Extreme October Blocking Event Hints at AO- Winter

[donsutherland1]

Today, the Arctic Oscillation (AO) stood at -4.057. That is the 6th lowest figure on record for October. 2014 is only the second year since 1950 during which the AO fell to -4.000 or below in October. Previously, only the great October blocking episode of 2002 during which the AO stood at -4.000 or below during the October 16-20 saw such an outcome.

 

More importantly, blocky Octobers following summers during which the AO averaged <0 typically saw a blocky winter. Since 1950, there have been 12 cases during which the AO averaged -0.500 or below during October with a summer AO < 0. 10/12 (83%) cases saw the winter AO average < 0.

 

Some detailed information from those cases is below:

 

 

Overall, especially if the ensemble guidance proves accurate through the medium-term, it appears that winter 2014-15 could see a predominantly negative AO.

[JBG]

Today, the Arctic Oscillation (AO) stood at -4.057. That is the 6th lowest figure on record for October. 2014 is only the second year since 1950 during which the AO fell to -4.000 or below in October. Previously, only the great October blocking episode of 2002 during which the AO stood at -4.000 or below during the October 16-20 saw such an outcome.

 

More importantly, blocky Octobers following summers during which the AO averaged <0 typically saw a blocky winter. Since 1950, there have been 12 cases during which the AO averaged -0.500 or below during October with a summer AO < 0. 10/12 (83%) cases saw the winter AO average < 0.

 

Some detailed information from those cases is below:

 

 

Overall, especially if the ensemble guidance proves accurate through the medium-term, it appears that winter 2014-15 could see a predominantly negative AO.

I thought 1960-1 was extremely "non-blocky."

[uncle W]

I thought 1960-1 was extremely "non-blocky."

the nao averaged on the plus side 1960-61...

[JBG]

the nao averaged on the plus side 1960-61...

For the uninitiated what is the difference between AO blocking and NAO blocking?

[donsutherland1]

I thought 1960-1 was extremely "non-blocky."

I'm using the Arctic Oscillation (AO).

 

The 1960-61 monthly AO figures were:

 

December: -0.343

January: -1.506

February: +0.621

 

2 major AO- periods:

12/7-23/1960

12/28/1960-1/29/1961

[donsutherland1]

For the uninitiated what is the difference between AO blocking and NAO blocking?

The NAO deals with height anomalies over the North Atlantic. The AO covers a larger area (20°N and northward). There is often, but not always, a relationship between the two. The AO has a larger hemispheric impact than the NAO.

[ORH_wxman]

I thought 1960-1 was extremely "non-blocky."

 

 

The NAO deals with height anomalies over the North Atlantic. The AO covers a larger area (20°N and northward). There is often, but not always, a relationship between the two. The AO has a larger hemispheric impact than the NAO.

 

 

Yeah '60-'61 had a big block on the Pacific side in the EPO region and even extending all the way over into northern Siberia. It phased with a +PNA ridge at times which gave some bouts of extreme cold.  

 

That winter is one of the few examples of where the NAO being solidly positive was out of phase with the AO as a whole.

[uncle W]

the nao is tanking also this week...1960-61 daily nao index...it was negative around the time of the three blizzards...

1960 12  8 -0.6601960 12  9 -0.7811960 12 10 -0.4381960 12 11 -0.3161960 12 12 -0.0781960 12 13 -0.1081960 12 14 -0.6561960 12 15 -0.7481960 12 16 -0.3501960 12 17 -0.3221960 12 18 -0.4531960 12 19 -0.4461960 12 20 -0.6471960 12 21 -0.7521960 12 22 -0.6361960 12 23 -0.2091960 12 24  0.4871960 12 25  0.8511960 12 26  0.7211960 12 27  0.3391960 12 28 -0.0031960 12 29 -0.2181960 12 30 -0.1701960 12 31 -0.2511961  1  1 -0.2681961  1  2 -0.2211961  1  3 -0.2521961  1  4 -0.0591961  1  5 -0.0061961  1  6  0.1641961  1  7  0.4801961  1  8  0.6091961  1  9  0.5671961  1 10  0.4041961  1 11  0.6371961  1 12  0.8401961  1 13  0.5321961  1 14  0.2171961  1 15  0.0321961  1 16  0.1971961  1 17  0.2431961  1 18  0.3271961  1 19  0.3921961  1 20  0.1861961  1 21 -0.2161961  1 22 -0.2631961  1 23 -0.0531961  1 24  0.0261961  1 25  0.0231961  1 26  0.1131961  1 27  0.4291961  1 28  0.3721961  1 29  0.2841961  1 30  0.5331961  1 31  0.4071961  2  1  0.1261961  2  2 -0.0481961  2  3 -0.1261961  2  4 -0.0791961  2  5 -0.0021961  2  6 -0.0101961  2  7 -0.1781961  2  8  0.0661961  2  9  0.1441961  2 10  0.2871961  2 11  0.5381961  2 12  0.496

ftp://ftp.cpc.ncep.noaa.gov/cwlinks/

[Isotherm]

I posted this in the NY forum earlier. A strong correlation exists for strongly negative AO Oct --> negative AO DJF, but not for the reverse scenario.

 

Below is a 500mb composite of the 11 cases in which a < -0.75 October AO occurred, and their ensuing winters. 9/11 or 82% of those cases featured a -AO average for DJF. The glaring failures were 74-75 and 06-07 here. What's interesting is that this correlation does not really exist for > +0.75 AO Octobers as their ensuing winters featured mixed results. At this point in time, it's still early in the season, and generally when we examine -AO or +AO Octobers, there isn't much of a correlation when including the weakly positive or negative cases.

 

[WxUSAF]

Great work, Don.

[Srain]

Houston recorded it's earliest snowfall on record in the winter of 2009-10 when a storm system brought snow to Houston, Lake Charles and Lafayette, LA on December 4th. Great work Don. It will be interesting to see how the winter develops.

[griteater]

Today, the Arctic Oscillation (AO) stood at -4.057. That is the 6th lowest figure on record for October. 2014 is only the second year since 1950 during which the AO fell to -4.000 or below in October. Previously, only the great October blocking episode of 2002 during which the AO stood at -4.000 or below during the October 16-20 saw such an outcome.

 

More importantly, blocky Octobers following summers during which the AO averaged <0 typically saw a blocky winter. Since 1950, there have been 12 cases during which the AO averaged -0.500 or below during October with a summer AO < 0. 10/12 (83%) cases saw the winter AO average < 0.

 

Nice work Don.  Here's a visual to go along with the numbers...

 

[Jonger]

Was that last October, where that massive Aleutian block formed?

[Amped]

Good sign, but the small sample size can be misleading.

[WEATHER53]

Today, the Arctic Oscillation (AO) stood at -4.057. That is the 6th lowest figure on record for October. 2014 is only the second year since 1950 during which the AO fell to -4.000 or below in October. Previously, only the great October blocking episode of 2002 during which the AO stood at -4.000 or below during the October 16-20 saw such an outcome.

 

More importantly, blocky Octobers following summers during which the AO averaged <0 typically saw a blocky winter. Since 1950, there have been 12 cases during which the AO averaged -0.500 or below during October with a summer AO < 0. 10/12 (83%) cases saw the winter AO average < 0.

 

Some detailed information from those cases is below:

 

 

Overall, especially if the ensemble guidance proves accurate through the medium-term, it appears that winter 2014-15 could see a predominantly negative AO.

Don-My analog years for my winter outlook did center around those early 80's winters.

[WEATHER53]

Good sign, but the small sample size can be misleading.

Saying that the sample size is "too small'  is abritray and thus also misleading. The sameple size is what the sample size is.  There can be a case that 1,2,3 examples may come up short but when you have 7,8,9 the idea that "7,8,9 out of the last 50 years is too small a sample size" is baloney.

[Mallow]

Saying that the sample size is "too small'  is abritray and thus also misleading. The sameple size is what the sample size is.  There can be a case that 1,2,3 examples may come up short but when you have 7,8,9 the idea that "7,8,9 out of the last 50 years is too small a sample size" is baloney.

 

Amped didn't say the sample size is "too small", but he's right, a small sample size is something that one should take into account. Don Sutherland understands this, of course, but it's still worth mentioning it I think.

[donsutherland1]

Good sign, but the small sample size can be misleading.

I agree. My use of "hints" in the thread title reflects a reasonable degree of uncertainty. Nothing is cast in stone yet, but I do like how some of the pieces appear to be falling into place.

 

That the OPI mentioned in this forum is also looking favorable right now adds some degree of confidence. Things seem to be leaning toward a potentially blocky winter at this point in time. We'll see where things stand at the end of the month. In an ideal situation, the historical analog set, OPI, and Dr. Cohen's Siberian Snow Advance Index would all be in agreement.

[donsutherland1]

Don-My analog years for my winter outlook did center around those early 80's winters.

Your analog pool from the Mid-Atlantic discussion certainly looked reasonable to me. It will be fun to see how the data falls in place in coming weeks. Hopefully, we'll have a nice winter (cold and snowy for those who don't know my preferences).

[Wow]

16 day avg 500mb ht anomalies 12z GFS.. this takes it to 10/24 - Neg AO/NAO nearly throughout the entire month.  GOA low is a strong signal as well.

 

[Chinook]

Historical note: the month of October 2002, which was a negative AO (-1.489) was similar to the 500mb anomaly maps of yesterday

 

[WxMidwest]

Good stuff Don....

[billgwx]

Historical note: the month of October 2002, which was a negative AO (-1.489) was similar to the 500mb anomaly maps of yesterday

 

october 2002 500mb anomaly.png

 

Funny you mention. Winter 2002-03 has been very much on my mind as a potential analog for the winter to come.

[stadiumwave]

A new atmospheric ENSO index is introduced to WSI clients, showing current state and what this may mean for winter:

 

http://www.wsi.com/blog/energy/introducing-a-new-atmospheric-el-nino-southern-oscillation-index-its-current-state-and-upcoming-winter-implications/

[Typhoon Tip]

It will be interesting to see where the averages for this October end up, considering the entire suite of GFS-derived members is super duper clustered around a new nadior...exceeding -4SD is looming later next week.  

 

I also wonder if there is a correlation coefficient when comparing specific departure values, year-to-year.

[millwx]

A new atmospheric ENSO index is introduced to WSI clients, showing current state and what this may mean for winter:

 

http://www.wsi.com/blog/energy/introducing-a-new-atmospheric-el-nino-southern-oscillation-index-its-current-state-and-upcoming-winter-implications/

 

Yup, and it looks absolutely horrific.  Binning ENSO into "El Nino" and "La Nina", regardless of your index (3.4 SSTA, MEI, their new index), as they do in that write-up is about the worst thing one can do.  Even the "weenies" on this board know that the strength of the ENSO signal can totally alter the pattern (e.g., both strong El Ninos and La Ninas yield warm North American patterns, in general; a weak La Nina up through a low-end moderate El Nino is the best cold signal).

 

To be clear, I'm not saying their new index is bunk.  I have no idea.  Perhaps it does, in fact, capture the atmospheric state - in ENSO terms - better than anything else we have.  But their blog write-up on it, where they simply differentiate between El Nino and La Nina states, is just a horrific misrepresentation of ENSO forcing.  I'd say that any meteorologist should be ashamed to attach their name to that, but, hey, what the heck... I, very sadly, see that done all the time.  I still hear/read that "El Nino = warm winter" all the time, without any regard for the strength of the event.  Sorry to get all revved up over this... but it's perhaps my biggest pet peeve in this entire field.  I'm astonished how many mets go to the ESRL site, look at the LINEAR correlation maps (which get tilted positive because the strong El Ninos outweight the strong La Ninas) and then blindly jump to the conclusion that "El Nino = warm winter".  The correlation is NON-linear.  Arrrgg!  Ok, off my soapbox.

[Mike.Ventrice]

Yup, and it looks absolutely horrific.  Binning ENSO into "El Nino" and "La Nina", regardless of your index (3.4 SSTA, MEI, their new index), as they do in that write-up is about the worst thing one can do.  Even the "weenies" on this board know that the strength of the ENSO signal can totally alter the pattern (e.g., both strong El Ninos and La Ninas yield warm North American patterns, in general; a weak La Nina up through a low-end moderate El Nino is the best cold signal).

 

To be clear, I'm not saying their new index is bunk.  I have no idea.  Perhaps it does, in fact, capture the atmospheric state - in ENSO terms - better than anything else we have.  But their blog write-up on it, where they simply differentiate between El Nino and La Nina states, is just a horrific misrepresentation of ENSO forcing.  I'd say that any meteorologist should be ashamed to attach their name to that, but, hey, what the heck... I, very sadly, see that done all the time.  I still hear/read that "El Nino = warm winter" all the time, without any regard for the strength of the event.  Sorry to get all revved up over this... but it's perhaps my biggest pet peeve in this entire field.  I'm astonished how many mets go to the ESRL site, look at the LINEAR correlation maps (which get tilted positive because the strong El Ninos outweight the strong La Ninas) and then blindly jump to the conclusion that "El Nino = warm winter".  The correlation is NON-linear.  Arrrgg!  Ok, off my soapbox.

 

Well thanks for being straightforward and brutally honest about your thoughts. I agree with you in the sense that binning El Nino and La Nina events is a poor thing to do, as ENSO is a spectrum, not A or B. The purpose of the index was to identify times where the low-frequency forcing is coming from the West Pacific (hence why I call La Nina), verses the East Pacific (El Nino). It's just a label, but may be more beneficial to put West Pacific where the "La Nina" string is and East Pacific where the "El Nino" string is.  This analysis is a bit more in depth than your typical ESRL analysis but the methodology is simple, I get that. The difference is the index is completely independent of what is occurring in the Ocean, and only cares what is going on in the atmosphere. Thanks again for your thoughts.

[Typhoon Tip]

Well thanks for being straightforward and brutally honest about your thoughts. I agree with you in the sense that binning El Nino and La Nina events is a poor thing to do, as ENSO is a spectrum, not A or B. The purpose of the index was to identify times where the low-frequency forcing is coming from the West Pacific (hence why I call La Nina), verses the East Pacific (El Nino). It's just a label, but may be more beneficial to put West Pacific where the "La Nina" string is and East Pacific where the "El Nino" string is.  This analysis is a bit more in depth than your typical ESRL analysis but the methodology is simple, I get that. The difference is the index is completely independent of what is occurring in the Ocean, and only cares what is going on in the atmosphere. Thanks again for your thoughts.

 

Just one lowly Met opinion here, but I have wondered for a long while why there was not a residence type teleconnector, and this "sounds" on the surface to be an attempt at that; and therefore intrigues me.  

 

What I mean by that is simply whether the atmosphere, subsequent of an oceanic-atmospheric coupled model suggestion, is really being usurped. Sometimes it more than merely appears that a given warm ENSO is having an effect.  I have not seen a "Pineapple Express" like that which took place in 1983 since, and that one (as I later read research) was keyed into a warm ENSO leading that winter.  Well that's happened since ...

 

It's like maybe a model could be derived that suggests a "forcibility index", or a given ENSO's ability to have much influence. There are other Global factors that can mute it.  Just like a damped out transient MJO wave space; only in the larger paradigm and extending over a longer period of time.   

[DTWXRISK]

100% strongly agree 
 

Yup, and it looks absolutely horrific.  Binning ENSO into "El Nino" and "La Nina", regardless of your index (3.4 SSTA, MEI, their new index), as they do in that write-up is about the worst thing one can do.  Even the "weenies" on this board know that the strength of the ENSO signal can totally alter the pattern (e.g., both strong El Ninos and La Ninas yield warm North American patterns, in general; a weak La Nina up through a low-end moderate El Nino is the best cold signal).

 

To be clear, I'm not saying their new index is bunk.  I have no idea.  Perhaps it does, in fact, capture the atmospheric state - in ENSO terms - better than anything else we have.  But their blog write-up on it, where they simply differentiate between El Nino and La Nina states, is just a horrific misrepresentation of ENSO forcing.  I'd say that any meteorologist should be ashamed to attach their name to that, but, hey, what the heck... I, very sadly, see that done all the time.  I still hear/read that "El Nino = warm winter" all the time, without any regard for the strength of the event.  Sorry to get all revved up over this... but it's perhaps my biggest pet peeve in this entire field.  I'm astonished how many mets go to the ESRL site, look at the LINEAR correlation maps (which get tilted positive because the strong El Ninos outweight the strong La Ninas) and then blindly jump to the conclusion that "El Nino = warm winter".  The correlation is NON-linear.  Arrrgg!  Ok, off my soapbox.

[GaWx]

Today, the Arctic Oscillation (AO) stood at -4.057. That is the 6th lowest figure on record for October. 2014 is only the second year since 1950 during which the AO fell to -4.000 or below in October. Previously, only the great October blocking episode of 2002 during which the AO stood at -4.000 or below during the October 16-20 saw such an outcome.

 

More importantly, blocky Octobers following summers during which the AO averaged <0 typically saw a blocky winter. Since 1950, there have been 12 cases during which the AO averaged -0.500 or below during October with a summer AO < 0. 10/12 (83%) cases saw the winter AO average < 0.

 

Some detailed information from those cases is below:

 

 

Overall, especially if the ensemble guidance proves accurate through the medium-term, it appears that winter 2014-15 could see a predominantly negative AO.

 

Don/folks,

Very nice work, Don, as usual.

1. One thing that I think can be easily overlooked by some is the -AO bias for DJF. For all DJF's since 1950-1, the mean AO has been -0.35 and the median has been -0.41 as opposed to near 0. So, the index is sort of tilted in the negative direction. So, a DJF's average AO of, say, -0.25 would actually mean that winter would be favored to get slightly less AO type blocking than average. What I'm getting at is that I feel that a DJF AO # of ~-0.40 should be treated as neutral rather than 0.

 

2. Related to this -AO bias, 62% of DJF months 1950-2014 had a -AO and only 38% had a +AO. Also, 33% of DJF months had an AO under -1 while only 17% had it over +1.

 

3. Even after taking this bias into account, the DJF mean of -0.779 in Don's analysis still suggests a pretty good -AO signal vs. -0.35. However, the median of -0.414 is totally neutral. The % of winter months less than 0 (75%) is still a pretty good signal vs. 62% though the 39% for -1 or lower seems pretty modest to me vs. 33%.

 

4. I realize the advantage of looking at Oct. vs. Nov. since it gives one more month of lead time in analyzing winter AO probabilities. However, and probably not surprisingly, Nov. gives a much stronger signal than Oct. Whereas the mean is -0.75 (vs. -0.35) and the median is -0.50 (vs. -0.41) for the DJF AO when the preceding Oct. AO is <-1, the mean is a whopping -1.14 (vs. -0.35) and the median is a whopping -1.25 (vs. -0.41) for the DJF AO when the preceding Nov. AO is <-1. So, the correlation with November is much stronger than that with October.

 

 I forgot to mention that the actuals as well as the GEFS mean forecast out two weeks strongly suggest a good bit better than 50% chance for a sub -1 Oct AO as I'm estimating a sub -1 average for 10/1-27 based on it.