psuhoffman

If we want to grasp at real straws this pattern isn’t that bad for going into March. Yea the pac is puke but in March it’s less important actually. This is very close. Get slightly more NAO and suddenly systems will cut west to east under the block. You can see the eastern ridge is mostly flat already there. It’s a typical error from a good look.

Last year was a nino. What’s it matter?

@leesburg 04 just for you. Last 24 hours for Thursday/Friday. Look at that 2” line retreat!!!

Lol. Nothings ever over. Crazy stuff happens. But I hate when people grasp for straws for false comfort. The 12z euro was way worse then 0z. Maybe not on the clown maps but in every way that actually matters.

@Ji sneakily euro trended colder for Tuesday the last 24 hours. One or two more bumps and we sneak into some snow. So you know what that means. Bet everything you have on a north trend next run!!!!

Euro trended colder Tuesday though. Isn’t that far from snow/sleet up here. But the euro has been running too far south all winter so having the model with the biggest south bias lately as the only one that’s even close isn’t likely a winning formula.

@Ji I was finally able to look at the run. It SUCKED. I didn’t look at the precip or pretty colors. None of those details matter at 140 hours out. This matters. This was 0z (which was slightly worse then the runs before even) That was a hopeful progression. That looked like a possible frozen event here. This is 12z thats puke. That looks like a big ass rain event. I didn’t look at the precip. I don’t care what it shows. Qpf and precip type at range is incredibly inaccurate and will adjust to climo for storm progression 90% of the time. That 12z progression was a major step towards rain from 0z regardless of what the pretty clown colors say.

That sounds like what I was afraid it showed. I know some might see some blue over them to start and celebrate but I don’t buy medium range guidance that shows that with a cutting system and the mid level flow is all from the south. Imo that’s unlikely to work out. I want to see signs the wave is suppressed south of us. That would be meaningful.

Thursday was never shown as a pure snowstorm and you know that. Euro shows snow to wintry mix to rain About 36 hours ago the euro was mostly a snow to dryslot event for us and kept the snow pack destroying warm sector to our southeast. There was no rain. It’s trended warmer every run since. The eps also. Had the max snow just NW of 95 a couple days ago but shifted up into central pa last night. I’m not that interested in a minor front end snowfall that’s immediately washed away by 45 degree rain. Those storms are depressing. I want to see signs the storm stays southeast of us. Also...in case you haven’t noticed models at range are always way too optimistic on front end frozen with a west track system. It can work but it’s rare. Models show that a lot at range but it typically is a mirage and the mid level warm layer comes blasting in before the heavy precip arrives. Without a suppressed flow it hard to prevent that. So I was asking doc the euro make a real trend towards a meaningful frozen event or did it continue trending towards a minor front end frozen before the main storm cuts way west and we get rain?

Jokes aside I can’t really look right now but I care more about the trend then verbatim what it shows. The last few runs went from a snowstorm to some snow to mix to rain for us. Did that reverse or is it still trending warmer?

Is it close to being close to being close to snow? Or is it only close to close to close to close to maybe kinda being close?

The problem with your "BIG SNOWSTORMS" correlation is that a 20" or even a 12" snowstorm in Baltimore is too rare an occurrence to be able to find those kinds of correlations. There aren't enough of them to see meaningful trends. The one meaningful trend is they mostly happen during a moderate or stronger Nino with a -NAO. That is a meaningful trend. If we lower the bar to 10" which is still a VERY significant snowfall and rare in Baltimore...the randomness is apparent. From 1958 to 1967 Baltimore had 7 such 10" snowstorms. In the next 10 years they had NONE. How is there a trend there? Baltimore had 3 of them in 1987 and only 1 the rest of that decade! They had 2 in 1996 and only 1 the rest of that decade! More recently Baltimore had 3 in one year in 2010 and none in the 3 years before and after that then 2 in 3 years from 2014-2016 and none since. It is totally random. You have to manipulate the data into very specific and rare and arbitrary numbers like "exactly this many inches at this exact location" to find correlations and those correlations are meaningless because snowfall is too fluky locally for such a specific thing to have meaning. And 10" is still honestly too rare an event to get meaningful data. If we do what Wes did for DC and lower the bar to 8" at Baltimore, which is still a very rare event that does not happen most winters, we get enough snowfalls to really see "trends". But there are none. in 10 years from 1958 to 1967 Baltimore had 17 8" snowstorms The next 10 years they had just 1. THen in 1978/79 they had 2 and the next 7 only 1. 1987/88 they had 4 then the next 7 only 1. 1996 they had 3 then the next 3 years none. From 2000 to 2016 there were 9 spread out pretty well...but then NONE since. You have at times focused in on too narrow of a window to find a false signal. If you only looked at the 2000 to 2016 period there seemed to be some regularity to our snowfall but when you pull back you see that was just a random short term coincidence in the longer term randomness. I have run the numbers...there is no predictability of snowfall based on recent history of past years.

I am not wasting my time digging into it but the UK on Pivotal counts freezing rain as snow. So all it has to be is cold by a few degrees over the other guidance to make that map happen. If it does actually show SNOW then its a divergent solution. Either way unless some other guidance jumps on board its not worth close examination. Any one model all by itself is going to be wrong 90% of the time.

I don't know if you are just screwing around with Maestro or not...I sometimes suck at picking up on that stuff... but this is not an accurate representation. 2010 was scorching...truly epically hot summer. 1999 and 1988 were above average but nothing spectacular, 1987 was hotter then 1988 for instance. 1966 and 1977 were about average. and 1922, 33, 44, and 55 were below average summers.

Excellent model. You should totally hold your breath for that!

You have to show the causality. What about years that end in 1 affects the weather?

@CAPE @Bob Chill this would lead to a workable look in March. Our pac issues can more easily be overcome with some blocking in March with shorter wavelengths. That’s why Nina’s (even more hostile ones) in general lose their correlation to warm/lack of snow in March. It’s not a guarantee. They don’t all end up producing but enough did that it’s obviously not wise to assume it’s over. Btw the top pattern analog that’s been showing up over and over lately is early March 1956. That pattern did turn cold and snowy later that month. But people in here would probably still complain because it was a general 3-8” snowfall across our area and 2 feet up near NYC so I know how that would go. We wouldn’t enjoy our 5” because we would be envious of the 20” to our north.

Fringed

When there is no blocking (right now there isn’t) they are. They press the cold too much. This winter most of the time they were too far north with the boundary from long range. But DC had a few perfect track rain due to boundary temps and a few systems sheared out totally. And yes a couple did track slightly too far north but that wasn’t the primary problem until this week imo.

We are getting into range where the errors we need are becoming less likely. Not impossible but it’s becoming a long shot.

We got a nice storm on the front end as the tpv rotated in. You and me got 6” and the cities rain from a perfect track coastal lol.

We’re living in a simulation so

on a more specific level wave spacing didn’t go our way either. The wave that popped up Monday prevented the cold from pressing behind this weekends wave. It ends up being one long wave train from Saturday to Tuesday with the front stalled to our NW. very very different then if the high had come over the top behind the wave Saturday (when there was no Monday wave). Then the spacing is too great between Monday/Tuesday and the Thursday/Friday wave. With the TPV gone and the NAO gone we needed good wave spacing to work. It’s not like that was impossible or dire we had good cold around to our NW and we’ve seen that work with some luck but we didn’t get any. Sounds too familial.

Actually from 2 weeks away they were just off by about 100-150 miles. Just to our NW is going to get crushed this week by several waves of frozen. The guidance was too far SE with that boundary but the error was well within what we should expect from those ranges. Once it got within the magic day 5 where details start to show it became apparent we were on the wrong side. If we want to parse why the guidance from range didn’t retrograde and break the block as fast which had the TPV and boundary pressed further east. But it’s a minor error that’s going to happen all the time at that range. It could have shifted equally the other way and pulled what it did in Dec 2016 when a similar look showed 35” from day 7 and became cold and dry with weak waves suppressed south of us. Both are equally likely and equally minor errors for long range. The problem imo is us pretending we have the ability to pin down the exact track of synoptic systems from range. Truth is there is only typically a snow “win” zone of about 100 miles with any synoptic system but from day 7 there is a typical error of hundreds of miles. So under no pattern or setup should we ever be confident in a snowfall from that range.

We’re discussing numerology. What do you think?