raindancewx

I was curious to see if the enhancement area you get from subtracting the "top -NAO winters" from the "top +NAO winters" for 1950-51 to 2020-21 resembled the snow pattern last year. The most enhancement is for the deep south to be sure. But the cold push was severe enough last year that the biggest enhancement was in places that see very little snow. You can interpret this as "Chicago averages 46% more snow in the top -NAO winters since 1950 compared to the top +NAO winters since 1950". Pretty sure the light blue area extends into northern Michigan, SE Canada, and northern New England, but I didn't really look. The deep blue area likely goes deeper into the South, but those places often average under five inches of snow or see snow in less than 70% of years, so the data gets weird real fast. My guess is if we have another winter with blocking, the snow would more closely resemble my map than the 2020-21 snow observations. Not a huge fan of the NAO as a temperature indicator for most of the US overall but it does favor snowy conditions in most spots nationally when negative. That's the main reason I look at it. My map is comparing total July-June snow in the years that had -NAO v. +NAO winters, i.e. not just the snow that fell in winter. Both sets of years include a number of warm and cool ENSO events, so shouldn't be a huge factor.

https://web.archive.org/web/20110902042909/https://www.cpc.ncep.noaa.gov/products/precip/CWlink/MJO/enso.shtml Hmm

Nino1+2 Nino3 Nino34 Nino4 Week SST SSTA SST SSTA SST SSTA SST SSTA 21JUL2021 22.1 0.6 25.2-0.4 26.8-0.5 28.5-0.4 28JUL2021 22.1 0.8 25.1-0.4 26.7-0.4 28.7-0.1 04AUG2021 21.7 0.7 25.1-0.2 26.7-0.3 28.6-0.2 11AUG2021 20.9 0.1 24.8-0.4 26.5-0.4 28.6-0.2 12AUG2020 19.9-0.9 24.8-0.4 26.4-0.6 28.3-0.5 Still behind last year, especially in Nino 1.2 and Nino 4. Subsurface warmed up a bit this week too on the weekly ENSO PDF the CPC releases each week. The subsurface is still a bit warmer than last year too. For my purposes, 'flat Neutral' in Nino 3.4 is 26.65C in August. So we've just crossed over to cold-Neutral in mid-August. SOI has also collapsed from over +16 in July to about +4 for the past 30 days. +4 or -4 in August is pretty reliable for what the next winter will do back to the 1930s - it's real close right now.

Warmth is pushing east from the west at all depths.

Fairly optimistic for a wet Fall-Spring for parts of the West with the PDO currently trending positive (rapidly cooling SSTs east of Japan). My research shows Nino 1.2 leads PDO base state changes if it is "counter" the PDO base state heading into the cold season. Nino 1.2 has been pretty warm. We'll see. This is pretty similar to how the PDO behaved in late 2013 though from the looks of it.

This tends to happen when Nino 1.2 is warm in cold ENSO late Summer/Fall - see that huge cold trend east of Japan? That ain't the negative PDO folks. Watch the north pacific.

The Australians think the Indian Ocean Dipole should weaken pretty quickly later this year into winter. Not necessarily a great sign for a big La Nina to continue late winter and spring. Negative IOD is generally the La Nina amplifier/signal. The super +IOD of 2019-20 coincided with a lot of the weird weather that fall/winter, and then it died super fast.

CFS has La Nina for Oct-Dec, cold-Neutral Jan, then back to pretty warm shortly after.

I'm more interested in the development and future decay of the event than the actual peak strength. The CFS and Canadian have different ideas with regard to structure. It's a shame the Jamstec isn't back yet. That thing had a good read on SST configurations. I am not trying to setup some super cold La Nina winter when I list the years that have rare outcomes. It's just stuff I look at along the way to winter. To me the big issue with 1967 is simply how cold it gets in Nino 3 relative to the other zones. It's a weird setup in an ENSO sense. I do think a top-five cold December is possible with regard to the 1991-2020 period nationally. A lot of years in the 1960s like December 1960, December 1963 and December 1967 are actually much colder than that. If you just went by current recency profiles in SSTs and US temps, something like 1967, 2011, 2013, 2017 is a decent start for winter. The hurricane season will provide more hints going forward. I don't really worry about the QBO as an index much because it is auto-correlated to the 11-year solar cycle. (~11 years x 12 months = 132 months. x 2 = 264 months) Every ninth 28-29 month QBO cycle should "match" with every second solar cycle by timing: 264 months / 29 months = 9.1 QBO cycles, or "laps" if you think of it as a race to the same spot around a 'track' of what is possible.) More generally, you tend to find good AMO/PDO sst matches at 60 year intervals +/-3 to adjust for ENSO differences I find. So I generally look at years that are 30, 60, 90 years prior (+/-3 per 30 year cycle) and then the years at 11-year intervals first. 1967-->1978-->1989-->2000-->2011-->2022 is pretty close and certainly meets the criteria given that solar cycle length varies from 9-13 years since the SILSO observations started. 1960-->2020 is the 60 year match, and 1960 was also a very hot NW Summer, especially July 1960.

At this point I'm generally looking at 20-30 years for the winter of the last 100. I'm already close to throwing out some La Ninas just based on hurricane activity which doesn't seem to be super amped, or super low in the Atlantic through mid-August. Last year had nine tropical storms by this date, ten by 8/13. A lot of years like 1983 or 1974 or 1933 or 2017 are at one extreme by ACE/total activity. I like to focus on rareness in various signals. So what has been rare this year? - Record +WPO in Jan/Mar (WPO >1.5 in March: 1977, 1985, 1991, 2010, and probably several years in the 1930s) - Record +PNA in August (PNA >1.5 in August: (1958, 1960, 1977, 1990, 1999, 2004, 2007, 2020) - Record heat and cold in the Southwest in Summer (Highs <=70F, AND >=100F June in ABQ: (0 but 1937 is closest) - Unusual heat Northwest in Summer (1960, 1967, 2013) - Rapidly Rising solar activity in a La Nina (1955, 1956, 1988, 1998, 1999, 2010, 2011) with many near Ninas too. Which years show up at least twice? 1960, 1977, 1999, 2010. That blend kind of looks like a 1967/2013, and those years have been decent SST matches.

Presumably no one is that interested in the record rain, and cold weather we've had at some of these sites since the brief June heat wave. Traditionally, the monsoon comes on strong in July when June is very hot. You had multiple days in Phoenix with highs in the low 80s in July and the month finished cold against the current normals. In the totality of observations, this will probably go down as a near average Summer for heat in the Southwest, with both intense high heat and cold highs, with the wetness as the more important feature. It's silly to pretend that the individual days of record shattering heat or unusual coldness (I had highs in the 60s in late June shortly after the near record heat, basically tied for coldest ever for 6/16-7/15) are more important than the overall averages if you're talking about climate. I understand it's very warm in the Northwest. But that's more relevant in a climate sense since that heat has persisted unusually long. Whereas our heat is pretty normal - highs are a bit cold overall if anything locally. 1/1-8/9 90F highs in Albuquerque for 1931-2021. 1 1994-08-09 69 0 2 2011-08-09 68 0 3 1980-08-09 67 0 4 1958-08-09 66 0 5 2012-08-09 63 0 - 1956-08-09 63 0 7 2020-08-09 61 0 - 2018-08-09 61 0 9 2001-08-09 60 0 - 1953-08-09 60 0 11 2016-08-09 59 0 12 1996-08-09 58 0 - 1969-08-09 58 0 14 1983-08-09 57 0 - 1978-08-09 57 0 - 1977-08-09 57 0 - 1952-08-09 57 0 - 1934-08-09 57 0 19 2005-08-09 56 0 - 2000-08-09 56 0 - 1981-08-09 56 0 - 1974-08-09 56 0 - 1964-08-09 56 0 - 1951-08-09 56 0 - 1942-08-09 56 0 26 2003-08-09 55 0 - 1993-08-09 55 0 - 1984-08-09 55 0 - 1979-08-09 55 0 - 1963-08-09 55 0 - 1954-08-09 55 0 32 1970-08-09 54 0 - 1959-08-09 54 0 - 1947-08-09 54 0 35 1966-08-09 53 0 - 1961-08-09 53 0 - 1960-08-09 53 0 - 1946-08-09 53 1 39 2017-08-09 52 0 - 1982-08-09 52 0 41 1989-08-09 51 0 - 1938-08-09 51 0 - 1936-08-09 51 0 44 2019-08-09 50 0 - 1987-08-09 50 0 - 1943-08-09 50 0 47 2014-08-09 49 0 - 2013-08-09 49 0 - 2002-08-09 49 0 - 1957-08-09 49 0 51 2021-08-09 47 0 - 2006-08-09 47 0 - 1995-08-09 47 0 - 1972-08-09 47 0 - 1945-08-09 47 0 - 1937-08-09 47 0 57 1939-08-09 46 1 58 1948-08-09 45 0 59 2008-08-09 44 0 - 2004-08-09 44 0 - 1975-08-09 44 0 62 2010-08-09 43 0 - 2009-08-09 43 0 - 1990-08-09 43 0 - 1967-08-09 43 0 - 1933-08-09 43 0 67 1935-08-09 42 0 68 1976-08-09 41 0 - 1968-08-09 41 0 - 1962-08-09 41 0 - 1950-08-09 41 0 - 1931-08-09 41 0 73 2015-08-09 40 0 - 1973-08-09 40 0 - 1971-08-09 40 0 - 1949-08-09 40 0 - 1940-08-09 40 0 78 2007-08-09 39 0 - 1965-08-09 39 0 - 1955-08-09 39 0 - 1944-08-09 39 0 - 1932-08-09 39 0 83 1985-08-09 38 0 84 1998-08-09 37 0 - 1997-08-09 37 0 - 1992-08-09 37 0 - 1991-08-09 37 0 - 1988-08-09 37 0 89 1986-08-09 33 0 90 1941-08-09 30 0 91 1999-08-09 20 0

There were a lot of signs last Fall that the pattern was going to be pretty special at times. 2007-08 never matched up super well for timing in the winter, but you did have a massive cold dump into the US from 1/16-2/15 that doesn't really show up if you just look at Jan/Feb individually. Main difference without looking is the ridging in the SE pushed it somewhat further West than in February 2021. I was annoyed at the severity of the cold, I thought I was pretty bullish with half the US 5-10 degrees below average for two weeks using the 2007-08 analog as the basis for late winter cold. We went though this last year too - but the cold ENSOs after two El Ninos all tend to feature several incredible cold snaps nationally in winter and spring. Last year was no exception. You have amazingly potent cold snaps in 1931-32, 1959-60, 1970-71, 1978-79, 1988-89, even 2016-17 if you know when to look by timing.

The patterns nationally are still reminiscent of 1967 to me - which makes sense given that it is probably the closest SST match for 2021 overall.

PNA over 1.5 in August from 1950-2020: 1958, 1960, 1977, 1990, 1999, 2004, 2007, 2020 if the month remains very positive. Some other close to +1.5 years include 2011 and 2014. PNA+ in August correlations: Sept: How NE US & SW US. Strong correlation to warmth SE Canada & NY state. Oct: Warm signal West. Strong warm signal for CA/AZ/NV. (similar to -WPO in Feb-Apr, although not in Oct) Nov: Warm signal nationally - but weak. (Note last year is on the list above and was very warm in November 2020). Dec: Cold signal for NM/CO/TX, Central Plains, Midwest, NJ/PA - all weak Jan: Warm signal SW US/New England - all weak Feb: Warm signal NW US / cold Northern Plains - all weak Mar: Cold signal Midwest - all weak Apr: Warm signal TX/NM, cold CA - all weak. May: Cold signal North & West, hot signal SE - strongest in Texas. Precip: Sept - erratic at best Oct - weak dry signal Rockies & Texas & Ozarks Nov - weak dry signal most of the US Dec - wet signal SW & Midwest / TN Valley (strong for Northern NM mountains) Jan - erratic Feb - wet signal SW & Plains. Looks like the +WPO signal really. Mar - wet SE US. dry Northern Plains (strong) and Rockies Apr - wet coasts, dry NM/AZ/TX May - wet NW/SE, dry SW

We haven't had it in a while, but there are a lot of older years when the SSTs in Nino 3/3.4 are very cold, but only very briefly. That's my fear for the La Nina consensus - it gets very cold again in Nino 3/3.4 in the Fall, and then it sharply weakens. It's more common for the sea surface temps to get very cold/warm, but only briefly when the peak anomalies are closer to Halloween or Thanksgiving than Christmas (ala the observations of the fisherman per normal ENSO).

One very simple rule for SSTs is that from 1950-2020, you've never had Nino 3.4 drop by more than 1C from September to December-February. We're at 26.7C on the weeklies week one of August. A moderate La Nina would be 25.5C or colder in winter. Will August finish at 26.5C or colder? Possibly. Will September be below 26.5C - I would say so. But keep in mind, even though a 1C drop has been observed from September to winter, most cooling trends from September to Dec-Feb end up far weaker than that in practice. Last September was 25.88C, and the winter was 25.58C. If you use last year as an example, the big spikes up/down in the subsurface tend to lead the surface peaks by 3-6 weeks. So the big drop off now, once it stops, indicates we might see another ENSO peak prior to 11/1. Against the long-term averages, Nino 3.4 was coldest in November 2020 (25.3C v. 26.5C long-term average) after the subsurface was coldest in October.

One thing I like to do this time of year to start to winnow down analogs is to look at objectively good tropical matches in Dec-Feb for the prior winter that are still objectively good matches in July. The winters of 1966-67, 2000-01, 2010-11 were close winter SST matches in the tropics, and 1967, 2001, and 2011 are still close in July. 1966-67 was probably the best of the three overall for US conditions, but we had an exaggerated version of that pattern in winter. Winter 2020-21 was kind of opposite the very cold Dec 2010/2000 pattern, which is I 'knew' would be unlikely heading into last winter. The current July, and August pattern so far pretty is close to 1967 nationally. The blend of 1967-68, 2001-02, 2011-12 is a very weak La Nina in winter that is coldest in Nino 3, warmest in Nino 4, and then weakens sharply in Feb-Mar from the east. You can see the big -PDO signature with very warm water near Japan, and some cold water by the equator in the Atlantic like we've had in recent weeks. I don't think this look is quite right though, and I don't think the US temperature pattern it produces is correct either. We're still not that close to long-term La Nina conditions at the surface. Nino 3.4 needs to be around 26.15C in August to be even -0.5C against 1951-2010 averages. Nino 1.2 remaining warm into October against a -PDO base state in March-August tends to precede a flip to a more +PDO state in Nov-Apr. Something that needs to be watched. Nino1+2 Nino3 Nino34 Nino4 Week SST SSTA SST SSTA SST SSTA SST SSTA 09JUN2021 23.2-0.0 26.7-0.0 27.8-0.0 28.9-0.1 16JUN2021 23.3 0.4 26.0-0.5 27.2-0.5 28.8-0.2 23JUN2021 22.9 0.2 26.3-0.0 27.6-0.0 29.0 0.0 30JUN2021 22.6 0.3 26.2 0.0 27.5-0.1 28.8-0.1 07JUL2021 22.2 0.1 26.0-0.0 27.4-0.0 28.9-0.1 14JUL2021 22.2 0.4 25.9 0.1 27.2-0.1 28.6-0.4 21JUL2021 22.1 0.6 25.2-0.4 26.8-0.5 28.5-0.4 28JUL2021 22.1 0.8 25.1-0.4 26.7-0.4 28.7-0.1 04AUG2021 21.7 0.7 25.1-0.2 26.7-0.3 28.6-0.2 05AUG2020 19.8-1.3 24.6-0.7 26.3-0.7 28.4-0.4 I do think these years are likely to continue to watch well to 2021 in Nino 3.4 through at least December. You can see August should be near the speculative 26.5C going by week one. At 26.1C in September that would be the initiation of the La Nina if the blend isn't too cold. The current big time drop off in the subsurface will likely flatten or reverse a bit in a few weeks, as that water works up to the surface. I think you'll see a pattern fake in late September-October when that happens before a transition in November. My guess is this is another cold ENSO with a relatively early surface peak. The 2020-21 event was much more impressive in the Fall than the Winter by historical standards. Something like a top 1/3 cold event in October or November, and then well below that by the end of the winter.

It's a bit disingenuous to say we're ahead of last year. The warmth from April-June has to wash out fully first. More relevant is how the trend is from last year anyway. The way the similar cold below the surface has arrived in July is completely opposite last year both at depth (the 0-300m readings) and in an 2D East/West, North/South sense. The La Nina of 2020-21 was really an April-March thing in terms of the subsurface. This event is probably an August or September start at this point, pending how long it lasts. The subsurface last year featured several giant rises and dips that coincided with huge changes in the seasonal US weather pattern. It's not a coincidence that the huge cold periods in the central/western US coincided with a flip from declining heat to rising heat both times (Oct and Feb were the flips in trend). One reason I had liked 2007-08 last year was for the double peak in the subsurface, in October and then January, it was very much like 2020-21. The idea that last year was a Moderate La Nina is still kind of silly to me, when the "-1.3" or whatever CPC uses for 2020-21 is based on a much warmer average for Dec-Feb than earlier La Ninas. In terms of actual SSTs, by coldest month or coldest month against constant averages, last year was warmer than roughly half of La Ninas. It's more of a weak event than a moderate one at the surface. Years like 2010-11 and 2011-12 had much more impressive peaks, especially given they followed a much stronger El Nino in 2009-10. The subsurface was colder at the peak of the 2011-12, 2010-11, 2008-09, 2007-08, 1999-00, 1998-99, 1995-96, 1988-89, 1983-84 events compared to 2020-21. That's almost every La Nina since 1980...

Compared to last year, Nino 3 (and 1.2 also) are still much warmer at the surface, while Nino 4 is colder. The ONI values / raw SST data for July updated, so the matches above are the closest objective matches. I then compared the observations to the 1951-2010 average for the blend, and 2021 itself. It won't last long in Nino 3.4, but 1986 is currently the top match for the three zones. The SOI/SST matches I've shown recently, including 1967, 2011, 2013, 1979, etc, have all shown up again. These years are generally cold Augusts, so it will be interesting to see how that plays out.

These are the top SOI matches to 2021 for May-July from 1931-2020 (+3.9, +0.0, +16.3 in 2021 for each month respectively). I had used Spring 1979 as an analog for Spring 2021, given that the cold-ENSO 1978-79 winter followed two El Ninos near/following the solar minimum (1976-77, 1977-78) and had ~record +NAO/AO values in November before transitioning to a huge cold period in late winter. NAO was +3.0 or something insane in Nov 1978 before dying later on. Was pretty similar in 2020 and I knew that by the time I did the Spring Outlook in February. The 7th closest SOI match (really a ~ three-way tie between 1936, 1939, and 1960) is 1960, and the 1959-60 winter was also very similar at times (Dec/Feb) for temp patterns nationally, and 1960 has worked sporadically as a match in Spring/Summer 2021. If you don't know, 1936 and 2011 are both very hot Summers nationally, so interesting to see them both show up as top matches. July 1960 had incredible heat in the Northwest, similar to 2021. The "look" that is supposed to happen in August nationally is fairly similar to August 1939, 1936 blended toward 1939, and then maybe a touch colder overall. SOI May June July Distance (less = closer to 2021) 1979 +4.0 +4.5 +13.6 (warm-Neutral winter) (7.2 off) 2011 +2.1 +0.9 +9.1 (La Nina winter) (9.8 off) 2010 +10.5 +1.3 +18.7 (La Nina winter) (10.3 off) 1974 +10.6 +1.7 +11.1 (La Nina winter) (13.5 off) 1939 -0.4 -1.8 +7.5 (El Nino winter) (14.9 off) 1936 +4.7 -1.8 +3.9 (warm-Neutral winter) (15.0 off) Blend: +5.3 +0.8 +10.7 2021 +3.9 +0.0 +16.3

Subsurface heat content for May-July 0-300m below 100-180W in the Tropical Pacific best matches 1981, 1987, 2014, 2014 as a blend. The blend of those four features big drops in the subsurface at similar magnitude to 2021 in May-->June, June-->July. Historically, for 1979-2020, most years around -0.3 to -0.4 in July subsurface readings do end up as La Ninas, but quite a few in that area are Neutrals. A lot of years are borderline too, with only 1-3 months 0.5C or colder than the 1951-2010 average in Nino 3.4. When I talk about a Neutral that's what I mean. I'm pretty sure the surface will drop to La Nina cold levels - just not convinced it will be long enough for an official event yet. Last year, you already had the subsurface very cold in Apr-Aug prior to the surface getting cold enough to be in a La Nina in late August. Nino1+2 Nino3 Nino34 Nino4 Week SST SSTA SST SSTA SST SSTA SST SSTA 07JUL2021 22.2 0.1 26.0-0.0 27.4-0.0 28.9-0.1 14JUL2021 22.2 0.4 25.9 0.1 27.2-0.1 28.6-0.4 21JUL2021 22.1 0.6 25.2-0.4 26.8-0.5 28.5-0.4 28JUL2021 22.1 0.8 25.1-0.4 26.7-0.4 28.7-0.1 We're seeing a very nice Summer locally, with cool to near average highs in May, June, and July, wet conditions and somewhat warm lows. Total year to date 90 degree highs here are running below the 100 year median and average. Part of that to me is that La Nina doesn't seem to switch on until ~26C in winter, and even in August it's ~26.15C (26.65C is average for 1951-2020). So we're just not that close yet at the surface, with 27.0-27.1C likely to be the monthly number for July.

One thing 2013 has going for it outside the tropics is the y/y transition in the Arctic. The 2012 and 2020 sea ice extent were both very low with dis-favorable patterns up there in Summer. We seem to be in a better pattern for the ice retention in Summer 2021 - so far anyway - similar to how Summer 2013 was. I had used 2012-13 as a double weighted analog last year ahead of winter, mostly for precipitation patterns nationally rather than temperature patterns, so it's been interesting to see 2013 is still a decent match. July 2013 is pretty similar to the temperature spatial pattern of July 2021 nationally, and drought nuking rain of July 2013 has been present in the Southwest. I kind of "knew" this was coming, which is why I was so annoyed in that "look it's hot SW" climate thread from June when it was obvious we are entering our wettest Summer pattern in literally years. Here is Albuquerque for July 2021 v. July 2020:

Right now, 2013 is a decent match for SSTs globally. It's amazing how different the weeklies, tropical tidbits, and these maps all look frankly. I would say we're behind 2013-14 still at the surface, and I don't consider that year a La Nina for winter. That cold water off western North America is the main issue with 2013.

My interest in looking at the NAO is mostly from a national temperature perspective rather than identifying periods for Nor'easters or something. In that sense, it is most useful from about mid-January to mid-April. For Nov-Jan, the AO is more important in terms of long-term correlations. For precipitation, the NAO actually has some pretty strong correlations in various places too. The WPO is actually pretty useful in Sept-Nov, and then in Feb-Apr for temperature patterns too, in a correlation sense. Locally, the +NAO/AO is actually a good signal later on if it happens in November from what I can see. The seven winters with 'rising' but still low solar are all -AO winters too, if that's your preferred measure. November NAO values are a good long-term indicator here for how wet late Spring will be, but otherwise it's not that useful until later in winter for national patterns for temps/precip. The long-term WPO correlations on the other hand correspond almost perfectly to some of the wild temperature pattern swings in the Fall and Spring in the past year. My point with the NAO/AO stuff isn't that I can see with 100% certainty that they'll be negative this winter. It's just we're nearing the end of the historical 'solar' time frame when odds of a -NAO winter are likely. To me it's worth figuring out whether we'll have one more winter with an extended -NAO/AO before we predominantly switch back to mostly positive NAO/AO winters. New Mexico is uniquely weird in that we get our top snow years in both the highest NAO/AO and lowest NAO/AO winters. So I'm actually pretty indifferent about what ends up happening typically, since most years are relatively neutral. You could see solar activity rise from 16 sunspots/month to 30, about the level of gain from 2019-20 to 2020-21. Or you could see it rise to 68 or something if we're going into a bigger cycle (more active sun) than last time. In the latter case, there is no particular signal in the solar data. April 2021 NAO was also negative like April 2020, which hadn't happened in forever prior to 2020, so that was my first tip off to pay attention to it again this year. Most people seem to use Nino 3.4 as an estimate for what the PNA is going to do, even though the correlation is ~0.3 r-squared for DJF, and much weaker in December. The Nino 3.4 to PNA correlation in August is literally stronger than December to Nino 3.4 for the PNA, so I always find the way it is used to be a bit strange. The image below actually supports a La Nina though, as the PNA correlation to Nino 3.4 is negative (+PNA = -ENSO) long term. https://psl.noaa.gov/data/correlation/table/corr.table_dec.txt https://psl.noaa.gov/data/correlation/table/corr.table_jan.txt https://psl.noaa.gov/data/correlation/table/corr.table_feb.txt

-NAO (DJF) winters (in aggregate): 1954, 1955, 1957, 1965, 1967, 1968, 1977, 1978, 1986, 1997, 2009, 2010, 2020. I would argue a year like 2000-01 or 1967-68 is more "-nao" in reality than a "fake" -nao year like 1997-98, but the point is, roughly half of years with rising solar since 1950 see a net -NAO in winter. The "rising" but still low (<55 sunspots per month July-Jun) years are: 1954, 1965, 1986, 1997, 2009, 2010, 2020. That's seven for seven for a net -NAO, whereas the "rising" but high years are six for twenty. The odds are not bad overall even if you want to regress the 7 for 7 with the x+1/y+2 rule with some Bayesian assumptions, i.e. assume the trend would go from 7/7 to 8/9. The 2021-22 July-June solar year is starting off with around 40 sunspots this month - going to be real interesting to see how rapidly solar rises this year. It's harder to predict than the decays from the peak solar I find.