psuhoffman

Don Sutherland said in his winter outlook that 2016-17 was a best case most snowy possible outcome and 2002 as one of his "less snowy" options. For the record DC had 3.2" in 2002 and 3.4" in 2017.

Why does the tone feel so down about last nights game with Washington fans? It's weird...the perspective from Philly is that Washington should feel good about where they are and they are going to be a legit team now v the joke they were for a long long time. They went up against an elite super bowl contender last night on the road and hung in the game for 3 quarters! There was nothing to be ashamed of with that outcome. They showed they are a legit solid team, probably have a good chance to make the playoffs this year, and with a few more additions and some progress from their QB could even become a super bowl contender very soon. I guess I have to wonder what the expectations were, did they actually think they were ready to win a road game v an elite team now?

I think we do see some periods similar. Some of my analogs features some colder periods. 2009 and 2013. But I doubt it locks in all winter like 2014. Also, most comparable patterns to 2014 don’t go as well to the same extreme. That’s another year where we outperformed the pattern some imo.

2022 but the QBO was opposite that year. Also pattern analysis would indicate we might have got lucky some that year. There was a rather small snow max wrt mean right over some parts of our area. We totally maxed out potential, ate all the meat off that bone, and it was still below avg across most of the area But the strength of the Nina and the QBO make that a bad analog anyways. 2018 was really before the PDO went super negative. We got a little unlucky that we got 3 relatively meh snowfall years 2017-2019 right before we entered a historically awful PDO cycle that slammed the door on our chances of a big snowfall winter for a while. The past few examples of similar PDO patterns were bookended by some blockbuster winters. This time we entered a bad phase already in a snow drought. That’s some bad luck maybe. But I could be wrong about how I identified analog years.

I've been super swamped with work and still am, I will try to do a full write up soon...but since I don't have time right now I figured I would at least share my top analogs. I set up the criteria a while ago and have been updating the data as it comes in month to month. This is probably the last update so these are it. Top analogs 1:2012-13 2:2001-02 3:2022-23 4:2008-09 5: 2020-21 Adjusted weighted predicted snowfall BWI: 5.9" IAD: 6.9" DCA: 3.7" Quick thoughts... I set up the parameters I will use to identify analogs and the methodology to weight snowfall months ahead of time to avoid any bias. After last years utter failure I did revamp my process some and did historical testing to verify the new methodology is sound. It was better but as is any long range predictive process it's not perfect. Mainly because using the data available in the fall is not always indicative since things can flip quickly sometimes. For example: if the PDO were to start going through a larger scale phase switch in the next month or two it would bring 2014 into play as a possible analog. I don't see that right now though. I hope I am as wrong this year as I was last!

It's possible, there are no analogs this extreme. You have to go back to the 1800s to find anything close to this level of -PDO. And I doubt we can even take much from what happened then given how much things have changed. Overcoming a hostile pacific was a lot easier in the significantly colder climate of the 1800s. More recently we had an extremely (not to this level though) -PDO heading into 2022 and that turned out "ok". I would take a repeat of that winter in heartbeat right now. It could be much much MUCH worse!

No, but it will probably stay mostly this way until the PDO comes back towards less ridiculous territory and there is no sign of that happening at all. If the PDO were to actually stay in the -3 area all winter I could see this being a no winter year. Not just no snow...but a winter that is just a prolonged October with highs in 60s many days.

My gut says the PDO (and there are various components to that) has been driving this bus recently and much of the forecast failures have been that other dominant predictors are not having much impacting in muting it. It's possible that having this off the charts PDO (as well as a pretty anomalous QBO) could lead to something extreme breaking our way. We're in uncharted territory. But personally I fear the "extreme" event it could lead to would be a ridiculously warm dry snowless winter that makes even some of the recent dreg seem like a 1960's winter. I was actually more optimistic a month ago than I am right now, and I wasn't very optimistic then! That said, I don't have a crystal ball and I hope you are right. And we always could get something like 2000, its actually one of the analogs right now, and it was also a warm dry fall, and for the most part the winter pattern was also god awful...but we got an amazing 7 day period and so it's remembered as a good winter. In reality we just got lucky during the one week all winter that had any chance at all of snowing. But that can happen any winter. We've had some years go the other way, where we probably should have done better than we did based on the predominant pattern but got unlucky. I'm starting to think our best chance is simply to hope we get lucky and fluke our way to some snow with one or two decent anomaly periods within an overall crap winter. I highly doubt this ends up a winter where we have weeks and weeks on end of a favorable pattern to work with. We have to hope we maximize the few opportunities we get imo. And I am not at all qualified to say if we will or not. Most of what I do is simply identify probabilities based on data.

If the PDO doesn’t chill some were in trouble. It’s damn near impossible to sustain an eastern trough with the PDO hanging out around -3.

None. There aren’t really any good enso and PDO matches. 1993 and 2017 are probably the best but neither was even close to this negative a PDO. 2020 is probably the best combo analog with a decent enso and PDO match but not great. I’m not sure how much that matters when the PDO is this negative. It’s off the charts. I think it’s driving the bus.

It’s not hard to luck your way close to climo when avg is only 14”. But I said it was a decent year. It was still below avg snow at all 3 official stations so let’s not pretend it was a good winter. It was just better than most lately but that’s a low bar considering we are in the worst snow drought in recorded history.

I am not sold this is a total dreg no snow year. There are some "decent" years within the set, but I think slightly below normal snowfall is the high end bar unless we just get freakishly lucky. A year like 2022 is probably the best case scenario here. Oddly, and I've mentioned this before, there are still some in the enso thread "rooting" for the collapse of the nina to save winter...when in reality the 3 snowiest winters in the set of 16 similar PDO years I identified were all Nina's! Our expected snowfall actually goes down according to my method if we end up enso neutral!

The high correlation is probably related to the degree to which the PDO has been negative. It's not just been negative, its been historically negative. This was my biggest failure in last winter's prediction. I did weight the PDO, but I found several -PDO years with other matching variables that ended up snowy. What I failed to properly weight was that all of those years were -PDO but significantly less so heading into winter. If we only look at deeply -PDO years, we see its VERY difficult to get a lot of snow in those regimes. A less -PDO has more variability and can even be good for snow in a -NAO regime. But if the PDO is deeply negative it is the dominant factor.

I think its probably easier for NWP to resolve dominant synoptic features like a category 4/5 hurricane or the superstorm in 93 than some of the discreet events significant to our snowfall in winter. A deep enough pressure system becomes a bully vs getting impacted significantly by other discreet variables.

The more I dig into this the more I think maybe its really simple. There have been 16 times since 1950 where we entered the cold season with a deeply -PDO regime, averaging below 1.5 over the previous season. Only once did that season end up above average snow at BWI, and that was 2000 which IMO was a total fluke. The longwave pattern that winter was garbage 90% of the time and we got lucky the one week we had any chance of snow all winter.

We’re not gonna make it

If we want hope we could always get lucky with something like 2000. It was a Nina but I think the same type thing could happen in a neutrals also because the snow we got was just pure luck. The pattern that winter was predominantly a dumpster fire shit the blinds pattern. There was only a 10 day period the whole winter with any hope at all of snow. And it wasn’t even that good. It was just ok. extremely east based -NAO, mediocre pacific and AO. It’s not awful but that doesn’t scream great chance of a big snow. And that was the best we got all winter. The rest was total no hope garbage. But we got a few waves during that window and they all hit the region and one was a blockbuster! And a positive bust no less! But it was just pure dumb luck. If we replay that same winter pattern out 100 times 99 end up below normal snow and 70 probably end up like 2020! There is a lot of luck involved. Even in the worst winters there will be a couple chances. Get lucky and they hit and we avoid a dumper fire no matter how bad the predominant winter pattern is.

I’m with you 100%. There are actually two predominant types of Nina’s. A more flat pacific ridge with a +AO and those are dumpster fires. But Nina’s with a more poleward ridge and a -AO are cold. We still struggle wrt snowfall due to the lack of STJ and NS dominant systems so we get miller Bd to death. But they’re way better than neutrals during -pdos tend to be.

There seems to be a correlation with enso neutral and +AO in strongly -PDO regimes. That might explain why they all end up so bad wrt snowfall. In some of the Nina years the pacific ridge actually extends into the AO domain and sometimes even links with the NAO. We still suffer from the lack of STJ so they haven’t been particularly snowy but they are close and “snowier” than the neural years.

Going back to 1950 there have been 7 enso neutral winters during a deeply -PDO regime. All 7 were well below average snowfall at BWI. 2020 was the most recent example. If we end up enso neutral it would reduce the expected snowfall for BWI from 11” to 7” using the method I’m experimenting with. Assuming all other factors stay the same.

Our last 2 neutral winters were awful.

2001 was a decent winter up here, but the 3 largest snowfalls that winter here were all very marginal. Two were rain to snow events and one was just wet snow the whole time. 2 of the 3 were mostly rain in DC/Baltimore. If you increase the temp at all...you can probably take away a significant % of that seasons snow up here.

Honestly I’ve posted that clip multiple times over the years when people ask for a prediction. It’s a running joke. I’m sorry if it came off as an insult, I didn’t intend it that way.

Are you ok? This doesn’t seem healthy. Whatever I did to you I’m sorry. I know we’ve had disagreements and debates but I don’t know you, it’s not personal, and I wish you no ill will.