raindancewx

The cold-ENSO winters following two El Ninos usually do not lead to multiple-La Ninas in a row. I lean toward a Neutral that flashes cold and warm at times for next year but it is pretty early for that. My guess is Nino 4 will be pretty cold early next year, but warming steadily. The other zones will probably warm initially, but might reverse cooler later on. The models have Neutral to near-El Nino conditions by late next year. The transitions after two El Ninos that went to a La Nina/near La Nina are these: 1929-30, 1930-31 --> two cold Neutrals for 1931-32, 1932-33 1957-58, 1958-59 --> two neutrals in 1959-60, 1960-61 1968-69, 1969-70 --> two La Ninas in 1970-71 and 1971-72 1976-77, 1977-78 --> two neutrals in 1978-79 and 1979-80 1986-87, 1987-88 --> a big La Nina in 1988-89, and then a neutral in 1989-90 2014-15, 2015-16 --> I consider 2016-17 a La Nina, but it is really a Neutral strictly speaking, and 2017-18 was a La Nina

The CFS look for next Summer is actually very similar in the Pacific to what the "cold ENSO" after two El Nino look is historically. Those are 1929-30, 1930-31 El Ninos followed by 1931-32 (Summer 1932), and then more recently Summer 1960, 1971, 1979, 1989, 2017. Idea is Nino 4 is still pretty cool, and it occasionally bleeds into Nino 3.4

Weatherbell stuck to their warmed up blend of 1973-74, 1988-89, 2010-10 (x2). The Joe D'aleo Pioneer Model has a lot of years I like as analogs: https://www.weatherbell.com/winter-2020-update I've seen a lot of people saying you can't really get a strong or moderate La Nina look with the North Pacific so warm. You really can if you blend in a year with a very strong La Nina with a year that had a very warm North Pacific. I think blending in 2019-20, 2012-13 in at a low weight with a strong La Nina does a pretty good job warming up the north Pacific a lot without diminishing La Nina strength too much. That's what I did in my outlook anyway. I didn't use all of these years in my outlook, but at this point I think the 10 best analogs for winter are probably 1954, 1961, 1973, 1988, 1995, 2003, 2007, 2012, 2016, 2019. This will be a good year for seeing which factors matters most. My general sense is the warm North Pacific, with low sea ice and a pretty strong La Nina means you'll get (sporadic) very strong troughs/cold dumps into the West, but because a lot of extra moisture sources exist (Arctic Ocean, East Pacific at mid-latitudes), you'll see more snow and wetter Western storms. In the Southwest, we're usually dry in La Ninas because storms run out of moisture by the time they get here if the storm track has to crash over mountains, through cold water and a ridge. I don't think that's the case this year. You had an almost identical SOI crash that preceded the late October snow storm here to 2007, but the system this year had far more moisture. To me, that's likely to keep happening with a warmer North Pacific / Arctic.

The years in the CPC 6-10 and 8-14 day analogs have been changing a lot lately. We're close enough to Dec-Feb now that I think you have to expect some of the commonly repeating years to show up in winter too. For the 6-10: 1958, 1961, 1968, 1973, 1975, 1988, 1989, 1990, 1994. I like 1988 and 1961 as second tier analogs - 1961/2010/2016 is a good QBO/ENSO blend. For the 8-14: 1958, 1973, 1988, 1994, 1996, 1998, 2001 2005 As much as I like 2007-08, it's a double La Nina. The models don't have that. So once we get past Dec/Jan, 1988-89 may actually be the right idea, if we're going to see a lot of weakening out a La Nina, assuming the cold does dump West like I think it will. 1973 is kind of interesting too, since 1974 was a La Nina but far weaker than 1973. I prefer 1988 over 1973 though, since it followed two El Ninos which is quite rare.

This is pretty close to my thought process, although I dislike half of his years. https://www.youtube.com/watch?v=MXg0srzblgk&ab_channel=EricWilhelm He likes 2016-17 as the main analog. I think a de-amplified 2016-17 is sort of the right idea (less extreme warmth East, less extreme wetness West). He mentions a pretty high likelihood of a top five warmest ever winter for his area (Youngstown OH). His blend is: 2016-17 (x4) 2010-11 (x2) 2007-08 1998-99 (x2) 1995-96 1988-89 1983-84 If you were to just do a QBO/ENSO blend, at this point I'd say a 2010/2016 blend is appropriate, and maybe you throw in 1961 to get some more Gulf Hurricane activity in there like this year. That's a 2C SST drop off in Nino 3.4 overall as a blend, which is at least ballpark for this year (I still think its 27.1C to 25.5C for the actual winter, but could be colder). The blend below looks like what I have, but colder in the East/Southeast. Year Jan Feb Mar Apr May Jun Jul Aug Sep Oct 1961 -5.47 -0.62 0.54 2.82 6.88 7.96 6.69 5.60 6.02 7.59 2010 -16.02 -16.98 -19.68 -23.57 -26.28 -25.05 -9.84 1.45 6.58 10.83 2016 9.34 6.77 3.16 0.64 2.37 3.86 6.25 10.07 10.48 12.83 Blend -4.05 -3.61 -5.33 -6.70 -5.68 -4.41 1.03 5.71 7.69 10.42 2020 -2.51 -3.20 -4.36 -5.03 -4.86 -2.78 0.34 4.78 7.95 10.80

The latest plume from the European still has rapid weakening of the La Nina late winter. The cool down is supposed to continue longest in Nino 4 with the reversal (warming) starting slowest there. I think there is a pretty cold period in Nov-Dec for Nino 3.4 (now) before rapid weakening sets in, so i drew in my expectations - aggressive cooling early, aggressive warming late for Nino 3.4. In Nino 4, I like slower cooling initially, and then slower warming late. Those seem to be the biases for the plume in La Nina years from my recollection. The biggest ever (1950-2019) drop off from September to DJF in Nino 3.4 is about 1.0C. So anything below 24.9C (-1.6C) for DJF just strikes me as very unlikely (<2% odds?).

Year DJF JFM FMA MAM AMJ MJJ JJA JAS ASO SON OND NDJ 2020 0.5 0.6 0.5 0.3 0.0 -0.2 -0.4 -0.6 -0.9 CPC lowered the September Nino 3.4 reading to 25.88C fro 25.96C initially. October was 25.48C - warmer than 2007 (25.32C) but still tracking close to it. Nino 4 came in at 27.97C in October - still pretty near the 1951-2010 average. But similar to October 2007 (27.91C). The reading in Nino 4 in October 2010 was already down to 27.15C. So the pattern is a lot closer in Nino 4 to October 2016 (28.32C) still than 2010 despite how cold Nino 3.4 is.

The only November category five in the Atlantic according to Phil Klotzbach (CSU) is the Cuba hurricane of 1932. So I think you'll see some hesitation by NHC to call Eta a five before landfall unless they are absolutely certain. It is interesting looking back, a lot of the strongest La Ninas have had fives - 2007 (Dean), 1988 (Gilbert), 1998 (Mitch), and so on. It did snow in October 1932 in the valleys of the SW, so the pattern that year is likely similar in an MJO sense. The record cold at times this Fall in the West is also similar to some of the great cold waves of the 1932-33 winter (-66 in Wyoming...seems unreal, but Montana set a record low for the continental US in October).

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.

Mitch killed 10,000 people in 1998 with 180 mph sustained winds at its peak if my memory is right. Similar path and time of year to this. Pretty sure the models didn't get the path right on Mitch - have to watch for that in the coming days.

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.

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.

Nino1+2 Nino3 Nino34 Nino4 Week SST SSTA SST SSTA SST SSTA SST SSTA 30SEP2020 20.1-0.5 23.8-1.1 25.6-1.1 27.9-0.7 07OCT2020 19.5-1.2 23.4-1.5 25.5-1.2 27.8-0.8 14OCT2020 19.6-1.2 23.6-1.3 25.3-1.4 27.8-0.9 21OCT2020 19.9-1.1 23.8-1.2 25.3-1.4 27.8-0.8 28OCT2020 20.0-1.1 23.5-1.5 24.9-1.7 27.9-0.7 Nino 4 and 1.2 warmed a bit this week. Big drop in Nino 3 and 3.4 though. The current La Nina is now stronger/colder in Nino 3.4 than 2007 at this point, but warmer elsewhere. I was expecting Nino 3.4 to come in around 25.55C since the monthly data usually runs above the weekly data, but it looks lower than that now. This week the data is about half way between 1995 and 2007 in the eastern zones, but now much stronger/colder than 1995 in Nino 3.4. The slow cooling in Nino 4 is a feature of both years. Nino 4 was already dropping below 27.0C in 2010 and a lot of older La Ninas by this point (November). 31OCT2007 19.7-1.5 23.2-1.8 25.2-1.5 27.8-0.8 25OCT1995 20.3-0.7 24.0-0.9 25.6-1.1 27.9-0.8 Pretty healthy La Nina in Nino 3.4 though - will likely be in "strong" territory in Nov-Dec before weakening.

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:

His analogs do have some interesting features though. He's got warm periods in there for almost everyone - often near record warmth. Think December 2014 for the NE, or 1995/2005 for the SW. But then he has incredible cold periods like Jan-Mar 2014 and 2015 for the Midwest and NE. The mid-Dec to mid-Jan period in 1978-79 is super cold nationally, and as I've showed previously in here, Jan 3-9 1971 is insanely cold in the West, 20 below average in New Mexico and Colorado, for a week, despite that month warming up a lot later. The 1995 and 2009 winters are also pretty cold/snowy for a lot of the US. But 2005 is not. So I think he is trying to get at how volatile the winter could be. Where I am he has December 1959 with two feet of snow in the valleys, and February 2015 with 10 inches of snow in the valleys. The mountains saw 50-100 inches in those periods in a week or so, and that does tend to happen every five years or so. The 1959-60 period also has a major thaw in the middle of the country in the mid-Dec to mid-Jan period.

Next storm looks to be around 11/7 or 11/8 for the SW. Good to see. Legitimately no rain or snow from 10/5 to mid-January in 2017-18 here.

Larry Cosgrove has a blend of 1952, 1959, 1970 (x2), 1978 (x2), 1995, 2005, 2007, 2009, 2013, 2014 as his winter analogs. One hell of a blend. You have years with record cold / record warmth nearly everywhere, including some big snow months nearly everywhere. I actually don't think it's a terrible blend. Locally: December 1959 is incredible. So is February 2015 for snow. January 3-January 9 1971 featured multiple lows below -10 in Albuquerque, but the month still finished warm. But the week was 20 degrees below average for a huge portion of NM and CO. I think mid-Dec to mid-Jan 1979 is probably one of the coldest periods on record for the West. 1995/2005 both featured very warm/dry conditions in the West. 2009, 2013, 2014 also all featured at least brief cold in the SW at times, same for the NE actually. US wide: May 2015 is one of the wettest months on record, top five I believe. The 1970, 1978, 1995, 2007, 2009, 2013, 2014 are all know for their severity in one way or another - cold or snow - in huge portions of the East and Midwest.

Solar activity rose to 14.4 sunspots for October 2020 - highest since June 2018. The 12-month average is now above the year ending October 2009, at 4.2 sunspots/month v. 3.5 sunspots / month. That isn't a big difference, but it is 20% higher. The six months ending October 2020 were 5.8 sunspots/month, compared to 4.9 for the six months ending October 2009 too. Oceanic-heat content for October for 100-180W fell to -1.08. That's somewhat colder than October 2016 and October 2017, but not much. https://www.cpc.ncep.noaa.gov/products/analysis_monitoring/ocean/index/heat_content_index.txt Closest simple blend to match 2020 trends for Aug, Sept, Oct is 1999, 2007, 2016, 2016. Not the type of winter most of you will want if that blend holds. Would be a very warm November too. Year Aug Sept Oct 2016 -0.71 -0.71 -0.92 2016 -0.71 -0.71 -0.92 1999 -1.21 -1.27 -1.07 2007 -0.68 -1.03 -1.19 Blend -0.83 -0.93 -1.03 2020E -0.81 -0.87 -1.08

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.

The new Canadian trended the north Pacific warmer and the La Nina somewhat warmer too. Atlantic actually trended a lot colder too, with the Indian Ocean warmer.

The Euro seems to be on board with another good storm in the time frame you'd expect with the big SOI crash. It's crazy to me that the SOI is going to finish October so far under +8 when the La Nina is so healthy looking on the SST maps.

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.

This snow event for Boston is a good sign for the snow part of my winter outlook - NYC got jack, Boston got snow. I think that's going to repeat quite often in the winter. We'll see I guess. Definitely wasn't expecting record October snow in Boston and Albuquerque, that's one hell of a storm to do that, even with the huge SOI drop from the La Nina base state earlier in the month.

The CFS has gone to my main winter analog for November: 2007. I would expect the CFS is too warm everywhere, but probably correct spatially for November 2020. Tropical Tidbits currently has these as the top SST matches globally: 1967, 1988, 2010, 2011, 2017. Weatherbell had a cold West look for November (1973, 1988, 2010, 2010) - will be curious to see if the warm West verifies of if they see something the models don't. I think we'll have some cold in the West in November but generally a warm month. The issue is 1973/1988/2010 are all way too cold/strong compared to where the current La Nina is, especially in Nino 4 where those years are near 27.0C, and this year is still near 28.0C.

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.