raindancewx

Still there. Time-frame on this is 6 pm Saturday. So 10/2 evening +17-21 days. Roughly 10/19-10/23 or 10/20-10/24 for a storm for NM & CO. Roughly speaking, Kamchatka functions as the Rockies in the Bering Sea Rule. So the storms south/east of Kamchatka can pull in Gulf of California and/or Gulf of Mexico moisture depending on the path they take.

PDO was positive in both 1995-96 for Nov-Apr and 2009-10. It was +0.59 in January 1996 with the blizzard. https://oceanview.pfeg.noaa.gov/erddap/tabledap/cciea_OC_PDO.htmlTable?time,PDO I do think it will be more negative. But more "weakly positive" than truly negative. October in my analogs was a split decision: 2017 and 1961 had +WPO, 2001 and 2020 had -WPO conditions, with 1974 sort of near normal. If you blend 1961, 2017 (x2 each) with 1974 and then add a degree for how old those years are, it does look pretty similar to October on the CFS. Do want to see the Canadian, it's usually better. For what it's worth, all five of the years I like have the WPO positive in November. We'll see how that goes. It's probably not going to be as active in the late hurricane season without cold dumping into the Plains / Gulf like if the WPO is negative in October. The ACE was pretty high in 2001 & 2020 late - 46 & 75 respectively (22 is average), while the others are lower - 1961 (51.5), 1974 (2.6), 2017 (20.5). I'm expecting 30-40 from 10/1-end of season. You guys want 35+ ACE after 10/1 to top 150 for the season. That's a good snow signal for the NE in La Ninas.

Final SOI matches to July-September are in (broad) agreement with the CFS for October. SOI July Aug Sept 1938 17.2 12.4 7.6 1974 11.1 6.3 12.2 1981 8.1 5.1 6.4 1996 6.6 4.8 6.9 1998 12.8 9.7 12.1 2011 9.1 2.6 11.1 Blend 10.8 6.8 9.4 2021 16.3 4.4 9.2 A blend of 1938, 1974, 2011 is basically an ideal/canonical +WPO look for October. Matches fairly well to September for an analog based on three numbers and nothing else.

There are some mixed signals going into this winter. More than last year in some ways. The lowest Arctic sea ice extents in September correlate pretty well since the data starts to the biggest +WPO winters. That's a horrible/strong warm signal for most of Canada and the eastern US. But with more sea-ice extent this September v. 2020 you can already see the WPO behaving differently. It was very negative last October (a very cold signal for the Plains), while the temperature profile on the CFS is almost ideal positive (of course the CFS is terrible at predicting October I find). Generally you want upper level low pressure (blue) by 180W, 60N to be north of upper level high pressure (red) by 180W, 30N to get the +WPO sign from eyeballing it. WPO tends to not flip super often. So +WPO October is usually a cold signal for the West in Nov-Dec, a wet signal in the Plains in November, and a warm signal for the east through at least December. You can see what last year looked like when it was negative. For what it's worth, you're not really "supposed" to get huge +WPO readings with La Nina - last year was the first to have it so consistently positive since 1950 (only La Nina over +1.00 in Dec-Feb). Part of why I liked 2007-08 is it was the strongest recent La Nina with pretty consistently positive WPO readings. I think Joe Bastardi and his crew expect the WPO to go negative in December and that's part of why they expect a colder outcome than I do for the East. Some of the analogs I have do show that, but not the recent years.

The PMM warming this month makes me think the cold will have much more trouble getting into Nino 4 than last year. A lot of the cold-ENSO winters in the 1959-1967 period are pretty cold around 120W at the equator and then very warm in the North Pacific. Those years are pretty interesting winters, also a lot of nasty Gulf Hurricanes and similar features. The hurricane season put up 75 ACE 10/1-11/30 last year (3rd highest since 1931). Don't really see us getting particularly close to that in 2021. After Sam finally dies, could be quiet for a bit. Very unlikely to get a four/five combo in November two years in a row. For reference, these are the biggest October & Novembers for ACE: 87.8 (October 1878) 71.0 (November 1932) Highest Oct-Nov ACE since 1850 is 95 in 1878. The last ten La Ninas (2020, 2017, 2016, 2011, 2010, 2008, 2007, 2005, 2000, 1999) average only 35 ACE (+a bit more December). The 30-year average is 22 for 1981-2010. I kind of look at 145 as a good guess for now. 110 through 9/30 and then 35 more? The pattern should shift to somewhat west of what it was last year for cold dumps if the ACE drops from ~180 to ~145 going into winter. https://psl.noaa.gov/gcos_wgsp/Timeseries/Hurricane/hurr.atl.ace.data

https://www.accuweather.com/en/winter-weather/accuweathers-2021-2022-us-winter-forecast/1022887 As Paul Pastelok said in the earlier link, the Accuweather winter outlook is out today (9/29). Some things are similar to what I have. Some stuff I look at locally makes me a little skeptical of a dry winter out here despite La Nina. Was amused to see they had a wet winter here. I don't buy ~130% of normal here like he has though.

The higher the ACE the more likely it is the entire West has heat and dryness in a La Nina. Not just part of the west seeing one or the other, or just the Southwest - the whole West. It becomes close to a guarantee somewhere between 160-200 ACE I find. Last year was 180, and most of the west (TX, NM, CO, WY excluded where it was fairly average to cold for temps, and varied for precip) was pretty warm and dry. There is ACE inflation though via measuring techniques, so I treat last year like a 165-170 year. The highest ACE La Ninas all feature 3-6 month periods with literally no rain or snow in places in the Southwest - extraordinary dryness even here. I've also mentioned that in NYC/Philly/Baltimore the La Ninas with over 150 ACE are close to a coin flip for a good snow year (+20% or more) while all other La Ninas have close to 0% odds of a snowy winter. Look at your high ACE years for snow in the NE with La Ninas: 1933-34, 1995-96, 2010-11, 2017-18, 2020-21 are all in there as examples.

Ruh Oh. The GFS still has a very powerful low in same spot as 2/24/2021. That spot, +17-21 days is trouble for New Mexico and Colorado storms. Should be something pretty big coming through in that period. 3/13 was the Rockies blizzard right? You can see the +17 to +21 days works pretty well from that spot.

I don't know if it is of any use, but for my outlook this year I am including a tele-connection table to lay out the patterns. I partly did it because I think there are two pretty violent pattern shifts in mid-November and mid-February in the upcoming cold season. Been struggling with the weighting for November in particular. But basically it will lay out my expectation for WPO, PNA, AO, NAO for Nov-Mar. October is going to be a little bit nutty with the re-curving typhoon, rapid recent SOI changes, and then the GFS showing a very deep low off the southern Kamchatka tip. It's the same spot as the 970 mb low 2/21/2021 that preceded the March blizzard out here. https://en.wikipedia.org/wiki/March_2021_North_American_blizzard The Bering Sea Rule tends to translate those lows in those spots to "Albuquerque Lows" in 17-21 days. Locally, most of the tele-connections are broadly useless as correlations to long-term temperatures in most months under 0.04 (r-squared). But the January-April period does respond more to the NAO and WPO locally, so I do look at them for that period here. The big late September NAO crash is interesting - you might want to look at the periods that follow given your region. I'm fairly confident that the "look" that often precedes severe cold in the Plains in February is going to show up again in November. Since at least the late 1970s, that seems to be a +WPO, +AO, +NAO look in November. That's there in 1978, 1992, 1993, and 2020, with several other less severe outbreaks are similar.

The CFS may be off it's rocker, but it has a very wet October in the Southwest. We're heading into a cold-ENSO winter. But we also should have a pretty high ACE value. It's not real common for a wet October + big ACE combo to happen down here, especially with a cold ENSO look. I'm real curious to see what happens in the Atlantic - 2020 October-November ACE was third highest since 1931, behind 1932 and 2016. Unheard of to be as active as 2020 was late two years in a row. So ACE might only hit 155 or 160 or something. My weighted analogs of 1974 and 2017 as a blend might actually be about right - 150 ish ACE, with 1.00 inch of rain in October. Some of the moisture shown is the CFS not understanding a lot of the incoming rain will fall 9/30 in actual American time and not UTC. But it probably does see an actual wet month even ignoring that issue.

This is my take on why this event won't catch up or surpass last year at the surface. In 2017, Nino 3.4 was well behind the prior year cold. Unlike this year though, it was much colder y/y in Nino 1.2 and 3 by this time. We're running warmer each week in all zones for at least two months now.

Pending the exact timing of the heavy rain for New Mexico & Texas on 9/30-10/1, the strong Jun-Aug 1974 match to June-Aug 2021 for precipitation looks like it will continue into September 2021. Will be curious to see how that goes going forward, since the Summer was also similar to Jun-Aug 2017, which was very dry here in winter, while 1974 was very wet. I'm including 1961 as an analog as it is almost like a split the difference between 1974 and 2017 for winter. Both 1974 and 2017 are weak La Ninas following La Ninas with the QBO in the right phase. I find low solar activity yields different precipitation patterns in Spring/Fall then high solar activity nationally. We're getting real close to my threshold for "high solar" in individual months now, with ~50 sunspots looking likely for September, most since 2016 for a single month. For the Southwest, high solar is a pretty strong wet signal in March. But you need around 55 sunspots annualized for the "high" solar activity to show I find. Still only at about 25 annualized if September does finish around 50. Odds of getting above the high solar threshold are increasing though.

The GFS has a 980 mb low south of the tip of Kamchatka on 10/3. It's been there, or similar for a few runs now. That's probably a good snow event for at least the higher terrain of NM and CO if that verifies in a few days. Would be 10/20-10/24 if the GFS timing is right.

With the SOI back down to +8 for the past 30-days, the closest (tentative) matches for July-Sept have changed a fair bit from what I expected earlier (+10 Sept). 1931, 1938, 1950, 1974, 1981, 1996. Not super different looking from what the CFS shows for the month. I suspect the CFS will change it's tune quite a bit for October by 9/30 though. I'm pretty skeptical of the CFS look though. SOI could also still change quite a bit by the end of the month.

This is the good PDO date - the JISAO/Nate Mantua method. https://oceanview.pfeg.noaa.gov/erddap/tabledap/cciea_OC_PDO.htmlTable?time,PDO

May not be apparent for a couple days or a few weeks even, but part of me thinks the subsurface must be warming up given the massive flip in the SOI in recent days. The subsurface trend flipping from cooling to warming also accompanied big SOI crashes last year, both times, in late Oct and mid-Feb. The rain timing out here is right on time with my analogs though - 2017 had big rains late September in the SW. (9/27, 9/28, 9/30) 1974 had a big rain event around 10/6. Will probably be rainiest here 9/30-10/1. So it's in between the main analogs. The 10/1 timing also matches the recurring 42-48 day precipitation cycle that tends to show up here in the Summer, and then repeats through Spring. Date Tahiti (hPa) Darwin (hPa) Daily Contribution 30 day Av. SOI 90 day Av. SOI 25 Sep 2021 1011.09 1011.45 -15.92 8.12 9.03 24 Sep 2021 1011.46 1012.20 -18.18 8.84 9.14 23 Sep 2021 1013.45 1012.85 -10.22 9.79 9.41 22 Sep 2021 1014.79 1012.45 0.12 10.53 9.64 21 Sep 2021 1015.81 1011.30 13.02 10.70 9.80 20 Sep 2021 1016.79 1013.10 8.14 10.38 9.73 19 Sep 2021 1017.61 1013.20 12.42 10.48 9.66 18 Sep 2021 1017.26 1012.55 14.20 10.52 9.37 17 Sep 2021 1016.61 1011.85 14.50 10.34 9.03 16 Sep 2021 1016.31 1012.35 9.75 9.74 8.78 15 Sep 2021 1016.76 1012.70 10.34 9.38 8.69 14 Sep 2021 1017.57 1012.75 14.86 9.17 8.68 13 Sep 2021 1018.34 1011.35 27.75 8.69 8.70 12 Sep 2021 1017.66 1011.15 24.90 7.61 8.67 11 Sep 2021 1015.55 1012.55 4.04 6.89 8.57 10 Sep 2021 1015.61 1013.50 -1.25 6.96 8.57 9 Sep 2021 1017.31 1013.70 7.67 7.35 8.60 8 Sep 2021 1016.73 1013.80 3.63 7.68 8.48 7 Sep 2021 1015.24 1014.00 -6.42 7.70 8.45 6 Sep 2021 1015.25 1013.20 -1.60 7.84 8.52 5 Sep 2021 1015.88 1011.10 14.62 7.60 8.42 4 Sep 2021 1015.82 1010.50 17.83 6.51 8.05 3 Sep 2021 1014.85 1010.85 9.99 5.31 7.61 2 Sep 2021 1014.58 1011.80 2.74 4.63 7.27 1 Sep 2021 1015.79 1011.95 9.03 4.76 7.09

Once the NAO value overall value is in for September I can see which years match best. I'm also curious to see if the huge dip actually verifies as severe as depicted. It's a little hard to eyeball a monthly value without knowing the severity of that dip.

I'm rooting against the September rain being heavy in Albuquerque. It's a terrible dry signal for the SW in La Ninas (heavy rain in Sept), even relative to the normal La Nina dry signal. But it's also hard to find a bad winter after a wet October here, regardless of ENSO state. Years like 1960, 1984, 2000, 2011 all had wet Octobers. The Euro had 1.4 inches of rain in Albuquerque on 10/1 on a recent run.

I don't what he means by harmonics. I use timing variables though for local stuff since we get so little rain. We had cool periods 6/29, 8/15 locally and now have another coming 10/1 with the rains (and high mountain snow) this week. So you have a 42-48 day recurring cold signal in the Southwest that's repeated multiple times already. Last year it was more like 6/10, 7/25, 9/9, 10/25 or something.

Hurricane Sam looks like it could push the number of major hurricane days above the 8.75 recorded last year. For the period with the ACE index, 1850-now, the 8.75 major hurricane day figure correlates to about 140 ACE long-term, despite the 180 that actually occurred from the parade of weak tropical storms. If Sam really last Sat-Thu (six days) as a major hurricane, we're already at 12 (!) days with a major hurricane this year. Then we could easily see a few more in October-November. Very simple way to estimate La Nina winter precipitation in Albuquerque is just 1.46" - (0.0522")*(major hurricane days). Probably not two cat five (ish) Nov storms though. Last year that formula implied 1.00 inches and we had 0.82 inches. Average error for the past 29 La Ninas is 0.40" from that formula. 90% chance of being within 0.70" historically. The trade off locally is with the rain and clouds late September, the month won't finish quite as hot as I expected. So that actually favors a bit more precipitation in January and February locally. The 10-wettest Septembers in Albuquerque heading into cold ENSO winters are a strong dry signal here for winter. They have average 0.88 inches of winter precipitation - about 33% below average, around double the normal dry signal for La Nina. The wettest Septembers average 2.39 inches of rain though. We're at 0.51". I don't think the rain arrives quite in time to make it by the end of September. A wet October would be welcome though - those are a strong signal for March fun.

I was looking more at the super-cold Plains centered Februaries today. You seem to have common look in November before those Februaries. Haven't really looked at the pre 1978 outbreaks. +WPO November +AO November +NAO November Didn't really see indication for commonality in other time frames. 1978, 1992, 1993, 2020 all had that signal though ahead of the nuclear cold dumps in February 1979, 1993, 1994, 2021. Nov WPO AO NAO 1978 +0.10 +2.47 +3.06 1992 +1.26 +0.72 +1.12 1993 +0.83 +1.00 +2.56 2020 +0.72 +2.09 +2.54

The current look in the Pacific is something like a basin wide very weak La Nina or cold Neutral. The subsurface readings may already be near their coldest values. A lot of the La Ninas with earlier peaks spend late Fall and Winter with retreating subsurface cold, it becomes more similar, by trend to a strengthening (warming) El Nino subsurface. That's why you can get a relatively cold signal in the Southwest in some of the more special cases. There is less cold at the surface by Peru this year but more is coming, and similar readings to last year in Nino 4 currently. I could see this event being a pretty normal basin-wide look rather than an east-based or Modoki look for winter. I'm not super hung up on it though because I think we'll be returning to Neutral conditions much faster than last year. Here is what happened after the subsurface trend reversals last year (flip from cold trend to warm trend). Scale on all of these is -7 or colder for deep purple, +7 or colder for deep red, in units of 2 degrees Fahrenheit. Opposite flip, from trending warmer from a cold start, to colder from a warm start:

I have a similar look to what you have for December. For November I expect that look to be transitional and not the entire month. I'm pretty different for January, and then I think February is pretty cold again. Not sure why you consider 2007 Neutral though? You'd get a different composite without it. The models have been trending toward a widespread rain event in the Southwest for late September / early October. That's something that happened in 2017 before the 96-day precip free period started in October - terrifies me a little bit actually. The big SOI crash recently is consistent with early October storminess out here, which also showed up in my analog blend when I attempted to time out the patterns. There are two "strands" where it has rained every 46-47 days here since June, and the next instance of the wet signal is about 10/1. For now, if the SOI finishes September around +10, the closest July-Sept match in the past 90 years is 1974, which is one of the analogs I like. I doubt the CFS has the right temperature for October at the moment, but it does look like October 1974 to some extent.

From the draft of my winter outlook.

This is from my winter outlook draft. Key word is draft. This is not necessarily representative of what I'll settle on by 10/10 when I finish up. If you're wondering if it is possible to get two severely cold, Plains centered, February cold outbreaks in a row, the answer is yes, and it's not actually that uncommon. The second year outbreak tends to be weaker though. I used 1936, 1978, 1994, 2018 as the match for last's year outbreak. Tentative idea is a more extreme version of Feb 1937, 1979, 1995, 2019 - i.e. similar spatial placement but much warmer in the south since you have strong long-term correlations to SE warmth in winter after -PDO Summers.