My thoughts about winter

[okie333]

The NAO has some persistence to it from winter to winter, and the Atlantic SST's are where it "saves" that persistence during the spring and summer. (Note that we are now in September, so the files have been loaded already, so to say... the current SST's are of no consequence). Granted, the persistence is fairly weak, but going with the same state as last winter is still slightly better than flipping a coin, given no other information... it is even better, though not overwhelmingly so, if the SSTs of the spring and summer match the pattern from the last winter. This year the April-August SST's are pointing negative once again, yet not as overwhelmingly so as last year. Sunspot area, which is what I go by (since it is a more reliable indicator of solar activity than sunspot number and the record goes farther back than 10.7 flux or Ap index), has gone up, but is still well below every cycle (since 1874) except for #14 (1910's... during which there was a pretty epic string of winters if COOP data is to be believed). What interests me more, though, is the 3-5 year lag in the long-term pattern... the magnetism had a step change in October 2005, and it wasn't until spring 2009 that we really started to see the effects on the AO. If this is true, then next winter corresponds to an even lower point in the magnetic field than last winter did, and it corresponds to the very depths of the SSN/10.7 minimum. Right now we are at the top of the 6-7 month pseudocycle that the Sun has experienced lately... I forsee a drop in solar activity starting in October and continuing through at least January, with the sunspot number bottoming out in January or February before rising again. Not quite as well-placed as 2010-2011's drop, but still a pretty good time to have one. Oh, and to anyone using the Holton-Tan relationships, you need to still count this as a solar minimum... it is getting closer to an in-between, but I don't think we will see true "maximum" (by late 20th century standards) conditions for a long time (at least 20 years, possibly much longer). Based on multiple factors (including the smashed-record --NAO summer), I believe there will be a -NAO block averaging about -1.5. It will be overwhelmingly west-based... the indicators for a WEST-BASED -NAO are nearly off the charts. Based on the same summer correlation (which is much higher at .35 for the AO [it was around .1 for the NAO]), and the summer AO, which was third only to 2009-2010 and 1958-1959, I think the AO will average around -2.5. The EPO is definitely the teleconnection to watch... based on the +OLR in the Indian Ocean, the EPO should be raging negative... but it's not, currently. However, when the genuine wildcard is whether there will be a double-block or a very rare triple-block dominant pattern, things are still looking up... especially because the EPO is notoriously difficult to predict.

EDIT: Also, a -PNA average is pretty much a given. The STJ will be more active than last winter but less so than 2009-2010. That leaves plenty of room for error lol. Something tells me the STJ will be more active than the borderline weak/moderate Niña would suggest, though.

tl;dr: It's gonna be cold.

[Quixotic1]

The NAO has some persistence to it from winter to winter, and the Atlantic SST's are where it "saves" that persistence during the spring and summer. (Note that we are now in September, so the files have been loaded already, so to say... the current SST's are of no consequence). Granted, the persistence is fairly weak, but going with the same state as last winter is still slightly better than flipping a coin, given no other information... it is even better, though not overwhelmingly so, if the SSTs of the spring and summer match the pattern from the last winter. This year the April-August SST's are pointing negative once again, yet not as overwhelmingly so as last year. Sunspot area, which is what I go by (since it is a more reliable indicator of solar activity than sunspot number and the record goes farther back than 10.7 flux or Ap index), has gone up, but is still well below every cycle (since 1874) except for #14 (1910's... during which there was a pretty epic string of winters if COOP data is to be believed). What interests me more, though, is the 3-5 year lag in the long-term pattern... the magnetism had a step change in October 2005, and it wasn't until spring 2009 that we really started to see the effects on the AO. If this is true, then next winter corresponds to an even lower point in the magnetic field than last winter did, and it corresponds to the very depths of the SSN/10.7 minimum. Right now we are at the top of the 6-7 month pseudocycle that the Sun has experienced lately... I forsee a drop in solar activity starting in October and continuing through at least January, with the sunspot number bottoming out in January or February before rising again. Not quite as well-placed as 2010-2011's drop, but still a pretty good time to have one. Oh, and to anyone using the Holton-Tan relationships, you need to still count this as a solar minimum... it is getting closer to an in-between, but I don't think we will see true "maximum" (by late 20th century standards) conditions for a long time (at least 20 years, possibly much longer). Based on multiple factors (including the smashed-record --NAO summer), I believe there will be a -NAO block averaging about -1.5. It will be overwhelmingly west-based... the indicators for a WEST-BASED -NAO are nearly off the charts. Based on the same summer correlation (which is much higher at .35 for the AO [it was around .1 for the NAO]), and the summer AO, which was third only to 2009-2010 and 1958-1959, I think the AO will average around -2.5. The EPO is definitely the teleconnection to watch... based on the +OLR in the Indian Ocean, the EPO should be raging negative... but it's not, currently. However, when the genuine wildcard is whether there will be a double-block or a very rare triple-block dominant pattern, things are still looking up... especially because the EPO is notoriously difficult to predict.

EDIT: Also, a -PNA average is pretty much a given. The STJ will be more active than last winter but less so than 2009-2010. That leaves plenty of room for error lol. Something tells me the STJ will be more active than the borderline weak/moderate Niña would suggest, though.

tl;dr: It's gonna be cold.

Nice write up. Why do you think the STJ will be more active?

[JoMo]

dude, I totally need some graphics or something. lol

Interesting write-up.

[SnowGoose69]

Nice write up. Why do you think the STJ will be more active?

I think he's leaning towards that idea because its a weaker Nina. Unfortunately though in my experience the 2nd year of many Ninas often sees a less active STJ. The reason for this is likely the fact that often a moderate-strong Nina is following an El Nino so sometimes the atmospheric memory of the previous ENSO state is not totally wiped out (This is likely why 98-99's temp profile in the N Plains looked more El Nino like). By the 2nd year of the La Nina you often see a true lack in the STJ. 2007-2008 was a La Nina but followed a seemingly endless string of El Ninos and there was a surprisingly active STJ in December and even to an extent beyond that with the massive southern Plains/MW ice storm and 2 other relatively big snow events in southern KS and MO (Wichita I believe had their snowiest December on record). 2008-2009 however saw a more classic La Nina pattern in areas which rely on the STJ. The one positive for those STJ dependent areas is that most La Ninas have had a big snow event at the end of the winter in those areas like KS, MO, TX, OK, AR which usually do not do too well in La Nina winters. Its almost as much an assured thing as the big late season snowstorm in March or April is in the Northeast preceding the formation of a moderate-strong El Nino.

[John S2]

.. the magnetism had a step change in October 2005

Would anyone be able to point me to a data source for this?

[okie333]

Would anyone be able to point me to a data source for this?

[BethesdaWX]

I think he's leaning towards that idea because its a weaker Nina. Unfortunately though in my experience the 2nd year of many Ninas often sees a less active STJ. The reason for this is likely the fact that often a moderate-strong Nina is following an El Nino so sometimes the atmospheric memory of the previous ENSO state is not totally wiped out (This is likely why 98-99's temp profile in the N Plains looked more El Nino like). By the 2nd year of the La Nina you often see a true lack in the STJ. 2007-2008 was a La Nina but followed a seemingly endless string of El Ninos and there was a surprisingly active STJ in December and even to an extent beyond that with the massive southern Plains/MW ice storm and 2 other relatively big snow events in southern KS and MO (Wichita I believe had their snowiest December on record). 2008-2009 however saw a more classic La Nina pattern in areas which rely on the STJ. The one positive for those STJ dependent areas is that most La Ninas have had a big snow event at the end of the winter in those areas like KS, MO, TX, OK, AR which usually do not do too well in La Nina winters. Its almost as much an assured thing as the big late season snowstorm in March or April is in the Northeast preceding the formation of a moderate-strong El Nino.

Does ENSO spiking warm this summer have any impact here? Summer 2010 & Summer 2011 were very similar with minor differences in timing...I cannot find any 1st yr Nina & 2nd yr Nina that have had summers all that close to one another.

[SnowGoose69]

Does ENSO spiking warm this summer have any impact here? Summer 2010 & Summer 2011 were very similar with minor differences in timing...I cannot find any 1st yr Nina & 2nd yr Nina that have had summers all that close to one another.

1998 and 1999 had some similarities in the summer temp anomalies but outside of that you're pretty much correct, there was little similarity between the summers of 54/55, 70/71, 73/74, 84/85, 88/89. At least in the south and east 98 and 99 were close. 1999 was the only 2nd year La Nina summer since 1950 where Texas was above normal and now 2011 joins that party. Typically 2nd year La Nina summers are cool in the south in the very areas which were ridiculously warm this year. The composite of June-August 1955, 1985, 1974, 1989, and 1999 across places like TX is -1 to -2 below normal...and if you remove 1999 its closer to -3 or -4.

[Roy]

Snow, that's interesting. Do you think that has implications for how Nina will interact with the south so to speak as we get closer to winter? There is a ray of hope for a more rainy pattern in West Texas this Thursday/Friday finally, the best chance of rain in a long time. Unfortunately, it won't help the rest of the state.

[tacoman25]

1998 and 1999 had some similarities in the summer temp anomalies but outside of that you're pretty much correct, there was little similarity between the summers of 54/55, 70/71, 73/74, 84/85, 88/89. At least in the south and east 98 and 99 were close. 1999 was the only 2nd year La Nina summer since 1950 where Texas was above normal and now 2011 joins that party. Typically 2nd year La Nina summers are cool in the south in the very areas which were ridiculously warm this year. The composite of June-August 1955, 1985, 1974, 1989, and 1999 across places like TX is -1 to -2 below normal...and if you remove 1999 its closer to -3 or -4.

However, the 1954-56 multi-year Nina summer composite shows a similar temperature pattern to summer 2011. 1954 and 1956 were both quite warm in the southern plains.

[BethesdaWX]

However, the 1954-56 multi-year Nina summer composite shows a similar temperature pattern to summer 2011. 1954 and 1956 were both quite warm in the southern plains.

We're looking at individual summers and comparing them, what is the point of mashing them all together? It is irrelavent (though there are better analogs than the winters in the 1950s anyway). Point is, summer 2010 & summer 2011 have been very similar to one another, perhaps moreso than any other back-to-back La Nina years.

That might be due to the weaker sun activity continuing magnetically, but even at the state of global SST anomalies there is very little difference in the positioning, PDO looks similar, as does the atlantic, IPO, and IOD. That is a sign that the global atmospheric state has not changed much.

[okie333]

All years where June, July, and August were all negative in both the AO and NAO:

[BethesdaWX]

All years where June, July, and August were all negative in both the AO and NAO:

I can't see how the Pacific Northwest will end up above average this winter, though oddly enough the best analogs (in my personal view), 1967-68, 1903-04, 1917-18, and 2008-09 all had a heavy positive along the west coast for a good portion of the winter.

[tacoman25]

I can't see how the Pacific Northwest will end up above average this winter, though oddly enough the best analogs (in my personal view), 1967-68, 1903-04, 1917-18, and 2008-09 all had a heavy positive along the west coast for a good portion of the winter.

Why do you think this analog is better than say 1956-57 or 1974-75?

[okie333]

Two years that fit the most according to the summer AO and NAO:

[tacoman25]

Two years that fit the most according to the summer AO and NAO:

I would love to see a repeat of either of those winters (Boulder had 112" in 1958-59 and a record 134" in 2009-10), but I don't think we can ignore the major PDO/ENSO differences that 2011 has.

[BethesdaWX]

Why do you think this analog is better than say 1956-57 or 1974-75?

The -PDO, second year Nina at similar strength to current, the weaker Sun activity magnetically, and a QBO of similar structure, not deep in all levels, puts it ahead of the 1950s in my opinion. Blocking at the high lattitudes is, in my opinion, where to look to after determining the ENSO/PDO phase.

[tacoman25]

The -PDO, second year Nina at similar strength to current, the weaker Sun activity magnetically, and a QBO of similar structure, not deep in all levels, puts it ahead of the 1950s in my opinion. Blocking at the high lattitudes is, in my opinion, where to look to after determining the ENSO/PDO phase.

1. As far as the PDO goes, 1967 is a decent match, though 1956 will be closer.

2. 1967 didn't have notably weak sun activity like the last few years.

3. 1967 followed a weak, negative neutral winter. 2011 is following a moderate/strong Nina. 1956 and 1974 are better matches in that respect.

[BethesdaWX]

1. As far as the PDO goes, 1967 is a decent match, though 1956 will be closer.

2. 1967 didn't have notably weak sun activity like the last few years.

3. 1967 followed a weak, negative neutral winter. 2011 is following a moderate/strong Nina. 1956 and 1974 are better matches in that respect.

1) The exact PDO number at the current timeframe isn't very important, the basic phase (structure-wise too) matters more as the number will change, the raging -PDO we had in August has backed off somewhat.

2) The Magnetic aspect was much closer in 1967-68 than 1956-57, no debate there, we're a bit lower still at this time but nowhere near as high as 1956-57.

3) The QBO matches up better in 1967-68

4) 1966-67 was about the same strength overall as 2008-09 was, but the consitions persisted a month or two longer, a 2nd year Nina-heavy atmosphere is what we have at this point, and had then (In 1967-68), the minute differences won't have a large scale impact the year after relative to other drivers/factors that will make any effect unmeasurable.

I'm not sure why you're arguing 1956-57 really, you're pointing out small flaws in 1967-68 that are minor relative to 1956-57, you seem to favor the 1950s which is ok but you can't trash analogs that are similar, or if anything, better.

[tacoman25]

4) 1966-67 was about the same strength overall as 2008-09 was, but the consitions persisted a month or two longer, a 2nd year Nina-heavy atmosphere is what we have at this point, and had then (In 1967-68), the minute differences won't have a large scale impact the year after relative to other drivers/factors that will make any effect unmeasurable.

I'm not sure why you're arguing 1956-57 really, you're pointing out small flaws in 1967-68 that are minor relative to 1956-57, you seem to favor the 1950s which is ok but you can't trash analogs that are similar, or if anything, better.

What does 2008-09 have to do with 1966-67? 1966-67 was much weaker -ENSO than 2010-11, that's my point.

[BethesdaWX]

What does 2008-09 have to do with 1966-67? 1966-67 was much weaker -ENSO than 2010-11, that's my point.

I'm drawing a relative comparison, 2008-09 is considered La Nina by many, 1966-67 was also -GLAAM dominant.

My point, you're dramatically overstating the importance of the ENSO phase in the previous year to a SECOND YEAR La Nina. In this case, a second year La Nina is a second Year La Nina, El Nino leftover is one thing before a 1st year La Nina but you won't see the effect in the second year 2 years after a moderate El Nino, at least not significantly, atmospheric memory of one event usually doesn't last 2 years.

[uncle W]

usually October or November gives hints to where the main storm track will set up...These are precipitation maps for October and November before NYC's snowiest winters...

[SnowGoose69]

usually October or November gives hints to where the main storm track will set up...These are precipitation maps for October and November before NYC's snowiest winters...

That one does work pretty well for determining snow but necessarily cold. If you plot some of the less snowy winters in NYC you see the axis of heavy precip in Oct/Nov runs from around LA/MS NNE'ward across TN, KY, IL, IN, and MI.

[icebreaker5221]

Two years that fit the most according to the summer AO and NAO:

The 4 years (since 1950) that featured both June and July below -1.00 NAO were 1958, 2008, 2009 and 2011. Note that all 4 of these years featured a negative AO for all 3 months of JJA.

However, only 3 years have featured June, July and August ALL below -1.0 NAO: 1958, 2008 and 2011.

(2008 was also a better ENSO match than 2009, but that's an aside.)

Long story short, I wouldn't necessarily throw out 09-10 as an analogue, but I would definitely throw in 08-09 when using NAO and AO.

Edit: doing so relaxes the cold in the south and east and the warmth in the west a bit. Takes the bite out of the central US cold, keeps the north central cold.

[okie333]

The 4 years (since 1950) that featured both June and July below -1.00 NAO were 1958, 2008, 2009 and 2011. Note that all 4 of these years featured a negative AO for all 3 months of JJA.

However, only 3 years have featured June, July and August ALL below -1.0 NAO: 1958, 2008 and 2011.

(2008 was also a better ENSO match than 2009, but that's an aside.)

Long story short, I wouldn't necessarily throw out 09-10 as an analogue, but I would definitely throw in 08-09 when using NAO and AO.

Edit: doing so relaxes the cold in the south and east and the warmth in the west a bit. Takes the bite out of the central US cold, keeps the north central cold.

I was basing this off the fact that those two years (along with 2011) are way ahead of all others when it comes to JJA combined totals on a renormalized AO/NAO composite. 2008 did not have nearly as strong of a AO. The AO has a much larger summer/winter correlation (about .35) than the NAO does (about .1), so mentioning the summer NAO without mentioning the summer AO is silly.

[tacoman25]

The 4 years (since 1950) that featured both June and July below -1.00 NAO were 1958, 2008, 2009 and 2011. Note that all 4 of these years featured a negative AO for all 3 months of JJA.

However, only 3 years have featured June, July and August ALL below -1.0 NAO: 1958, 2008 and 2011.

(2008 was also a better ENSO match than 2009, but that's an aside.)

Long story short, I wouldn't necessarily throw out 09-10 as an analogue, but I would definitely throw in 08-09 when using NAO and AO.

Edit: doing so relaxes the cold in the south and east and the warmth in the west a bit. Takes the bite out of the central US cold, keeps the north central cold.

What are the top 4 -ENSO summers (or summers that preceded -ENSO winters) in terms of -NAO? Well, obviously 2008 is number one, how about the others?

[uncle W]

We might see another Neg A O winter coming up...Since 1950 Aug. 2011 has the third lowest AO index...all the analogs below had a lower AO value in the winter months.....

year...Aug.AO index...Winters low value

1977.....-1.412.....-3.014 Feb...

1964.....-1.207.....-2.084 Feb...

2011.....-1.063.....-???

1960.....-1.008.....-1.506 Jan...

1950.....-0.851.....-1.928 Dec...

1987.....-0.836.....-1.066 Feb...

1986.....-0.826.....-1.473 Feb...

[SnowGoose69]

What are the top 4 -ENSO summers (or summers that preceded -ENSO winters) in terms of -NAO? Well, obviously 2008 is number one, how about the others?

To burst everyone's bubble, the summer of 1974 had all 3 months all negative though the most negative out of the 3 was -0.76 in August. Still 74-75 was pretty miserable for the eastern part of the country, and that was even seeing the Sept/Oct reversal to a +NAO which is usually a good sign. The summer of 1998 also had a strongly negative NAO in June and July of -2.70 and nearly -0.50. August was more or less dead neutral. The NAO though stayed negative in 1998 up until November when it went positive the rest of the winter.

[okie333]

To burst everyone's bubble, the summer of 1974 had all 3 months all negative though the most negative out of the 3 was -0.76 in August. Still 74-75 was pretty miserable for the eastern part of the country, and that was even seeing the Sept/Oct reversal to a +NAO which is usually a good sign. The summer of 1998 also had a strongly negative NAO in June and July of -2.70 and nearly -0.50. August was more or less dead neutral. The NAO though stayed negative in 1998 up until November when it went positive the rest of the winter.

From my last post, "The AO has a much larger summer/winter correlation (about .35) than the NAO does (about .1), so mentioning the summer NAO without mentioning the summer AO is silly."

[nzucker]

To burst everyone's bubble, the summer of 1974 had all 3 months all negative though the most negative out of the 3 was -0.76 in August. Still 74-75 was pretty miserable for the eastern part of the country, and that was even seeing the Sept/Oct reversal to a +NAO which is usually a good sign. The summer of 1998 also had a strongly negative NAO in June and July of -2.70 and nearly -0.50. August was more or less dead neutral. The NAO though stayed negative in 1998 up until November when it went positive the rest of the winter.

1998 and the following years are bad analogs because that El Niño had such an overwhelming impact on the atmosphere, including the substantial warming of global temperatures/heights it produced, and we were also in much more of a +PDO phase.