Winter 20-21 Discussion

[Typhoon Tip]

I wonder if it might be worth considering these QBOs at relative values - comparing  +1, vs +10, vs +20 ... -1 QBOs ...etc, to the solar min, the PDO/AMO multi-decadal oscillation and ENSO.

I keep reading 'a +QBO during this' ... or aa -QBO during that' ... but it seems one should not auto-applicate the QBO just based on whether it is positive or negative.

As we know, the QBO is a top down mass oscillation phenomenon... where +20 equates to a strong westerly wind flux, .. +1 is a weak...  Moreover the index moves from +20 to +1 to -1 and -20 ... and back, repeating every 20 to 36 months or so.. 

Is  a +5 QBO during a solar minimum the same correlation result as a +5 during a solar max... ?  It's obviously hugely complex - requiring a team of Aspergers types with a penchants for pattern recognition to ferret out, granted...but, it seems quite questionable to me to just apply a +7 QBO to a solar min that is extraordinarily strongly negative.   This latter may "overcome" the anti-correlation of +7 with SSW ( for example) and drive either a SSW anyway...or, maybe it is an SSW with a slow downwelling... I mean, jesus - there's a science fiction novel there that is actually too plausible to just be pure fiction!

Or not. Earlier this year ( I think it was a 2nd time this has ever happened ) the QBO abruptly broke cycle.. It slipped back negative as January turned page and continued, only resuming positive over this summer..and still is so.. Boo ya!  But will it remain that way... Since 2015, we've suddenly registered two of these occurrences ...which never happened since the 1950s.  This adds uncertainty ... ouch

 

[raindancewx]

Weather.com and Accuweather are showing ~record October snow for Albuquerque Monday-Tuesday. The models are showing upper level features supportive of snow banding, so someone could get 4-8 or 5-10 inches of snow. The record is 3.2 inches in Albuquerque for 1931-2019 in October.

Weather.com has 1-3" Monday, 1-3" Monday Night, and then 0-1" Tuesday Night/Wednesday AM. I look at that as 2-7" for the city. Accuweather has about 4 inches.

My snow forecast is probably going to bust low here if we actually do get 3,4,5 inches of snow in October. The La Nina Oct-May average is only 6 inches in Albuquerque, but October only averages 0.1 inches of snow long-term, and the median is 0.0", not even Trace.

 

 

 

[raindancewx]

Pretty good system for southern snow pack if it verifies. Chama averages ~100 inches so this would put them well on their way.

[andyhb]

Good start at hacking away the drought across the High Plains as well.

[raindancewx]

I've never seen a situation where a tropical storm on the Atlantic side could feed a NM/TX snow event, but it looks like that could happen on Tuesday/Wednesday. A couple years ago, you did have a dying pacific tropical storm enhance snow in October in CO/NM/TX, but it was a much warmer system than this one is. This was an earlier depiction for late Tuesday.

Actually, the whole "-NAO reverts to neutral", and rage of god cold descends into Montana for an extended period, does remind me of late January to February 2019, even though I dislike 2018-19 as an analog for the winter. My winter analog blend had a high of 73 or 74 for Albuquerque in October. The cold in the next week should push Albuquerque from 80 degrees Fahrenheit for the month to date high to about 74 or 75, so should end up being pretty close, even though we got there via record heat and record cold, while I had expected just general (weakly) above average temperatures. 

There was another huge SOI crash in recent days, so I expect the pattern to stay stormy in November, even if the US is warm overall, at least to about Thanksgiving. I'd expect the -NAO part of the pattern to return later in the month, but we'll see if I'm right about that. The SOI isn't really in line with the "this is a very strong La Nina" look you see in the Pacific right now. The SOI says this is a borderline Neutral/La Nina pattern. That may be temporary, but the next days looked like they'd have pretty -SOI readings too.

Date	Tahiti (hPa)	Darwin (hPa)	Daily Contribution	30 day Av. SOI	90 day Av. SOI
25 Oct 2020	1011.65	1009.80	-6.71	6.96	7.51
24 Oct 2020	1012.72	1009.65	1.16	7.74	7.60
23 Oct 2020	1013.92	1009.60	9.22	8.18	7.65

Weather.com is actually more optimistic than the local NWS for snow here too - keep in mind, in 90 years, the record for October snow in Albuquerque is 3.2 inches. Coldest low of October is also highly correlated with total days to see a low <=32F here from Oct-May.

 

 

[StormchaserChuck!]

I was thinking raindance knows some good relevant Winter correlations to major snowfall early this far south.. it's been a great localization experience. 

[raindancewx]

I don't think it really means anything. I did look at the warmest Oct highs 10/1-10/25 in Albuquerque that were also Octobers with accumulating snow in Albuquerque. Not exactly a big list. These are the warmest years 10/1-10/25.

1	1950-10-25	81.1	0
2	1979-10-25	80.8	0
3	2020-10-25	80.1	0
4	1991-10-25	79.7	0
5	1952-10-25	77.9	0
6	1992-10-25	77.4	0
7	1978-10-25	77.2	0
-	1963-10-25	77.2	0
9	1954-10-25	77.1	0
10	1955-10-25	76.9	0
11	2012-10-25	76.8	0

Out of those years, I had 2012 as an analog. But 1979, 1991, 1992, 1978, 1963, 1954, and 2012 are all average to great winters here. The 1979, 1991 temperature matches above did have snow in Albuquerque in October, despite their warmth. In the predictive periods i look at for the NAO, 1979 was a strong match, and 1991 has the fun and games in late October for the NE kind of like this year.

 

[LibertyBell]

On 10/22/2020 at 8:42 PM, raindancewx said:

Pick three similar years in Nino 3.4 in winter and see how close any of them are nationally when you ignore the other factors. These are the winters I consider La Ninas from 1931 to 2019. Look at 1984 and 1933, or 2011 and 1995 as an extreme example:

La Nina	3.4 DJF
1973	24.63
1988	24.83
1975	24.90
1999	24.95
2007	24.98
1998	25.07
1970	25.09
2010	25.21
1955	25.22
1949	25.26
1942	25.31
1950	25.41
1933	25.51
1984	25.55
1954	25.56
1964	25.69
1938	25.72
2017	25.72
1995	25.74
2011	25.76
1971	25.76
2008	25.79
2005	25.80
2000	25.87
1974	25.93
1983	26.00
1956	26.10
2016	26.30

wow I thought 1933-34 was one of the most extremely cold winters here in NY, looks like most of the rest of the country west of here wasn't cold at all.

 

[LibertyBell]

44 minutes ago, raindancewx said:

I don't think it really means anything. I did look at the warmest Oct highs 10/1-10/25 in Albuquerque that were also Octobers with accumulating snow in Albuquerque. Not exactly a big list. These are the warmest years 10/1-10/25.

1	1950-10-25	81.1	0
2	1979-10-25	80.8	0
3	2020-10-25	80.1	0
4	1991-10-25	79.7	0
5	1952-10-25	77.9	0
6	1992-10-25	77.4	0
7	1978-10-25	77.2	0
-	1963-10-25	77.2	0
9	1954-10-25	77.1	0
10	1955-10-25	76.9	0
11	2012-10-25	76.8	0

Out of those years, I had 2012 as an analog. But 1979, 1991, 1992, 1978, 1963, 1954, and 2012 are all average to great winters here. The 1979, 1991 temperature matches above did have snow in Albuquerque in October, despite their warmth. In the predictive periods i look at for the NAO, 1979 was a strong match, and 1991 has the fun and games in late October for the NE kind of like this year.

 

what about 2011 as an analog?  Looks like this storm might produce snow in the northeast too, first here in October since 2011....

[Caveman]

Hello all...I was wondering if anyone knows about the MAD WEATHER Teleconnections website?  Has it been moved to a different web address or has it been discontinued?  It was very helpful and one I often looked to during the cold season.  Any insights would be greatly appreciated!

[griteater]

5 hours ago, Caveman said:

Hello all...I was wondering if anyone knows about the MAD WEATHER Teleconnections website?  Has it been moved to a different web address or has it been discontinued?  It was very helpful and one I often looked to during the cold season.  Any insights would be greatly appreciated!

Yeah it looks like that page was taken down.  You can do a google image search on madus weather teleconnections and it will show some of those teleconnection pages scattered throughout the internet where folks have pasted them into a blog, etc., like here - https://dcstorms.com/2019/01/06/capital-weather-teleconnections-mjo-indicate-a-return-to-winter-january-2019/

[griteater]

On 10/23/2020 at 10:05 AM, Typhoon Tip said:

I wonder if it might be worth considering these QBOs at relative values - comparing  +1, vs +10, vs +20 ... -1 QBOs ...etc, to the solar min, the PDO/AMO multi-decadal oscillation and ENSO.

I keep reading 'a +QBO during this' ... or aa -QBO during that' ... but it seems one should not auto-applicate the QBO just based on whether it is positive or negative.

As we know, the QBO is a top down mass oscillation phenomenon... where +20 equates to a strong westerly wind flux, .. +1 is a weak...  Moreover the index moves from +20 to +1 to -1 and -20 ... and back, repeating every 20 to 36 months or so.. 

Is  a +5 QBO during a solar minimum the same correlation result as a +5 during a solar max... ?  It's obviously hugely complex - requiring a team of Aspergers types with a penchants for pattern recognition to ferret out, granted...but, it seems quite questionable to me to just apply a +7 QBO to a solar min that is extraordinarily strongly negative.   This latter may "overcome" the anti-correlation of +7 with SSW ( for example) and drive either a SSW anyway...or, maybe it is an SSW with a slow downwelling... I mean, jesus - there's a science fiction novel there that is actually too plausible to just be pure fiction!

Or not. Earlier this year ( I think it was a 2nd time this has ever happened ) the QBO abruptly broke cycle.. It slipped back negative as January turned page and continued, only resuming positive over this summer..and still is so.. Boo ya!  But will it remain that way... Since 2015, we've suddenly registered two of these occurrences ...which never happened since the 1950s.  This adds uncertainty ... ouch

 

Tip - you have a lot to unpack here, haha.

IMO, the QBO to La Nina/Cool ENSO relationship is one of the better winter seasonal forecast correlations that exists.  In a nutshell, when the QBO is positive at 50mb, this favors a more poleward North Pac ridge.  When the QBO is negative at 50mb, this favors a more suppressed North Pac ridge.  This correlation has been successful in 21 of 29 cases (72%).  For cases where the QBO was positive, it has been successful in 14 of 18 cases (78%).  2 years were thrown out (83-84 and 85-86) because the Jul-Oct Global AAM prior to the winter was positive, which is one indicator that the La Nina is not very well coupled with the atmosphere.  It is very rare to see the Jul-Oct Global AAM positive leading into a La Nina / Cool ENSO winter...the avg AAM is solidly negative in recent months. For QBO designation I like using the Free Univ of Berlin QBO chart to eyeball the QBO progression for each winter along 50mb, during the period Nov-Feb - https://www.geo.fu-berlin.de/met/ag/strat/produkte/qbo/qbo_wind_pdf.pdf.  For the cases that failed, there was no bias as to whether the QBO was rising or falling, or whether the amplitude of the QBO was high or low.  The key differentiating factor was whether the QBO was positive or negative.

HM previously had a detailed post on his blog about this correlation (from 2012), but his blog is no longer active / posts are no longer visible...but here is what I believe is at least part of the science behind the correlation.  During the +QBO (Westerly), there are warm anomalies in the lower stratosphere along the equator with cool anomalies off equator into the subtropics.  The higher tropopause height off equator allows for deeper, more robust tropical convection there (compared to normal) - this is termed the "QBO Induced Meridional Circulation"

 

Images below contain OLR for a mix of weak to strong La Ninas.  On the left are -QBO Ninas.  +QBO Ninas on the right.  We can see the +QBO composite showing enhanced off-equator convection in the 10N to 25N zone.  Note: on these images, the color scheme is reversed from the typical OLR composites...that is, the warm colors are showing -OLR anomalies (enhanced convection), with the cool colors showing +OLR anomalies (reduced convection).

 

It's my belief that the enhanced off-equator convection aids more poleward ridge building in the North Pacific via the typical process (image below from: https://blog.timesunion.com/weather/how-el-nino-influences-our-weather/1247/)

 

Looking at daily Singapore sounding data, we can see that the +QBO is behaving as expected this fall as temperatures in the lower stratosphere along the equator are showing plus anomalies (far right side of image)

 

Also, we can see in this loop of 50mb temperatures, the warm anomalies along the equator, with the cool anomalies off equator

 

Here are composites of -QBO / "South" / Suppressed North Pacific Ridge winters VS. +QBO / "North" / Poleward North Pacific Ridge winters during La Nina / Cool ENSO (I used SON to JFM averaged ONI of -0.20 or lower for selection of the years)

 

 

Here is the +QBO "North" composite with trends noted for years since 1990

 

Having said all of that, there are years where the correlation doesn't work.  2 reasons can be: 1) The lower stratospheric temperature profile in the tropics and subtropics doesn't match the QBO Induced Meridional Circulation pattern discussed earlier, and 2) The polar vortex / AO / NAO are overly positive/strong and zonal flow is too strong to allow development of poleward ridges in the mean pattern.

Lastly, there were 2 other things you mentioned that I wanted to touch on.  One was the interrupted QBO cycle.  Indeed, one interruption occurred in the winter of 15-16 and the other occurred last winter.  The QBO does however appear to be back to a more typical cycle (at least for now).  We can see the tan colors here showing the +QBO stretched out from roughly 20mb down to 70mb on the far right

 

You also mentioned utilizing the QBO in terms of its movement thru the QBO cycle/progression.  Sam Lillo has done work in this arena.  Here is one of his images showing the occurrence of strat vortex weakening events at various times in the QBO cycle.  You can see at various points in the cycle where there is preference for early winter vs. late winter weakening events.  My guess is that we are currently somewhere between W30 to W50 on this image.

[raindancewx]

November on the CFS looks a lot like November 1954 if you want a cold ENSO year to match it to. Very hot in the West all Summer & Fall, and then cold in winter. It's the cold Western composite look, ala 2007 or 1988 that Griteater has.

Here is a look at some snow totals for the storm down here. A lot of the mountain communities have already had 20-30% of their annual totals with this storm and the September storm. I don't see any reason to not expect a recurrence of these types of storms occasionally through next Spring. They seem to be tied to the NAO flipping phase via changing MJO forcing. The "extremely low snow" in the Southwest idea already looks pretty wrong. We only get ~0-5% of our snow through October in the sites I'm familiar with in NM. I think Albuquerque officially had 3.9 inches - pending the official data compilation tonight. I had 7.5 inches with the storm. But it's certainly looking like a record snow event for October here.

[raindancewx]

Quick update on some ideas for the winter after checking the stats here:

- All years from 1931-2019 with measurable snow in Albuquerque in October (10/10) saw measurable snow in at least five months of the eight month Oct-May cold season. This is compared to only 24 out of 79 years when it doesn't snow in October. The odds of six months of measurable snow are also much higher: 3 out of 10 in years with snow in October, and 4 out of 79 without snow in October. These are both highly statistically significant differences. The city has only seen measurable snow from October-May from 1931-32 to 2019-20.

- March also saw snow 10/10 times following measurable snow in October, but only 48 of 79 Marches did without snow in October. 

So I'm expecting snow in at least four more months of snow here in the valleys of the Southwest. That means the mountains should get hit pretty hard. The question is which four? January/February actually see somewhat lower frequency of accumulating snow than usual after it snows in October. So I lean toward November, December, March, and then one-two months of January, February and April. 

Western snow pack is incredible in the high terrain at the moment. I'm sure it will let up for a bit, but if the mountains get enough snow, you can create cold air that prevents the valleys from warming up much.

Years with unusual snow pack well to the south in the Plains/Rockies in October will often see some snow events pretty deep into the south later in winter.

The official number still isn't in yet, but I think the airport had 3.9 inches of snow. The long-term average is 9.6" for October-May, but we only average 0.1" in October. It's pretty likely that the city will finish above average for snow - that's very rare in a La Nina, but the odds are close to a coin-flip for it now. 

When I did my winter forecast, total precipitation came out to 0.30" for October in Albuquerque with a high of 73.2F in the raw analog blend. We're likely to finish at 73-74, with 0.22", so it's on the right track. The blend of years implied a mostly dry pattern through May, but one month in Nov-April should see 2-3x normal precipitation (around one inch). There are hints on the GFS of storms entering southern California in the long-range too, which is consistent with November evolving to a colder pattern for December in the West if I did this right.
 

  

[DTWXRISK]

On 10/22/2020 at 2:37 PM, griteater said:

40/70 - thing is, at 500mb, a composite of the last 2 winters (below) looks like La Nina instead of El Nino with NPac ridge and -PNA.  So, when I composited the last 4, it came out to kind of what I had in my head for this winter in the big picture. 

+QBO this winter would favor a more poleward NPac ridge.  That composite of mine does have a poleward ridge (no negative anomalies in Alaska), but it's biased to the NW.  4 of the last 5 mod-strong Ninas with a poleward NPac ridge were biased to the NW ('56, '71, '74, '11) .  '89 is the lone exception.  One thing that would support the idea of a suppressed NPac ridge would be a strong +AO.  The seasonal models favor a suppressed ridge at the moment.  AO/NAO are always a challenge to figure out.

Anyway, those are some thoughts on it...hope I will have time to put more into it the next 2 weeks...my work has been busier than normal of late.

 

   YOU   do  realize that  the  idea  that  the La Nina   is  going to stay  Moderate  to strong all winter  long is  total    crap right ? 
 that None of the enso  data / models supports that  idea ?
 

[DTWXRISK]

On 10/27/2020 at 12:53 PM, griteater said:

Tip - you have a lot to unpack here, haha.

IMO, the QBO to La Nina/Cool ENSO relationship is one of the better winter seasonal forecast correlations that exists.  In a nutshell, when the QBO is positive at 50mb, this favors a more poleward North Pac ridge.  When the QBO is negative at 50mb, this favors a more suppressed North Pac ridge.  This correlation has been successful in 21 of 29 cases (72%).  For cases where the QBO was positive, it has been successful in 14 of 18 cases (78%).  2 years were thrown out (83-84 and 85-86) because the Jul-Oct Global AAM prior to the winter was positive, which is one indicator that the La Nina is not very well coupled with the atmosphere.  It is very rare to see the Jul-Oct Global AAM positive leading into a La Nina / Cool ENSO winter...the avg AAM is solidly negative in recent months. For QBO designation I like using the Free Univ of Berlin QBO chart to eyeball the QBO progression for each winter along 50mb, during the period Nov-Feb - https://www.geo.fu-berlin.de/met/ag/strat/produkte/qbo/qbo_wind_pdf.pdf.  For the cases that failed, there was no bias as to whether the QBO was rising or falling, or whether the amplitude of the QBO was high or low.  The key differentiating factor was whether the QBO was positive or negative.

HM previously had a detailed post on his blog about this correlation (from 2012), but his blog is no longer active / posts are no longer visible...but here is what I believe is at least part of the science behind the correlation.  During the +QBO (Westerly), there are warm anomalies in the lower stratosphere along the equator with cool anomalies off equator into the subtropics.  The higher tropopause height off equator allows for deeper, more robust tropical convection there (compared to normal) - this is termed the "QBO Induced Meridional Circulation"

 

Images below contain OLR for a mix of weak to strong La Ninas.  On the left are -QBO Ninas.  +QBO Ninas on the right.  We can see the +QBO composite showing enhanced off-equator convection in the 10N to 25N zone.  Note: on these images, the color scheme is reversed from the typical OLR composites...that is, the warm colors are showing -OLR anomalies (enhanced convection), with the cool colors showing +OLR anomalies (reduced convection).

 

It's my belief that the enhanced off-equator convection aids more poleward ridge building in the North Pacific via the typical process (image below from: https://blog.timesunion.com/weather/how-el-nino-influences-our-weather/1247/)

 

Looking at daily Singapore sounding data, we can see that the +QBO is behaving as expected this fall as temperatures in the lower stratosphere along the equator are showing plus anomalies (far right side of image)

 

Also, we can see in this loop of 50mb temperatures, the warm anomalies along the equator, with the cool anomalies off equator

 

Here are composites of -QBO / "South" / Suppressed North Pacific Ridge winters VS. +QBO / "North" / Poleward North Pacific Ridge winters during La Nina / Cool ENSO (I used SON to JFM averaged ONI of -0.20 or lower for selection of the years)

 

 

Here is the +QBO "North" composite with trends noted for years since 1990

 

Having said all of that, there are years where the correlation doesn't work.  2 reasons can be: 1) The lower stratospheric temperature profile in the tropics and subtropics doesn't match the QBO Induced Meridional Circulation pattern discussed earlier, and 2) The polar vortex / AO / NAO are overly positive/strong and zonal flow is too strong to allow development of poleward ridges in the mean pattern.

Lastly, there were 2 other things you mentioned that I wanted to touch on.  One was the interrupted QBO cycle.  Indeed, one interruption occurred in the winter of 15-16 and the other occurred last winter.  The QBO does however appear to be back to a more typical cycle (at least for now).  We can see the tan colors here showing the +QBO stretched out from roughly 20mb down to 70mb on the far right

 

You also mentioned utilizing the QBO in terms of its movement thru the QBO cycle/progression.  Sam Lillo has done work in this arena.  Here is one of his images showing the occurrence of strat vortex weakening events at various times in the QBO cycle.  You can see at various points in the cycle where there is preference for early winter vs. late winter weakening events.  My guess is that we are currently somewhere between W30 to W50 on this image.

OUTSTANDING   POST ANALYSIS....  WOW    

I like the  north pacific Riidge     +QBO map   sicne 1990   a LOT 

[raindancewx]

None of the lowest sea-ice cold ENSO years (2007, 2011, 2012, 2016) that are closest to the current observations look anything like the 'North' North Pacific Ridge composite since 1990 in winter. I can't imagine the lack of sea ice not having some kind of effect on placement of highs/lows in the North Pacific / Arctic zones. An example of this would be 2016, when DT had expected a big trough in the East/Midwest for winter based on the QBO/ENSO progression.

More generally, the seven winters themselves are all very different. Say you lived in the Northeast and believed the North North Pacific Ridge composite was right. Would you go hot or cold for the Northeast? Average? The "look" of the seven post 1990 winters don't really match this October either. 1995 is closest for October of the bunch but it could easily be warmer in the Plains/East this November. 2008 is second closest for Oct-Nov, but it had the South very cold in October, and the Plains near average, but it's probably a good match for November. 2010, the La Nina most of you in the South want, is a god awful match for October in particular, with the entire middle of the US warm, and California cold. The 2013 October was extremely cold in the entire West, which is not like this year at all, even with the recent cold shot. The 2016 October is hot in the middle of the US too. 2017 is warm outside the NW. I'm not saying the composite or these years won't work somehow, but I am skeptical given how close we are to winter with almost all of these winters behaving differently. I consider 1995 a decent analog, and 2008 is probably fine for the eastern third of the US, but I'm not a fan of the other years at all.

 

 

[griteater]

^ Raindance - it's just one piece of the puzzle.  You need to be able to predict whether the North Pac ridge will be poleward (North) or suppressed (South)...and whether it will be biased to the east or to the west (from a seasonal standpoint).  You also need to be able to predict what the AO/NAO are going to do.  All of that impacts the forecast.  For example, for 16-17, if you could forecast that the N Pac ridge was going to be biased to the NW and that the AO/NAO were going to be positive, you'd get the general idea correct of cool in the PNW and warm in the eastern half of the U.S.

[raindancewx]

My thing with La Ninas is they aren't really Pacific driven for most of the US, outside the NW. It's much easier to figure out what will happen in a La Nina by focusing on the Atlantic. That's part of why I'm so anti-2010 as an analog. That was a very active hurricane season without any landfalling Gulf Coast hurricanes, and without anything like the late activity this year has had. 

To me, the pattern has three main components.

1) You have transient, but extremely powerful cold dumps into the Southwest that occur around major changes in arctic pressure patterns, likely triggered by big ENSO/MJO changes. These have differed in how they've been delivered, but the cold snaps actually have been recurring every 45 days or so for months already, despite the persistent heat. Early June, Late July, Early Sept, Late Oct all featured these cold shots in NM.

2) You have major cold dumps into the Northwest and Northern Plains that don't reach the SW or East at all. These cold dumps will likely shift West if/when the SE ridge expands in power as the La Nina peaks, and then snap back east in Spring. My analogs had a cold Spring in the east after a pretty mild winter. 

3) You also have periods of intense Western heat corresponding with cold/warmth duking it out pretty evenly in the East with Arctic/NAO help fighting the Pacific. This 10/1-10/10 pattern is essentially the same pattern the CFS sees for November (it's essentially Phase 4 MJO for Nov, v. Phase 5 Oct).

Broadly speaking, the big dumps of cold into the SW have occurred shortly after a period of La Nina weakening or during a period of slower strengthening. The models have the La Nina weakening a lot starting Dec/Jan...part of why I don't buy a hot winter here. I lean toward a 45-35-20% split for the three patterns: with the middle image the 45, the near average east 35, and the 20 the SW cold dump. But the middle image should shift West as the La Nina strengthens and/or a SE ridge grows potent. Locally, the hottest, driest Summers in La Nina tend to see sporadic intense cold shots in mid-Fall develop and gradually become more consistent by the winter, so everything is pretty on target for what I've been expecting. 

[raindancewx]

The Canadian has a very warm patch of water between Alaska and Hawaii which to me implies at least some resemblance to your 'South' North Pacific Ridge look.

[raindancewx]

Cold ENSOs with ongoing MJO waves around phase 7 on 11/1:

2017 - started but died, i.e. lower magnitude

2016 - a week late for phase 7

2000 - pretty close to the Euro outlook by timing intensity (phase 2 by 11/15 like in 2000)

1996 - 10 days late for phase 7, and higher magnitude

That's really it since 1975. As a blend, it's pretty close to what the Canadian showed for November.

 

My unweighted winter analog blend was pretty close in October, and looks like the models for November. But of course, there is a much simpler blend to get Oct-Nov...1954:

 

[40/70 Benchmark]

On 10/30/2020 at 4:50 PM, DTWXRISK said:

OUTSTANDING   POST ANALYSIS....  WOW    

I like the  north pacific Riidge     +QBO map   sicne 1990   a LOT 

I second this....incredibly informative. Still trying to wrap my mind around the QBO stuff.

 

[40/70 Benchmark]

On 10/30/2020 at 9:06 PM, raindancewx said:

None of the lowest sea-ice cold ENSO years (2007, 2011, 2012, 2016) that are closest to the current observations look anything like the 'North' North Pacific Ridge composite since 1990 in winter. I can't imagine the lack of sea ice not having some kind of effect on placement of highs/lows in the North Pacific / Arctic zones. An example of this would be 2016, when DT had expected a big trough in the East/Midwest for winter based on the QBO/ENSO progression.

More generally, the seven winters themselves are all very different. Say you lived in the Northeast and believed the North North Pacific Ridge composite was right. Would you go hot or cold for the Northeast? Average? The "look" of the seven post 1990 winters don't really match this October either. 1995 is closest for October of the bunch but it could easily be warmer in the Plains/East this November. 2008 is second closest for Oct-Nov, but it had the South very cold in October, and the Plains near average, but it's probably a good match for November. 2010, the La Nina most of you in the South want, is a god awful match for October in particular, with the entire middle of the US warm, and California cold. The 2013 October was extremely cold in the entire West, which is not like this year at all, even with the recent cold shot. The 2016 October is hot in the middle of the US too. 2017 is warm outside the NW. I'm not saying the composite or these years won't work somehow, but I am skeptical given how close we are to winter with almost all of these winters behaving differently. I consider 1995 a decent analog, and 2008 is probably fine for the eastern third of the US, but I'm not a fan of the other years at all.

 

 

I agree. I don't buy the N Pac ridge composite, however, I don't think it's going to be as flat as some the eastern US terds, like 2011, either.

[40/70 Benchmark]

On 10/30/2020 at 11:04 PM, griteater said:

^ Raindance - it's just one piece of the puzzle.  You need to be able to predict whether the North Pac ridge will be poleward (North) or suppressed (South)...and whether it will be biased to the east or to the west (from a seasonal standpoint).  You also need to be able to predict what the AO/NAO are going to do.  All of that impacts the forecast.  For example, for 16-17, if you could forecast that the N Pac ridge was going to be biased to the NW and that the AO/NAO were going to be positive, you'd get the general idea correct of cool in the PNW and warm in the eastern half of the U.S.

I hate to be redundant, but 2007 represents a great compromise both with respect to the ep vs cp la nina continuum, as well as the N PAC ridge vs S PAC ridge composites.

[40/70 Benchmark]

@griteater No coincidence that most of the N PAC ridge composite winters are ep la nina seasons...and most S PAC members are co la nina seasons. 

[raindancewx]

There is a real chance that Nino 3.4 will be top ten for cold readings since 1950 in both November & December. Especially if the European continues to underestimate how cold it is in Nino 3.4. Anything under 25.4C in Nino 3.4 in November would be top ten for 1950-2019. From there, only a small drop to 25.25C or so would be top ten in December. I do expect substantial warming to begin mid-January. You can sort of see the origins of it with Nino 1.2 and Nino 3 flatlining/warming slowly.

The cold-ENSO years with a +0 to +8 SOI in October are also a pretty interesting bunch (+4 in Oct 2020):

1931    4.3
1933    4.1
1934    4.7
1948    6.6
1949    6.0
1954    2.2
1959    4.7
1961    4.7
1983    4.7
1996    5.2
2007    6.1
2012    2.3

It's pretty unusual to have an SOI value at +4 in a pretty healthy La Nina, but 2007 did. This group of years is primarily average to very cold in the West (1931, 1934, 1948, 1949, 1954, 1959, 1961, 1983, 1996, 2007, 2012) or 1933, with much more varied results in the East.

[40/70 Benchmark]

On 10/31/2020 at 9:34 PM, raindancewx said:

The Canadian has a very warm patch of water between Alaska and Hawaii which to me implies at least some resemblance to your 'South' North Pacific Ridge look.

Canadian is also the most central-based of all guidance, IMO...so that makes sense that it would support the southern PAC ridge.

[raindancewx]

Hurricane Eta reminds me of Mitch in 1998. Mitch was a god damned beast, especially for late Oct/early November. I sincerely hope this doesn't kill 10,000 people in five days with 75 inches of rain in the mountains and hills of Central America, before ejecting out and smacking Florida like Mitch did. Mitch peaked at 180 mph sustained winds, so I don't think we'll see that. Pretty sure it also stalled forever and did not behave according to model tracks. That's what I remember from back then.

Eta does push the ACE into the 150s though. That's likely inflated compared to historical years with systems like the storm that hit Portugal that would have never been named in prior years. But anyway, this is the composite for ~about 150 ACE in a La Nina. Not too different from the 80-160 composite overall. 

 

 

[40/70 Benchmark]

Not sure ACE is done for the season.

[raindancewx]

It may not be. But I think older years would have had 10-20 extra points the way storms are tracked and observed today. So I think the final number today minus 10 is a better estimate when compared to historical seasons. I gave a range of 95-175 for the season in my winter forecast for a reason. I fully expected 135, but there are some La Ninas with a lot of late activity.