February Medium/Long Range Discussion Thread

[psuhoffman]

36 minutes ago, mitchnick said:

I was commenting on the trustworthiness of the MJO forecasts.

 

Valid point. But keep in mind that even though the trend towards more amplitude in the mjo wave hurts us now it also makes it even more likely it gets into 8 later and helps. It's trending stronger not weaker. 

[psuhoffman]

49 minutes ago, paulythegun said:

EPS has several MASSIVE hits, all in some form or another after the cold front associated with the Low of Sauron day 9-12. Some anafront action. Also some members with the trailing piece of energy out of the gulf after the aforementioned low.

One member has what looks like a bull's-eye of 30"+ all from one storm (day 9-12).

Sent from my Pixel using Tapatalk
 

 

 

 

Yea I saw that. The mean jumped up a decent amount but it's not from a lot of hits. It's from a weird number of massive snow hits. Almost all or nothing. Very few with some snow. It's 6+ or nothing. There must be a divergent camp and the means are muted enough to see how there could be a gefs like camp within the members. 

[Wonderdog]

14 minutes ago, psuhoffman said:

Yea I saw that. The mean jumped up a decent amount but it's not from a lot of hits. It's from a weird number of massive snow hits. Almost all or nothing. Very few with some snow. It's 6+ or nothing. There must be a divergent camp and the means are muted enough to see how there could be a gefs like camp within the members. 

This corresponds with what you have said for the last week that the pattern would be best after this weekend.

[Wonderdog]

LP in South Dakota at 12z now in Colorado at 18z at hour 96.

[CAPE]

EPS continues to have a rather conflicted look in the long range. Looks like a decent bet of a warm period early next week lasting a few days, then a cool down later in the week. No window with even a decent signal for a snow event that I can see over the next 10 days, despite the mean snowfall looking okay. Late in the run, who knows. Looks ambiguous as usual, so might as well go with persistence.

[Thanatos_I_Am]

Give me E13 and you can take next winter too

[Amped]

One we'll get one of those post frontal cutoffs to compete with our top KU storms.  It probably won't resolve until 36hrs out.   Until then, stay thirsty my friends.

[WinterWxLuvr]

The ns system is dying on the gfs.  Too bad the ss system is also on life support.

[DC2Winston]

Just want one potential storm to make it within like D4-5.  Even if it doesn't happen or in a lesser degree.

D7-8 is our glass (concrete) ceiling this winter. 

[NorthArlington101]

GEFS still interesting for Friday. I'm going to look at it closely before I make a definitive statement but the trend looks decent.

[NorthArlington101]

Mean improved. There are some better hits down south, which seems to be the theme this year. 

8/20 show .5" at DCA at 18z compared to 3/20 at 12z.

Think I'm allowed to post this image, let me know if I'm not...

 

[Maestrobjwa]

I have a gut feeling that if we get anything of any significance this season...it'll happen by President's Day week (surprise, surprise, lol It may be the almanacs going to my head, as well). That's the only thing keeping me interested in the LR and if anything appears to be changing!

[NorthArlington101]

Really big rain-to-snow signal on the 18z GFS somewhere days 8-10. Couple of solid "events" in there. 

[kurtstack]

If we can get our first -ao of the winter this might be just in time for a late save... I know another football lucy lol


Sent from my iPhone using Tapatalk

[psuhoffman]

14 minutes ago, NorthArlington101 said:

Really big rain-to-snow signal on the 18z GFS somewhere days 8-10. Couple of solid "events" in there. 

It's a step back from 12z. Honestly the gefs snow mean is the worst it's been in a long time this run so I'm not enthused. EPS had more of a signal. 

[87storms]

Looks like this is one of those winters where we're gonna need to rely on timing.  We certainly haven't wasted much cold because there hasn't been much to begin with and this is coming from someone who dislikes the cold.  All things considered it's not a surprising result.  Mostly mild and mostly dry.

[WinterWxLuvr]

1 minute ago, psuhoffman said:

It's a step back from 12z. Honestly the gefs snow mean is the worst it's been in a long time this run so I'm not enthused. EPS had more of a signal. 

If you look at the Fri/Sat deal, 18z is definitely a step up but not a big event even if it happens.  Then if you look ahead to Sunday, another marked improvement.  I didn't look beyond that.  I looked at the ind members, not the mean.

[psuhoffman]

4 minutes ago, WinterWxLuvr said:

If you look at the Fri/Sat deal, 18z is definitely a step up but not a big event even if it happens.  Then if you look ahead to Sunday, another marked improvement.  I didn't look beyond that.  I looked at the ind members, not the mean.

My reply was to the comment about the 8-10 day. The signal was worse that period.  Gefs for like 4 runs in a row is showing the pattern we want toward the end. Trough over the southeast with the nice banana of higher heights across Canada. It's not a very cold look but stormy with coastal chances. But we have been chasing it all year. If it starts moving closer and shows up on EPS and geps then maybe I'll buy that it's a result of the mjo and strat. 

[WinterWxLuvr]

6 hours ago, dtk said:

Nonsense.  You just have to have a better understanding of the tools, what they represent, and have a grasp on the practical predictability at those lead times.  I'll agree with you that people shouldn't be looking at the deterministic details at those time ranges, but models are skillful enough at sniffing out patterns and setups.  You should also look at guidance in the 6-14 day range in terms of something like a time-lagged ensemble instead of independent, deterministic forecasts each cycle every 6 hours.

I wasn't model bashing.  I don't do that.  In early Dec, the models predicted that about a half inch snow would happen here.  They also said the snow would split my county in half with the places about 10 miles to my south getting none.  And that is exactly what happened.  Amazing.  I've said this before and in fact you have commented on it.  It's amazing that some plastic, metal wiring, and electricity can tell me something like what I wrote above will happen and then it does.

But right now, for whatever reason, they throw out a look that 3 days later is remarkably different.  We've speculated on why in this thread.  You're right about one thing.  I don't have the expertise needed to make use of all this data.  About all I can do is watch for trends.  When I see a look like the first pic below, I can say hey, that looks pretty good.  What do I think 3 days later when the same ensemble, not op, has the look of the second pic?  How should it be interpreted?  This has been going on for some time now.  These pics are at 10 days and 7 days from now.  Now if you want to say there's inherent risk in looking at 10 day forecasts even if they are ensembles, then you're basically saying the same thing that I was saying.

[WinterWxLuvr]

12 minutes ago, psuhoffman said:

My reply was to the comment about the 8-10 day. The signal was worse that period.  Gefs for like 4 runs in a row is showing the pattern we want toward the end. Trough over the southeast with the nice banana of higher heights across Canada. It's not a very cold look but stormy with coastal chances. But we have been chasing it all year. If it starts moving closer and shows up on EPS and geps then maybe I'll buy that it's a result of the mjo and strat. 

I honestly think those last two are playing a huge role in what we are seeing in the 10+ day modeling.

[kurtstack]

I wonder if the 10+ models perform better in a -ao than a +ao. Has this been examined statistically? One consistent all winter has been the +AO


Sent from my iPhone using Tapatalk

[SomeguyfromTakomaPark]

58 minutes ago, WinterWxLuvr said:

If you look at the Fri/Sat deal, 18z is definitely a step up but not a big event even if it happens.  Then if you look ahead to Sunday, another marked improvement.  I didn't look beyond that.  I looked at the ind members, not the mean.

GEFS gives some hope for super bowl?

[BristowWx]

3 minutes ago, DCTeacherman said:

GEFS gives some hope for super bowl?

Unless the high is over New England, the NS weakens and the SS strengthens not seeing how we can pull this off.  12z yesterday was about the best look we could have hoped for.  It's been down hill since then as you know.  The SS at 500 is quite weak sauce.  Even as the NS is weak as well.  

[psuhoffman]

37 minutes ago, kurtstack said:

I wonder if the 10+ models perform better in a -ao than a +ao. Has this been examined statistically? One consistent all winter has been the +AO


Sent from my iPhone using Tapatalk

They perform better in stable blocking patterns because it slows everything down and it promotes stability in the location of certain features. So that would be a byproduct. They aren't going to do as well in a progressive pattern with vorts flying all over. That's why our snows in 2014 weren't long lead time threats but mostly trended better inside 5 days. But in 2010 or last year we saw hints of the storm 2 weeks out and pretty close at 10 days. 

[psuhoffman]

1 hour ago, WinterWxLuvr said:

I honestly think those last two are playing a huge role in what we are seeing in the 10+ day modeling.

It's the only reason I still have hope those two factors are real influences and they both have strong signals right now. That gives me some pause before just saying it's over.  But it's getting late. This next reshuffle is our last chance probably.  So we have to get it right. 

People have been saying how the models Lucy us all winter. And I think I sometimes write a paragraph of analysis of what's going on behind the scenes and some simple think "yea but it didn't snow" and while I want snow too I also like trying to learn and figure this out. 

There are a few observations I've made. The overall basic patterns have been fairly well advertised on the guidance. But there are two very important details they keep screwing up that hurts us in the end. The first is simply the extent of the cold. Even when we have a fairly impressive trough it's just not that cold. Cold has been pretty pathetic most of the winter except two very brief shots.  The baroclinoc boundary keeps shifting north as we get closer  

The other is the nao and AO. The models keep showing it and it is muted or non existent when it comes. While some then say the models are useless I'm more interested in the analysis. Obviously there are factors telling the guidance the nao should go negative. There is some driver in the pattern lobbying for that. But the model is seeing that. But there is something else that the models aren't giving enough weight that keeps overriding it. It could be the qbo. I'm honestly not sure. But I'm far more interested in that discussion then the "models suck" crap we're getting too much of lately. 

So it's not snowing. Sucks yea. But constant complaining won't help and there is still weather to analyze and figure out even if it's not snow. 

[WinterWxLuvr]

21 minutes ago, psuhoffman said:

It's the only reason I still have hope those two factors are real influences and they both have strong signals right now. That gives me some pause before just saying it's over.  But it's getting late. This next reshuffle is our last chance probably.  So we have to get it right. 

People have been saying how the models Lucy us all winter. And I think I sometimes write a paragraph of analysis of what's going on behind the scenes and some simple think "yea but it didn't snow" and while I want snow too I also like trying to learn and figure this out. 

There are a few observations I've made. The overall basic patterns have been fairly well advertised on the guidance. But there are two very important details they keep screwing up that hurts us in the end. The first is simply the extent of the cold. Even when we have a fairly impressive trough it's just not that cold. Cold has been pretty pathetic most of the winter except two very brief shots.  The baroclinoc boundary keeps shifting north as we get closer  

The other is the nao and AO. The models keep showing it and it is muted or non existent when it comes. While some then say the models are useless I'm more interested in the analysis. Obviously there are factors telling the guidance the nao should go negative. There is some driver in the pattern lobbying for that. But the model is seeing that. But there is something else that the models aren't giving enough weight that keeps overriding it. It could be the qbo. I'm honestly not sure. But I'm far more interested in that discussion then the "models suck" crap we're getting too much of lately. 

So it's not snowing. Sucks yea. But constant complaining won't help and there is still weather to analyze and figure out even if it's not snow. 

I really like that next to last paragraph.  Fascinating to think about what that overriding factor is and why the models don't recognize it as an overriding factor.  Is it short range MJO forecasts that go awry and then cause changes in the longer range pattern? Strat? QBO?  

Its really kind of cool.

 

[psuhoffman]

18 minutes ago, WinterWxLuvr said:

I really like that next to last paragraph.  Fascinating to think about what that overriding factor is and why the models don't recognize it as an overriding factor.  Is it short range MJO forecasts that go awry and then cause changes in the longer range pattern? Strat? QBO?  

Its really kind of cool.

 

That's the stuff that hooks me. The good mystery to solve. It's why part of me hopes they never get too good at model guidance such that we remove the element of surprise and mystery.  

I think the mjo can be ruled out. It's been stuck in the COD 95% of the winter. I think it's playing havoc now but it wasn't before. It's been modeled as a non factor and has been a non factor. 

I doubt it's the strat. The guidance has been pretty good both on showing the consolidation and recovery of the PV in December and the warming episode that started recently and looks to really go strong. So while the strat is a huge influence it's been modeled well. 

The qbo is the obvious one that sticks out. There is a correlation between west qbo and a positive AO. Then if you look at cold enso states and low solar apparently that correlation is even stronger. Give credit to Joe dealeo for pointing that one out. (Side note. Funny how Joe d pointed out a strong correlation between low solar cold enso west qbo years and very positive AO with eastern ridge yet his colleague jb went cold and snowy anyways. Funny huh). Then factor in this is a record qbo.  I think it's very possible the models can't correctly see the impacts of such an uncharted phenomenon.

But this is just speculation. It could be some other influence. The sst. Or a weird combination of multiple factors in a delicate balance that the models have little hope of calculating properly. But it interests me a lot and would love more discussion on that. 

ETA: I get that probably 1/3 people who read this are going to go "who cares it's not snowing" lol

[WinterWxLuvr]

10 minutes ago, psuhoffman said:

That's the stuff that hooks me. The good mystery to solve. It's why part of me hopes they never get too good at model guidance such that we remove the element of surprise and mystery.  

I think the mjo can be ruled out. It's been stuck in the COD 95% of the winter. I think it's playing havoc now but it wasn't before. It's been modeled as a non factor and has been a non factor. 

I doubt it's the strat. The guidance has been pretty good both on showing the consolidation and recovery of the PV in December and the warming episode that started recently and looks to really go strong. So while the strat is a huge influence it's been modeled well. 

The qbo is the obvious one that sticks out. There is a correlation between west qbo and a positive AO. Then if you look at cold enso states and low solar apparently that correlation is even stronger. Give credit to Joe dealeo for pointing that one out. (Side note. Funny how Joe d pointed out a strong correlation between low solar cold enso west qbo years and very positive AO with eastern ridge yet his colleague jb went cold and snowy anyways. Funny huh). Then factor in this is a record qbo.  I think it's very possible the models can't correctly see the impacts of such an uncharted phenomenon.

But this is just speculation. It could be some other influence. The sst. Or a weird combination of multiple factors in a delicate balance that the models have little hope of calculating properly. But it interests me a lot and would love more discussion on that. 

ETA: I get that probably 1/3 people who read this are going to go "who cares it's not snowing" lol

I find it interesting.  You won't get a who cares from me.  If I'm getting continuously kicked in the nuts I like to know why.

[kurtstack]

It's the only reason I still have hope those two factors are real influences and they both have strong signals right now. That gives me some pause before just saying it's over.  But it's getting late. This next reshuffle is our last chance probably.  So we have to get it right. 

People have been saying how the models Lucy us all winter. And I think I sometimes write a paragraph of analysis of what's going on behind the scenes and some simple think "yea but it didn't snow" and while I want snow too I also like trying to learn and figure this out. 

There are a few observations I've made. The overall basic patterns have been fairly well advertised on the guidance. But there are two very important details they keep screwing up that hurts us in the end. The first is simply the extent of the cold. Even when we have a fairly impressive trough it's just not that cold. Cold has been pretty pathetic most of the winter except two very brief shots.  The baroclinoc boundary keeps shifting north as we get closer  

The other is the nao and AO. The models keep showing it and it is muted or non existent when it comes. While some then say the models are useless I'm more interested in the analysis. Obviously there are factors telling the guidance the nao should go negative. There is some driver in the pattern lobbying for that. But the model is seeing that. But there is something else that the models aren't giving enough weight that keeps overriding it. It could be the qbo. I'm honestly not sure. But I'm far more interested in that discussion then the "models suck" crap we're getting too much of lately. 

So it's not snowing. Sucks yea. But constant complaining won't help and there is still weather to analyze and figure out even if it's not snow. 

To be honest we are probably trying to understand something that we may not possibly be able to comprehend. We use models that digest millions of data points and progress all of those pieces of data forward in time based on complex multi-variable thermodynamic and fluid dynamic equations over a spherical three dimesnional surface. You say there must be a driver causing the models to predict a -ao errantly, but perhaps there are thousands of drivers across the globe all influencing it to some extent. So i dont think it gives it justice to point to one index as the driver of any given errors in the model output but rather believe it is a complex combination of multiple variables across the globe that are affecting the forecast indices for better or worse.


Sent from my iPhone using Tapatalk

[BristowWx]

6 minutes ago, psuhoffman said:

That's the stuff that hooks me. The good mystery to solve. It's why part of me hopes they never get too good at model guidance such that we remove the element of surprise and mystery.  

I think the mjo can be ruled out. It's been stuck in the COD 95% of the winter. I think it's playing havoc now but it wasn't before. It's been modeled as a non factor and has been a non factor. 

I doubt it's the strat. The guidance has been pretty good both on showing the consolidation and recovery of the PV in December and the warming episode that started recently and looks to really go strong. So while the strat is a huge influence it's been modeled well. 

The qbo is the obvious one that sticks out. There is a correlation between west qbo and a positive AO. Then if you look at cold enso states and low solar apparently that correlation is even stronger. Give credit to Joe dealeo for pointing that one out. (Side note. Funny how Joe d pointed out a strong correlation between low solar cold enso west qbo years and very positive AO with eastern ridge yet his colleague jb went cold and snowy anyways. Funny huh). Then factor in this is a record qbo.  I think it's very possible the models can't correctly see the impacts of such an uncharted phenomenon.

But this is just speculation. It could be some other influence. The sst. Or a weird combination of multiple factors in a delicate balance that the models have little hope of calculating properly. But it interests me a lot and would love more discussion on that. 

ETA: I get that probably 1/3 people who read this are going to go "who cares it's not snowing" lol

I sometimes wonder if JB puts out the same hype-ish forecasts to his core business traders.  It's one thing to tell a bunch of weenies it's going to be cold and snowy and then bust but another to tell those who pay for accuracy because they have skin in the weather (oil and gas futures, retailers etc) the same thing and be wrong.  My guess is no.  Those forecasts are likely his real thoughts and probably far more steeped in the facts. He might be a wishful weenie at heart but business is business and cash rules.Those type clients are not interested in snow and cold because they are winter weather enthusiasts.   Just my 2 cents.