% of our snowfall over the last 100 years and the trend is continuing. That’s it. I’m depressed enough about that reality without listening to some nut that takes it too far and says it’s never gonna snow again.

I’m 99% sure the gfs was simply right because of its typical errors being right in the correct direction for once. But I do wonder if the AI is less prone to errors due to estimating conditions in data sparse regions.

I only said what I said because that guys post was kinda ridiculous and hostile right off the bat. You probably could have made the point in a better way and I wouldn't have been that way.