December 2023

[Typhoon Tip]

10 minutes ago, dendrite said:

We all fall in love with the fancy hires QPF and clowns now at d5-7.

The models are really damn impressive. As the decades go by our expectations change and we demand more from them and want specific deets many days out. I think many of us forget what they were like in 2010 or 2000. They've missed plenty of storms in the past. Heck I remember when the euro was gaining its king status back when it only ran once per day, most of us did not get to see QPF, and we basically forecasted using H5-7-85 and 2m/1000mb. We were toggling coarse SLP contours on unisys.

Just imagine a day when all that technological infrastructure is rendered meaningless by the advent of the 'Weather Actualization Grid'   ... something like a HARP starts guiding the atmosphere, such that modeling the way we know it is instantly outmoded. 

Dystopia -

[Ginx snewx]

11 minutes ago, dendrite said:

We all fall in love with the fancy hires QPF and clowns now at d5-7.

The models are really damn impressive. As the decades go by our expectations change and we demand more from them and want specific deets many days out. I think many of us forget what they were like in 2010 or 2000. They've missed plenty of storms in the past. Heck I remember when the euro was gaining its king status back when it only ran once per day, most of us did not get to see QPF, and we basically forecasted using H5-7-85 and 2m/1000mb. We were toggling coarse SLP contours on unisys.

You know I have been obsessed with this stuff for 60 years. Even with old modeling I feel I did better forecasting than now. It may be they are too fine tuned and the damn butterflies screw up output. People can point to scores but here on the ground we know intuitively something has changed for the worse op wise. Specifically talking inside 96 hrs.

[Ginx snewx]

6 minutes ago, Ginx snewx said:

Congrats Dendrite on the foot plus. Huge hit for CNNE.

That's an impressive Euro storm. Ski areas gonna go belly to belly this year.

[TauntonBlizzard2013]

This result shouldn’t be surprising given the calendar. 

[ORH_wxman]

Euro was trying for 12/6 too but it just doesn’t have enough space. We need 12/4 to lift out quicker. 

[STILL N OF PIKE]

8 minutes ago, Ginx snewx said:

That's an impressive Euro storm. Ski areas gonna go belly to belly this year.

I was hoping this would at least hit them 

this was never more than a long shot for NE mass to me 

next 24 -36 hours hopefully there is some model consistency 

[ORH_wxman]

14 minutes ago, Lava Rock said:

It's quite obvious to me models were better back then, but why don't they either go back to them or tweak the ones they have to make them like they were? Would they deny that the models are worse now than they used to be? What benefit is there to providing a frcst based on models that goes back and forth sometimes within <48hr?

The models were a lot worse back then. I forecasted with them for years…we knew some of the biases better, but their skill in every single metric was worse.

I think many forget that we used to weenietag people for getting too excited about a 144 hour storm threat…now people actually take them seriously. And there’s a good reason for that….models actually sniff out storms at 144 hours way better than they did in 2008 or 2009. Hell, I remember we tracked the 12/17/20 storm from like 200+ hours out, lol…it was consistently there on guidance. That type of stuff almost never happened back then. 

[dryslot]

4 minutes ago, ORH_wxman said:

Euro was trying for 12/6 too but it just doesn’t have enough space. We need 12/4 to lift out quicker. 

Yes, Spacing is a problem right now, It needs more room to get the wave on the 6th to amplify, I don't think its impossible, That air mass behind the 4th would support snow into SNE.

[CoastalWx]

I'm 

4 minutes ago, ORH_wxman said:

The models were a lot worse back then. I forecasted with them for years…we knew some of the biases better, but their skill in every single metric was worse.

I think many forget that we used to weenietag people for getting too excited about a 144 hour storm threat…now people actually take them seriously. And there’s a good reason for that….models actually sniff out storms at 144 hours way better than they did in 2008 or 2009. Hell, I remember we tracked the 12/17/20 storm from like 200+ hours out, lol…it was consistently there on guidance. That type of stuff almost never happened back then. 

Yeah, the models aren't worse. The updated versions would never be allowed to be released if so.

 

I think people are falling into the example above that you stated and also, the gap has closed with the GFS and Canadian vs the Euro.

[ORH_wxman]

1 minute ago, CoastalWx said:

I'm 

Yeah, the models aren't worse. The updated versions would never be allowed to be released if so.

 

I think people are falling into the example above that you stated and also, the gap has closed with the GFS and Canadian vs the Euro.

Yeah the relative performance of the euro has declined but not the absolute performance. Maybe it will reclaim some mojo in El Niño where it seems to have a strength with a bunch of interacting shortwaves. 
 

Despite all the handwringing, this threat is starting to evolve kind of close to expectations…best chance of snow is NNE….CNE is precarious but has a legit shot…SNE is a longer shot. 
 
12/6 doesn’t seem to have enough room but too early to totally write it off. 
 

This all seems pretty consistent with our original thoughts…we just have a bit more confidence that 12/4 is the main show now and not 12/5-6. 

[Spanks45]

14 minutes ago, ORH_wxman said:

Euro was trying for 12/6 too but it just doesn’t have enough space. We need 12/4 to lift out quicker. 

Here in SNE, we actually have some cold air to play with too...just goes to show how perfectly we need to time something this time of the year to get anything more than a cold rain, at least down here 

[CoastalWx]

Let’s get DC some snow before Methuen.

[ORH_wxman]

3 minutes ago, Spanks45 said:

Here in SNE, we actually have some cold air to play with too...just goes to show how perfectly we need to time something this time of the year to get anything more than a cold rain, at least down here 

Yep…first week of Dec is pretty hostile still…esp outside of interior elevations. Climo rapidly gets better though as we move beyond that. Hopefully after the mid-month warmup, we can start cashing in during the 3rd and 4th weeks of the month…would be nice to have a snowy holiday period. 

[Ginx snewx]

Just now, CoastalWx said:

Let’s get DC some snow before Methuen.

Speaking of falling into the trap...

[ineedsnow]

Euro still has some snow here but I don't like that trend

[Lava Rock]

8 minutes ago, ORH_wxman said:

The models were a lot worse back then. I forecasted with them for years…we knew some of the biases better, but their skill in every single metric was worse.

I think many forget that we used to weenietag people for getting too excited about a 144 hour storm threat…now people actually take them seriously. And there’s a good reason for that….models actually sniff out storms at 144 hours way better than they did in 2008 or 2009. Hell, I remember we tracked the 12/17/20 storm from like 200+ hours out, lol…it was consistently there on guidance. That type of stuff almost never happened back then. 

maybe I didn't ask the right question. if the models were worse, were mets better at making more accurate frcsts with what they had? the other thing i'm not considering which speaks to your 2nd point is that in the "old" days (1990-2015) models didn't pick up on storms way out there (right?) and mets were working within a storm range of say 96hr, so those storms, if they panned out would make it seem the frcsts were better back then. it just seems we've spoiled ourselves with models that extend beyond 300hr, so one can get a lot of disappointment if what they're tracking doesn't work out. maybe i'm just babling, but was trying to make more sense of it to a weather weenie like me

[WinterWolf]

Lol call it what you guys want…nobody is falling in love with dumb ass clown maps, or saying because the euro showed a snow storm yesterday and not today it’s worse. Ultimately it isn’t as good anymore…ya can take the scores for the Himalayas and for the Gobi dessert and burn em..the model has lost its mojo. 
 

Sure, the GFS is a lot better now, but it’s still not as good as the Euro was 8 yrs ago. As somebody said..to sensitive now is probably the case, I agree.  
 

Moving on…I guess we wait a little longer in SNE..

[Ginx snewx]

18 minutes ago, ORH_wxman said:

The models were a lot worse back then. I forecasted with them for years…we knew some of the biases better, but their skill in every single metric was worse.

I think many forget that we used to weenietag people for getting too excited about a 144 hour storm threat…now people actually take them seriously. And there’s a good reason for that….models actually sniff out storms at 144 hours way better than they did in 2008 or 2009. Hell, I remember we tracked the 12/17/20 storm from like 200+ hours out, lol…it was consistently there on guidance. That type of stuff almost never happened back then. 

It's not that 5 plus day modeling that is the problem.  I agree the Euro is still king for the big dogs/ 12/20 and 1/22 being the latest long term hits. Overall you guys can disagree about performance but so many people have seen the degradation inside 96.

[CoastalWx]

1 minute ago, Ginx snewx said:

It's not that 5 plus day modeling that is the problem.  I agree the Euro is still king for the big dogs/ 12/20 and 1/22 being the latest long term hits. Overall you guys can disagree about performance but so many people have seen the degradation inside 96.

Euro wasn’t great in 1/22. Too far west.

[ORH_wxman]

Just now, Lava Rock said:

maybe I didn't ask the right question. if the models were worse, were mets better at making more accurate frcsts with what they had? the other thing i'm not considering which speaks to your 2nd point is that in the "old" days (1990-2015) models didn't pick up on storms way out there (right?) and mets were working within a storm range of say 96hr, so those storms, if they panned out would make it seem the frcsts were better back then. it just seems we've spoiled ourselves with models that extend beyond 300hr, so one can get a lot of disappointment if what they're tracking doesn't work out. maybe i'm just babling, but was trying to make more sense of it to a weather weenie like me

Yeah I think that’s part of it. We also have more model runs than ever. Everything seems to run 4x per day now…so that gives a better chance of seeing model guidance change. Or we bother to scrutinize every shift and pretend it’s not noise when sometimes it is. 
 

Like dendrite said earlier, we used to wait until 7-8pm for the 12z euro to come out and that was the only euro run…it didn’t even run at 00z until like 2003 or 2004 maybe? 

[Ginx snewx]

4 minutes ago, ORH_wxman said:

Yep…first week of Dec is pretty hostile still…esp outside of interior elevations. Climo rapidly gets better though we we move beyond that. Hopefully after the mid-month warmup, we can start cashing in during the 3rd and 4th weeks of the month…would be nice to have a snowy holiday period. 

Let's get a 12/5/03 again with those 2 to 3 foot totals. 

[dryslot]

12z EPS backs the OP as well on the 4th.

[ORH_wxman]

9 minutes ago, Ginx snewx said:

It's not that 5 plus day modeling that is the problem.  I agree the Euro is still king for the big dogs/ 12/20 and 1/22 being the latest long term hits. Overall you guys can disagree about performance but so many people have seen the degradation inside 96.

The thing that the euro used to be great at was being the first to sniff out a distinct trend. But it no longer has that superiority. It’s still the best model, but now we often get headfakes from the euro because unlike 10 years ago, when it shows a big change from other model guidance, it’s not always correct because other model guidance isn’t as inferior to the euro than it used to be. 
 

Years ago, if models were showing an exciting solution and all of the sudden, the euro shows a big trend at 108 hours, other guidance would almost certainly cave to that way more often than not because the euro was such a better model. Nowadays, if the euro shows a unique trend that other guidance isn’t showing, we stop and say “is euro out to lunch or is it trying to score a coup like the old days?”….and the reason we say that is models like the GFS aren’t pathetically inferior to it any longer. Only modestly so. 

[Ginx snewx]

10 minutes ago, CoastalWx said:

Euro wasn’t great in 1/22. Too far west.

Huh? Euro and EPS were actually a bit too far East from day 5/6

[dryslot]

The thing i find with the Euro of today then back several years ago is from run to run it would only make subtle changes towards a solution whereas today's euro make some bigger swings on some storms from run to run.

[CoastalWx]

Just now, Ginx snewx said:

Huh? Euro and EPS we were actually a bit too far East from day 5/6

Yeah and then at 72 hrs it buried the Berks.

[CoastalWx]

Just now, dryslot said:

The thing i find with the Euro of today then back several years ago is from run to run it would only make subtle changes towards a solution whereas today's euro make some bigger swings on some storms from run to run.

I find the GFS and Euro are close and a close compromise works inside 5 days. Maybe that changes this winter.

[Ginx snewx]

Just now, CoastalWx said:

Yeah and then at 72 hrs it buried the Berks.

1 bad run. Overall a stealler forecast for our white Juan

[powderfreak]

28 minutes ago, ORH_wxman said:

The models were a lot worse back then. I forecasted with them for years…we knew some of the biases better, but their skill in every single metric was worse.

I think many forget that we used to weenietag people for getting too excited about a 144 hour storm threat…now people actually take them seriously. And there’s a good reason for that….models actually sniff out storms at 144 hours way better than they did in 2008 or 2009. Hell, I remember we tracked the 12/17/20 storm from like 200+ hours out, lol…it was consistently there on guidance. That type of stuff almost never happened back then. 

This times 1000.  Models are way better now.  Maybe it’s a situation of information overload when 20 years ago or more you’d look at like 3-4 models.  Now you can look at dozens of you want to.

You’d look at the NGM, ETA and MRF and call it a day as a hobbyist and not see anything past Day 5 lol.

[CoastalWx]

Just now, Ginx snewx said:

1 bad run. Overall a stealler forecast for our white Juan

Was meh I felt. GFS was also maybe a tad east..but had the mid level features correct imo.