• Hello, please take a minute to check out our awesome content, contributed by the wonderful members of our community. We hope you'll add your own thoughts and opinions by making a free account!

Pattern The Great December Dump

But don't ignore the horrible GEFS cold bias either. I recommend you take the GEFS and warm it up considerably if you're going to use it for a forecast.
Dude! I think we can pretty much all agree on this by now. It has a cold/cold pattern bias every year. Has for years.
 
Dude! I think we can pretty much all agree on this by now. It has a cold/cold pattern bias every year. Has for years.

And dude, this needs to be emphasized as long as the maps are shown without the "cold pattern/bias" disclaimer. Not everyone is aware of this or wants to accept this and there are also new readers.
 
And dude, this needs to be emphasized as long as the maps are shown without the "cold pattern/bias" disclaimer. Not everyone is aware of this or wants to accept this and there are also new readers.
I think the people who are going to get the concept get it by now. That's kind of like having to remind people not to believe a 10 day map every single time a 10 day map is posted. Or like saying the models flip flop every time the models flip flop. Some people aren't going to care about anything but whether or not there's blue cold or white snow being shown over their house.
 
I think the people who are going to get the concept get it by now. That's kind of like having to remind people not to believe a 10 day map every single time a 10 day map is posted. Or like saying the models flip flop every time the models flip flop. Some people aren't going to care about anything but whether or not there's blue cold or white snow being shown over their house.

I always consider the "no snow for the SE" bias on all models. If it shows cold and snow, it's corrupted and should be discounted. Always. ;)

I'm not ready to say the GEFS has got the wrong pattern. It may be overdoing the degree of cold, but the pattern could be correct IMO. EPS versus GEFS both have their strengths and weaknesses I think.
 
But don't ignore the horrible GEFS cold bias either. I recommend you take the GEFS and warm it up considerably if you're going to use it for a forecast.
I use a blend of both ensembles, it's never a good idea to take one model at face value. I was just verifying if that massive Bermuda high was there on the GEFS. Since that is 300+ hours out, that Bermuda high does not have a chance at verifying at this time, considering that the GEFS does not have it. If other ensembles don't show that Bermuda high, the chance of that Bermuda high is slim. Of course, things can always change.
 
I always consider the "no snow for the SE" bias on all models. If it shows cold and snow, it's corrupted and should be discounted. Always. ;)

I'm not ready to say the GEFS has got the wrong pattern. It may be overdoing the degree of cold, but the pattern could be correct IMO. EPS versus GEFS both have their strengths and weaknesses I think.
I always view anything 10 days or over with skepticism. What I try to use to give me a little more confidence is agreement between model/ensemble suites on the general pattern, as well as pattern recognition. IMO, like Larry has been saying, the GFS tends to push lower heights too far south in the east in the LR. That leads to colder temps than what will eventually occur. It's been that way for a long time. BUT, it doesn't mean it doesn't have the general pattern right.

The take-away is that, even if there is a cold bias, it more or less means that it will average warmer than it's showing. But when it comes to specific daily temps or specific storm chances, there is no appreciable skill way out there. But if the general pattern is favorable, it tells you that there is at least a period of interest coming up....if that general look holds as we move in and especially if there is agreement between model suites and especially if there are features that are present that suggest a stability to the pattern.
 
Part of the problem is that any model when over analyzed past day 10, especially day 15, is going to show some large inconsistencies...it seems the GEFS biased cold or pattern biased. EPS has large errors day 10+ too.

But, to start December both the GEFS and EPS are showing a -EPO with an eroding -NAO. Question is how far east/west does the actual ridge setup in the pacific. Is it out over the GOA like the 0z EPS or is it splitting the Yukon like the GEFS.

gfs-ensemble-all-avg-nhemi-z500_anom_5day-5460800.png
 
Part of the problem is that any model when over analyzed past day 10, especially day 15, is going to show some large inconsistencies...it seems the GEFS biased cold or pattern biased. EPS has large errors day 10+ too.

But, to start December both the GEFS and EPS are showing a -EPO with an eroding -NAO. Question is how far east/west does the actual ridge setup in the pacific. Is it out over the GOA like the 0z EPS or is it splitting the Yukon like the GEFS.

View attachment 26332

Obviously, not all -EPOs are created equally, you can have the index crash to several sigma below climatology and end up really mild around here esp if the anticyclonic wave break is suppressed significantly equatorward of the arctic ocean.
 
Obviously, not all -EPOs are created equally, you can have the index crash to several sigma below climatology and end up really mild around here esp if the anticyclonic wave break is suppressed significantly equatorward of the arctic ocean.

You mean like this train wreck? Kind of what the EPS has been showing for early Dec.

compday.O33AuMbZPQ.gif
 
Still think we have a small shot at something first week of December, at least maybe us upper South folks (NC and Canada Lol sorry @NoSnowATL you started it)

View attachment 26335

I think you northerners have a shot for sure. Heck I might see some 31 degree rain out of it. That time period has been popping up with potential so it’s worth watching.


Sent from my iPhone using Tapatalk
 
Great example here of how you can have a -EPO & a -NAO simultaneously but still end up warmer than average in spite of the fact that you're in February of a NINO winter.

View attachment 26334

I assume you'd agree that the Feb 2019 Nina-like strong -PNA (-1.08), which is consistent with a strong SER, had a lot to do with this warm SE:

https://www.cpc.ncep.noaa.gov/products/precip/CWlink/pna/norm.pna.monthly.b5001.current.ascii.table

I know you know this but for those who don't: with a solid -PNA, persistent cold in the SE, especially far SE, is hard to get even with a strong -EPO like Feb 2019 had. With El Nino, a +PNA is usually favored over a -PNA thus helping to make Nino's chilly on average in the SE.
 
Last edited:
I assume you'd agree that the Feb 2019 Nina-like strong -PNA (-1.08), which is consistent with a strong SER, had a lot to do with this warm SE:

https://www.cpc.ncep.noaa.gov/products/precip/CWlink/pna/norm.pna.monthly.b5001.current.ascii.table

I knwo you know this but for those who don't, a solid -PNA, persistent cold in the SE, especially far SE, is hard to get even with a strong -EPO like Feb 2019 had. With El Nino, a +PNA is usually favored over a -PNA thus helping to make Nino's chilly on average in the SE.

It's not entirely accurate to say the -PNA had a lot to do w/ the warm SE or any mid-latitude teleconnection for that matter, it's important to remember that they're intended to provide simpler, one-dimensional output that represents important physical processes that are forcing them to behave in a certain manner. Not to mention, when the PNA is actually calculated by either z500 box averaging or regressing northern hemisphere 500mb heights onto the PNA pattern, the SE US is one tripole that's directly included in its calculation, so the stronger the SE US ridge gets, the lower the PNA value & vis versa.

Teleconnections are meant to simplify complex interactions between mid-latitude waves, the waveguide, tropical convection, & mechanical forcing from mountain ranges, etc. in a way that may be more easily digestible, but they shouldn't be a replacement for stringent analyses of the actual, physical processes that are forcing the atmosphere to express itself in a particular manner nor are they a "forcing" per say in the sense many here often allude to.
 
Folks,
Even though the GEFS has been cold in the 11-15 for several days, always beware its strong cold bias mainly due to too cold a pattern! Per Maxar, the last 30 days have been averaging a whopping 6 too cold just in the 6-10 at ATL and that even incorporates the very cold early Nov when the model was not cold enough!! So, it would have been even warmer without that very cold period. Even the EPS has been 3 too cold at ATL the last 30 days.

Nationally the GEFS has only been 3 too cold, but that's because cold biases are less in the Midwest and near 0 with even a slight warm bias in some western areas. The EPS has been 1 too cold nationally. I had thought the EPS had averaged slightly too warm but that hasn't been the case.

As an example of how bad the GEFS cold bias has been on a national basis, I looked at the nationally weighted heating degree days (US HDD) of the 12Z GEFS from one week ago today and compared its week 2 to today's 12Z GEFS week 1. The results are horrible. I mean fail/flunky bad. Every day of 11/23-30 was predicted by the GEFS of a week ago to be solidly colder than normal. The current GEFS has all of those days near normal.


US HDD
Date.....12Z 11/15 GEFS.....12Z 11/22 GEFS

11/23.......….......21......…...............……..19
11/24...…….……22.......................……....19
11/25.........……..23.............…......……….18
11/26......…….....22.........….……...…...….16
11/27......…….….22................…...…...…..17
11/28......………..23......…....…...………….19
11/29......…….….23.............…...…...........20
11/30...…...……..24......….......…..............19

11/23-30 avg...22.5.................….......18.5

The 22.5 US HDD forecasted a week ago is 22.5, which is a whopping 4.5 higher (colder) than the normal of only 18. The current run has only 18.5, which is 4 lower (warmer) than the run from a week ago and is only 0.5 higher (colder) than the normal of 18. It really stinks, folks. To make matters worse, even today's GEFS prediction of near normal for 11/23-30 will come in quite a bit too cold as a warmer than normal US is expected. So, the too cold miss of 4 HDD on the week ago run will verify even further off once 11/23-30 actuals come in. SMH. Fake extreme GFS/GEFS cold never ends!

I just did the same kind of analysis of the EPS from a week ago vs its latest run. Whereas the current GEFS is warmer by an average of 4 US HDD/day for 11/23-30 vs what the GEFS from one week earlier showed, the current EPS is only just over 1 US HDD warmer. This is consistent with the idea that the EPS has been a more accurate model than the GEFS with less cold bias.

By the way, the 12Z EPS came in a good bit colder for the SE in the 6-15 day period vs the 0Z. It has the SE colder than normal for 12/3-7 vs near normal on the prior run and vs much colder than normal on the 12Z GEFS/GEPS. I think the safest bet right now is to go with the EPS and colder than normal for 12/3-7 though not much colder than normal like the GEFS/GEPS suggest. (GEPS also has a strong cold bias for those who don't know.)
 
Last edited:
Back
Top