psuhoffman

@Terpeast looking at this made me think about our conversation wrt upstream v downstream cause/effect of the SER. But the gulf/atl isn't really downstream. As any wave crosses the CONUS there will be some southerly flow ahead of it. We obviously need that flow to be suppressed to some extent by the NS or to have cold air in place as that flow gets started...but south of us is off the Florida coast! And as any wave ejects from the Rockies the southerly flow it initiates in the plains is off the gulf of mexico. If those areas are on fire...it just seems logical that the southerly flow will be able to press north more than it would have otherwise. Hence...we get a stronger SER than historic analogs indicate we should have. None of that is to say I think the effect of the pacific doesn't have a part here...yes a ridge in the central pacific favors attempts at a SER. But historically there were things that could mitigate that SER. Lately...NOTHING mitigates it. When there is a central pac ridge its just game over and the SER goes ape no matter what else (-4 AO/NAO even) is going on. I think the warmer SST's in the gulf and SE coast are part of that equation. That added warmth just make sit even harder to beat down the SER and perhaps means we can no longer expect to mitigate a hostile pacific through other means. That is of course problematic since it eliminates another way to get snow and would doom us to atrocious results in even more winters that previously. It would not mean it cannot still snow when we get a favorable pacific though. I do think a favorable pacific can still overwhelm and suppress the SER.

GOOD the AC is broke in my classroom and it does not sound like it will be fixed anytime soon.

The CFSs verification is barely above random chance once past it’s month 1 forecast. Frankly none of our seasonal guidance has much success past month 1! We just aren’t there yet. As for the SER, it did respond to a dying Nina, there is just a misconception that I’ve pointed out before wrt what a Nina heading to neutral means. There is the belief in some circles a dying Nina is good but there is no data to support that. Nina’s that enter neutral by March show no increase in snow in the mid atl v Nina’s that stay strong. Furthermore look at the h5 comp for all Nina’s in the last 30 years… there actually isn’t a strong SER signature. That’s another misconception that a Nina=SER. But look at the enso neutral years following a Nina. There’s the SER! This winter actually behaves more like a neutral following a Nina than an actual Nina. In the end the early dying Nina might have killed our winter more than had the Nina not faded.

@Maestrobjwa another issue with the way you use patterns is not normalizing the data to account for changing norms. This is when you try to use past snowfall results as predictive. Let me illustrate. From 1884-1970 Baltimore never went more than 5 years without a 20” snow season. But that was never predictive that it couldn’t. The first issue is looking at the data closer no one season showed a significant predictive link to the next. The fact it had never gone 6 consecutive years was simply the fact each year had a 58% chance Baltimore got 20” so the odds of getting a 42% result 6 straight times was very low. But once you got to 5 the odds were still 42% it would happen again. It just never had. Until it did. It was only a matter of time! But that fact is even more irrelevant now. Because the odds have changed. Since 1970 it’s happened 3 times! That’s because the odds of 20” in a season dropped from 58% between 1884-1970 to 23% the last 30 years. And running a linear regression shows it’s likely about 18% now! So the odds of going on long runs with less that 20” is increasing because the odds of getting 20” in any given season went from 58% to 18% making that older data even more irrelevant as a predictive factor. So now…Baltimore stands at the end of a 7th season without 20”. And it has never had 8. But that means nothing. With the lowering probability of snow each season it’s only a matter of time. It’s almost inevitable given the current probabilities that eventually Baltimore will get 8 years without 20”. And it’s likely to be next year since the data says in any given season it’s now less than a 20% chance of 20”. Before you jump off a ledge that data is blind to things like enso. Right now we are predicted to get a Nino which would significantly improve our odds. But that’s a completely independent variable. For example if the models are wrong and we end up with an enso neutral next winter then the data says it’s probably about 80% likely we don’t get 20” and go 8 years. Regardless of the fact it’s never happened yet in our records.

you really need to stop attributing predictability and causality to patterns within random chaos. Patterns happen within any string of numbers from random chance. 73 wasn’t really a super Nino so it’s questionable if your pattern is even legit. But a pattern without a logical causality or link is just random chance. There is no logical causality to that pattern you cited. It’s way more likely to just be random chance. Like flipping heads/tails every other flip for a run.

Isn't that pretty much what we should expect any random day in winter?

Everything you say is true, but I am comparing our snowfall now to our snowfall in the past. Not another location. All those issues existed then too but we still snowed more than lately. We never snowed as much as Philly or NY or Elkins WV, those limiting factors you mention were always a problem. But that’s a different issue.

true... but there is a reason for that. Yea a central/west based nino gives us the best chance to get the small minimal amount of cold that still exists at the mid latitudes to be over us (other then siberia which is sheltered from the impacts of the on fire global SSTs) but it does nothing to the larger scale problem that is the reason we are struggling overall, which is if 75% of the mid latitudes are above average temperatures...the odds are already stacked way against us before we even worry about any specific longwave pattern.

We used to be able to do ok in a -PDO. We've had some blockbuster -PDO seasons. 1962, 1966 and 1979 were all -PDO seasons with huge snowfall totals in our area. But notice...none of them are recently. The problem is that over the last 30 years or so a -PDO almost always leads to a horrible snow season.

But if that onslaught only happens once or twice a decade and the rest of the time is mostly snowless dreg horrible periods, it's not worth the investment of time to track every winter and get your hopes up every season. We have already lost several regulars who decided its just not worth tracking anymore and they simply pop in when there is snow eminent but the rest of the time they are gone. I am starting to think they are right...it just doesn't snow often enough anymore to be worth tracking all winter long anymore. The fact is we are wasting our time the vast majority of the time lately.

Scary thing is it was a crap winter for much the same reason...there were several other perfect track storms that winter...but it was just too warm.

No way to know, it’s cyclical but not at predictable intervals. It’s was negative almost 80% of the time from 1945-1980 though. It was then positive 70% from 1981-2016. It’s been mostly negative since.

@Terpeast my post above also highlights what I was saying, that I fear our snow climo has been degrading more than we know because it was masked by the last +PDO cycle. Before 1980 we had a lot of snowy -PDO seasons. Since 1980 a -PDO season has almost always been a complete disaster. But that trend was masked by the fact we were in a +PDO most of the time since then so it didn't matter as much. But if we go back into another period where we have a -PDO most of the time like from 1945-1980 what would that mean?

@Terpeast I wanted to illustrate my point regarding our exchange a few days ago.... yes the Aleutian ridge is the main problem...but my point is... how do we mitigate an Aleutian ridge to get snow in years where that is a permanent feature? We had a jet extension into the EPO domain with an extremely -AO/NAO twice this winter and both times it did absolutely nothing to mitigate the SER effect of the pacific patter. That used to be how we got snow when the pacific isn't in a favorable configuration so my question is...when we get a season where there is crap pacific like this year...how do we get snow? We can't afford to just chalk every season that features an Aleutian ridge off as a total and complete fail loss. An Aleutian ridge is a predominant permanent feature during a -PDO, and a -PDO cycle is a reality about half the time. People are acting like this -PDO we are in now means we just can't get snow...but my point is we've had -PDO cycles before and it did manage to snow. Yes slightly less than during +PDO cycles...but it was nothing like this last 7 years. The worst part is the previous -PDO cycles that had 7 year periods that were close to this bad there was a convergence of a -PDO and an extreme +AO. That is NOT the case now...yea we've had a hostile pacific but the high latitudes have actually been pretty favorable much of the time and yet its done us no good at all. Let me illustrate my point... this is the composite of 12 above normal snowfall seasons from 1949 to 1980. These made up actually the vast majority of our snowy seasons during that period. Look at the pacific! DC got above normal snow in all 12 of those seasons with the mean h5 above! So why are we acting like an Aleutian ridge means we have a SER from hell and no chance of snow? And here is my point...and why we can't afford to just accept that... if you remove those 12 seasons from that period, our snowfall mean for that whole period, 30 years, a whole climo period...becomes as dreadful as its been the last 7 years. We would only have had 6 above normal snowfall seasons in the whole 32 years if you remove all those seasons where it snowed a lot despite a bad pacific base state. I am not disagreeing with you that the reason for the SER is predominantly what is going on in the Pacific...but in the past there were ways to compensate and still give us snow even in that pattern. But lately nothing seems to work, when there is an Aleutian ridge we torch no matter what else is going on.

@Ji you’re right about what the biggest issue was but were talking about 17 years so it’s not a fluke. It’s hard to quantify this but it seems to me that we used to get a larger % of our snow from northern stream systems. Remember clippers lol.

Yes we got plenty of snow during that era but remember while we did "ok" during that period...our average actually was close to historical norms during the 2000-2016 period... the problem is we needed to do VERY GOOD because that was actually what should have been an extremely snowy period if you look at the predominant phases of the PDO and NAO during that time period. Look at the mean h5 northern hemisphere pattern. We had an extended period where the Pacific, high latitudes, and Atlantic patterns all were in a favorable phase at the same time! Yet all we did was average what was historically a normal amount of snow over that period. And places not far to our north did experience one of their snowiest periods, on par with the 1960s. Yet we did not. Unless you accept that there is a new normal and that Dulles averaging 23" (which they did from 2000-2016) is now extremely snowy for our area. Yet when I moved to northern VA in 1994 23" was what IAD's normal snowfall was. Since 2016 Dulles has averaged 10.9" of snow. Yes we are in a down cycle now, the high latitudes are still OK overall but the pacific has entered a long term hostile cycle. And guess what...we are also doing worse than past previous comparable periods where the pacific was equally hostile. My point is, in my opinion, we also were doing worse than we should have from 2000 to 2016 only it wasn't getting much attention because we were doing fine wrt snowfall...but the fact is we should have been doing way better than fine...that was the up period and all we got our of it was what should have been normal mid range snowfall. Now we are in a legit bad period and its really dreadfully awful even by bad cycle standards. So yes you are right many of those storms did produce a decent amount of snow. But we needed MORE than decent during that period if that was the "snowy" cycle to maintain what we used to consider "normal" climo here. We needed that storm in January 2011 to be 20" plus like it was to our north not 6-10" of slop. We needed some of those northern stream system to dig a little more. I can't quantify exactly which storms were caused by what...just saying that we had a 17 year period with an extremely favorable pattern in every way, and places to our north set record snowfall on par or even surpassing the 1960's period...and all we got out of it was what we considered "normal" snowfall. That seems like a problem to me.

Yes I meant Nina. I’m not strongly inclined to say the SER is a cause but I’m also not sure it’s wholly an enso effect. I’ve noticed a disconnect between the high latitude and mid latitude pattern lately. The mid latitude impacts could still be pac driven but it’s not just a typical Nina thing is all I’m saying.

Except a Nino does not = constant huge SER. Actually if you look at the mean of all ninos there is only a very minor SER signature. This isn’t as simple as just “Nino” imo.

You bring up a legit hypothesis and I can’t say for sure. But a counter point…it was a bit of both. There were some storms during that period where temps did limit snowfall. Several storms in 2001 that hit NYC did give DC appreciable precip but we’re too warm. Dec 2003, Feb 2006, Feb 2007, March 2007, Jan 2008, Jan 2011, Dec 2012, March 2013, all featured storms that would have been more significant in DC if it was a few degrees colder and that’s why NYC got way more snow. But yes there were a lot of NS systems that just tracked to our north in that period. But is that actually uncommon or did we used to simply hit on more NS systems back when the mean NS was further south decades ago? The “miller b” thing is overblown Imo. It’s not some weird thing. Or some travesty. We don’t get “jumped” it’s just that they track to our north and the snow west of us is upslope on the westerly flow hitting the 4500 ft mountains. A west flow here is downslope. But the actual synoptic precip from a pure NS wave simply goes north of us because the storm tracks too far north for our latitude 90% of the time. Pure NS waves were never the best here but when I was doing that case study if all warning events at BWI I noticed what seemed a lot more of them back in the 50s, 60s and 70s than recently. If we went from a 30% hit rate to a 15% that makes a huge difference. I don’t know though. That would be a headache to quantify.

@Heisy something I have been keeping an eye on, and it likely doesn't matter for snow now since its so late...but it would most definitely matter if this was still winter... over the last 2 weeks, and it happens again in a few days which is what kills that hail mary threat you were peeking at...even when the pac jet extends and gets the ridge into the EPO domain...even rolling it into western Canada at times...it can't do much to the SER. The energy just cuts off under it and dives into the SW anyways. At times the SER gets beat down as a system crashes over top if it, but it immediately reloads. That would NEVER work for us because its not the wave that beats down the SER that is a threat for us...it would be a wave AFTER. If nothing can squash it for more than a day or two there is no legit threat. Look at the evolution the next 7 days...even with the pac ridge rolling into Canada and a beautiful west NAO block...the SER can't be suppressed much. I was curious so I looked at the pattern analogs. And sure enough almost all of them have way less SER than the guidance shows now. Why would that be.... simple...because everything else about the pattern says there SHOULDNT BE A SER...so when the model looks for analogs the only way to get as strong SER would be to have things different in every other way and so it would be a way worse match overall than simply picking the analogs where everything else is similar but with less SER. I noted this a couple other times recently...where the analogs were also saying basically "why is there such a strong SER". I wish this would be more discussion...instead of simply chalking the SER up to the pac pattern...discuss why is the SER so much stronger than history says it should be at times...including the pac pattern into that equation.

Oh god this again. I wasted an hour of my life years ago proving that the results from one season had no statistically significant impact on the chances of snow the next winter. And it resulted in a bunch of “but that doesn’t feel right” replies. I gave up on that. Math is hard.

I remember playing this game back in 2020. You were sure it HAD to get better. You cited statistics like this. Then you got like 7” the next year, 10” last year and nothing this year. So…3 years later I ask. Has it REALLY gotten better?