raindancewx

If you try to reconcile trends in the QBO and trends in the ENSO pattern year/year you get the same outlook as my actual (tentative) winter analogs more or less. Last year the y/y trend in Nino 3.4 and the QBO matched with 2004, and the winter was pretty similar, so I think I have the right idea for what to expect this year given that the QBO/ENSO trend blend matches my analogs. The up/down you see is whether the QBO trend was moving up or down from the prior month. You can see each trend was the same for Jan-Aug in 2004 and 2019, and the ENSO trend was El Nino after warm Nino 3.4. I think the ENSO trend matters a lot, because 2002 was actually a very close QBO/ENSO match, but it had a very different prior year ENSO. So 2019-20 looked nothing like 2002-03 nationally. These are from my winter outlook draft - Something like this for 2020-21. I'l refine it if September comes in very different on the monthly QBO index though. You can see, the trend in each month in the blended average matches 2020. So you'd expect the QBO to be around +6 for September, or at least higher than August. The year over year ENSO drop is probably more like 27.1 to 25.5 than 27.9 to 26.1, but it's close enough. I went with a blend because the La Nina / near La Nina +QBO winters following El Ninos are all pretty poor matches for either QBO intensity, or monthly trends. 1966 is way out for -QBO intensity early on. 1973 has the wrong month/month trend in most months. 1980 was trending up all year, so the timing is off. 1998 is off a lot by timing and magnitude. So the best matches are 2010 and 2016, but even they aren't that strong really. We're not going to see the 3C drop off in Nino 3.4 those two years had, and the 2020 QBO is way more positive each month compared to 2010.

This is a top five cold start to September here since 1931, through 9/13 (~43% of the month). Will be very curious to see if we can finish with a cold September. It does look warmer for the next week, but no 90s in the forecast. Any incoming smoke from the West Coast fires could also hold down temperatures during the day if it makes it out here in a thick plume.

The ACE index in the Atlantic is still remarkably low for how many storms we've had named this year. Even if the second half of the season was ~50% above average, we'd finish around 125 ACE. Pretty promising development for the West. After Tuesday/Wednesday, everything may either be dead or on land or extra tropical, except for Teddy. I don't really think Sally or Paulette are going to hurricanes real long either. The La Nina years with near average ACE in the Atlantic and low ACE in the East Pacific are pretty good winters for the West generally (1974, 1975, 1988, 2000, 2007)

My take is a weaker version of 2007 is the right idea. But I'm open to a stronger La Nina if the event can keep pace this month with 2007 in September. The bottom really fell out in 2007 in the Fall. That's the real test. The warm western Nino 4 subsurface was still there in 2007, even well into September - The models missed how strong it got: -1.5C ONI (forecasts were generally -0.5 to -1.0) I mentioned earlier, if you plot August subsurface against Dec-Feb SSTs in Nino 3.4, you get 25.75C as the estimated strength of the La Nina, give or take 0.75C, at about 80% certainty. So below 2007-08 (25.0C in winter in Nino 3.4) is pretty unlikely. But it wouldn't really be too shocking if it got close to that strong. My instinct is this event will be strongest in Fall, maybe as low as 25.5C or something in September, October or November given that week one of September was 25.8C on the weeklies. Nino 4 and the PDO are both the coldest they've been in a while, which is almost like an indication that the system is less resistant to a stronger La Nina than it has been in a while.

I had kind of forgotten about this since it didn't apply last year, but La Nina after El Nino is a pretty consistent and reliable cold signal for the interior Northwest. In the last 10 cases, an area by Idaho, Wyoming, and Oregon is always either cold, or near average, even in cases when the US is mostly torching. If you look at the areas north of the San Francisco to SD-NE line, that zone is never completely warm in any of the ten most recent La Nina after El Nino winters. There is also a pretty strong warm signal for Texas - 7/10 years feature +3 or hotter winters somewhere in Texas. Even Florida and the Southwest don't have that. The three composite looks that show up by month - cold interior West, cold Northwest, cold NW to Midwest - I actually have analogs that support all three to some extent. So it looks pretty reasonable to me. I think the January look is the closest to what will happen of the three months, but still want to see how the observations in September in the tropics pan out.

I'm with the one guy on the ice storms. I think there will be 2-3 big ice storms for the east this winter. I can't help but notice that despite very different setups since 2006, the lowest sea ice extent Septembers (<4.3m square km) tend to favor the West for cold in cold ENSO years (2007, 2012, 2016) while the higher years do not (2008, 2010, 2013, 2017), with 2011 (4.3m almost exactly) a NM/West TX special (it was very cold that December here). The east of course is warm in those lowest ice years, while the higher ice cold ENSO years are pretty cold for the east and very hot for the west. We don't really have sea-ice extent data for the 1950s/1960s, but the historical look backs I've seen imply that those years had sea ice around 5m in some Septembers when the Atlantic was very warm especially for that era. A lot of those cold ENSO years are also pretty cold in the West, 1954, 1955, 1961,1964 as examples. In other words, I don't think it's the volume of sea ice itself, I think its something being near the cyclical physical ice-extent minimum in a cold ENSO that may favor it (I say cyclical, because eventually we'll beat 2012, but not soon). The Northwest has a very strong tendency to be cold in La Ninas following El Ninos (1970, 1973, 1983, 1988, 1995, 1998, 2005, 2007, 2010, 2016 is the most recent composite for that). Every one of those years has at least someone cold (or relatively cold in national blow torches like 2005-06, and 1998-99) in the Northwest, between San Francisco and the Nebraska/South Dakota border. Not a single exception. I don't think it's the La Ninas themselves favoring the NW for cold those years, it's the big time cool down y/y that favors the cold for them up there, just like the big warm ups favor cold down here rather than the actual El Nino strength or type. If you look, there is no direct correlation between cold ENSO strength and temps in most NW cities, just like there isn't for warm ENSO and temps down here, it's the trend up or down that favors cold for the West. You can find other similar years, the current temperature anomalies for September in the US look like a lot like a blend of 2008/2011 for instance, but I'd still favor the La Nina following El Nino years overall, even though 2011 is the run away top match for similar temperatures here since June, and fairly close in Nino 3.4 for a few months now. The problem is a lot of the recent La Ninas are pretty similar to this May-August in Nino 3.4. The NAO and PDO both generally look wrong to me for Eastern cold, but I wouldn't expect the NAO to be completely positive for the winter. I do think there will be a moment with either rapid changes in MJO forcing or La Nina rapidly weakening, along with slowly rising solar activity that will allow the NAO to go negative. Winters like 2009-10, 2010-11, 2012-13 that had -NAO conditions at times and were generally trending more active in terms of solar conditions, even though 2009-10 and 2010-11 were both pretty low for solar activity. We hit a 12-month minimum of 1.7 sunspots/month (bottom 1% for all 12-month solar periods back to 1749) through February 2020 and have been (very) slowly trending up since. This is solar activity for 2009 and 2020 by month. 2009 1 2009.042 1.3 2009 2 2009.123 1.2 2009 3 2009.204 0.6 2009 4 2009.288 1.2 2009 5 2009.371 2.9 2009 6 2009.455 6.3 2009 7 2009.538 5.5 2009 8 2009.623 0 2009 9 2009.707 7.1 2009 10 2009.79 7.7 2009 11 2009.874 6.9 2009 12 2009.958 16.3 2020 1 2020.042 6.4 2020 2 2020.124 0.4 2020 3 2020.206 1.5 2020 4 2020.288 5.4 2020 5 2020.373 0.2 2020 6 2020.455 5.8 2020 7 2020.54 6.3 2020 8 2020.624 7.6

Using the 1951-2010 averages, Nino 4 has only been cooler than average for four months since mid-2012. It has been tracking similarly to 2007 this year. Will be interesting to see if that can continue for a while. A much cooler Nino 4 in September means the US should be a lot warmer than last October in most places too. 2007 28.97 28.66 28.66 28.73 28.84 28.90 28.68 28.48 28.11 27.91 27.61 27.42 2020 29.16 28.97 29.07 29.15 29.00 29.09 28.89 28.50 average 28.08 27.97 28.07 28.33 28.63 28.65 28.55 28.41 28.40 28.42 28.39 28.27 My sense is Nino 4 will have trouble falling below 28.0C this Fall-Spring. But we'll see. Still warm below Nino 4 after all.

High was only 55 today once again. So around ~30 degrees below average for the high. This is one of the coldest starts to September here on record, at least for highs. I think it's 6th place - 81.5F (down from 89.1F as recently as 9/8). Based on the historical 9/11-9/30 data for 1931-2019, roughly ~40% chance of a cold month (2F below average high or colder), less than 5% chance of a hot month (2F above average high or hotter), but still most favored for +/-2F compared to average. The high tomorrow is supposed to be in the 70s, so should drag the monthly average down some more.

Euro seems to like a -1.0C La Nina, with a somewhat warmer Nino 4. I'd probably go warmer on both. I use the colder 1951-2010 basis anyway, and on those figures these are more like -0.8C / -0.3C anomalies anyway. CPC is going to switch to a 26.65C or so base line for 1991-2020 Nino 3.4 average next year (+0.08C from the current baseline), so there is little doubt this event will be classified as a La Nina at any rate.

To me the experts have been half right on the season. A lot of the forecasts I saw had a lot of tropical storms - that's clearly right. One of the years I had for Summer though was 2007, which had only 74 ACE despite 15 tropical storms. We could easily end up with 20 or more in the Atlantic by the end of the season, which is pretty historic by any measure, even with the naming of the fish storms. It looks like about ~30% of seasons since 1851 have above average ACE, so I guess it's interesting to me, as someone who works with statistics a lot because you don't tend to see ~30% outcomes five times in a row in real world situations, even if they are independently determined. You can think of it like getting a La Nina or an El Nino five times in a row, which are also 30% outcomes realistically, since you have about 20 La Nina and 20 El Nino winters in the last 60 years. You get ~four same sign ENSO events once in a blue moon, but we've never seen five. If you think of 26.0C or colder in winter as La Nina in Nino 3.4, then 2010-11, 2011-12, 2012-13, 2013-14 is damn close to four La Ninas in a row as a recent example. But five would be pretty incredible wouldn't it? The Atlantic is different from ENSO, so it's not a guarantee that the season won't be active. But I do lean to below average.

The Atlantic ACE index averages 105 for 1981-2010. So there hasn't been a below average ACE year since 2015. I wouldn't be shocked if it happened this year. Statistically, Sept 13 is the half way point for total ACE in the 1981-2010 period. We're at 48 ACE as of the last update, and it goes up ~1.5 points per day in September on average, which is about a day of 1.5 tropical storms as a very rough estimate. I don't see any long-lived spurts of major hurricanes for at least 2-3 days, so it's not going to change dramatically though 9/13 when the season will be half over. For the data on the CSU site I can't find a five year period since 1851 where the ACE was above 105 each year for five years in a row in the Atlantic. Four years is fairly common: 1891-1894 1998-2001 2016-2019 I understand something unexpected could happen, we could have three category five hurricanes Sept 20-Oct 10, but I definitely lean toward an ACE index of 80-120 at this point rather than 120-160, or 160-200. Historically, September finishes around 80, and then you get 20-30 more points Oct-Nov. The fast start has really slowed since late August with the MJO help fading. It's pretty hard to get a complete crap winter in the West if the ACE is under 160 - so I'm definitely rooting for that but we'll see how it goes.

My issue with 2008 has always been it was warming up in the tropics that year instead of cooling. So the fact that we're cooler in Nino 3.4 already, after coming off an El Nino, makes me think the La Nina is stronger than 2008-09. The bold in Nino 3.4 in August. In a lot of ways, 2008-09 is a poor man's 1995-96, so it's not terrible by any means. I've flirted with using it at times. The 2008 hurricane season was a lot more active than this season seems like it will be in the Atlantic. The NE tends to have higher snow totals roughly from DC to Maine in the higher ACE La Nina years. I can also tell you that generally, 2008 was a cold/wet summer in the Southwest, and this summer is opposite. For whatever reason, the hottest Summers in the Southwest in La Nina tend to precede the colder winters. The NE getting hit by several tropical storms this summer is a good sign for Nor'easters I think (kind of like 2003/2012). I just think a lot of them will be snow to rain or rain to snow events without enough cold air for the entire event. 2008 24.86 25.08 26.07 26.83 27.09 27.04 26.99 26.72 26.47 26.37 26.25 25.74 2020 27.14 27.11 27.76 28.17 27.65 27.38 26.99 26.30 -99.99 -99.99 -99.99 -99.99

Today was the second coldest September high in Albuquerque back to 1892 - only 47F. Trailing only 9/27/1936. No high had ever been below 61 during Sept 1-15 in Albuquerque, from 1892-2019. The average high for the last 100 years on 9/9 is 85F - so legitimately close to 40 below average. Monthly high here dropped from 89.1 through 9/8 to 84.4 through today. It's actually almost unprecedented in the records to go from a hot August to a cold September - I'm a little skeptical yet that September will be able to remain below average.

My guess is the map is based on 1961-90 averages given the date listed at the bottom. But it is big and fairly detailed. I just assumed SE Canada would be somewhat lower in a more recent period. But there actually aren't a lot of good maps of Canadian snowfall out there from what I've seen.

Albuquerque had its second coldest September high since 1892 today, trailing only September 27, 1936 (46F). It was 47 today.

Whoops wrong map. This is Canadian snowfall - https://www.ccin.ca/home/sites/default/files/snow/snowfall_map.pdf

For the Chama to Moriarty zone, snow amount in NM so far seem to be 1-5" in populated zones from 6500-8500 feet, with the mountains seeing more.

I do like 2007, and I'm going to use it as one of my winter analogs but to me, there are still a couple problems with it. That said here is what I like about it: - La Nina with a relatively low-ACE Atlantic (currently the ACE is near average for 9/8, but the little tropical storms out there now only add around 1 point a day). Historically, the half way point for ACE is 9/13. So it's definitely becoming harder, fast, for a very active (~160+) ACE season to develop. - La Nina with a low-ACE NE Pacific - the NE Pacific is well below average. It's not real common to have both basins below average in a La Nina, but 2007 had both. - A lot of the recent La Ninas (1995-96, 2000-01, 2005-06, 2016-17, 2017-18, as well as cold Neutral 2013-14) had weakly positive or flat out positive PDO configurations Nov-Apr. I don't see that this year. The PDO could actually be the most negative it has been in 30 years for Nov-Apr from what I'm seeing. Will rival 1990-91, 1999-00, 2007-08, 2008-09, 2011-12 for a top five spot anyway in the last 30 years. - 2007 followed a 27.3C El Nino in winter 2006-07. A La Nina this year would follow a 27.1 El Nino in 2019-20. - 2007-08 was near the solar minimum. This is actually a bit of a wildcard since solar activity has been rising, slowly, since early 2020. - You did have pretty low sea-ice in September 2007, like this year. A lot of the recent La Ninas had higher sea ice (relatively). The cold ENSO years when sea ice drops below 4.3 million square km seem to be better in the West (2007, 2012, 2016) than those that don't (2010, 2013, 2017), with 2011-12 (4.3m) on the border, as it was cold in NM/TX but not elsewhere. The issues with 2007-08 are still substantial to me though: - I don't think this La Nina will be as strong as 2007-08, especially in Nino 4. - The QBO isn't a big factor for me, but I don't like the fact that it is near opposite of this year. The 2007-08 La Nina was very negative, trending up. This winter should be positive, trending up. If the trend matters more than the positioning, it is fine, but if the positioning (negative v positive) matters more, than it is a problem. - The Atlantic is warmer than 2007. - The 2007-08 La Nina continued into 2008-09. I'm not super convinced this will be a two year event, especially if it comes in a lot weaker than 2007-08 in Nino 4.

Happy snowy September 9th from the Southwest. In NM, the snow got down to about 7,000 feet. Only fell to 41 in Albuquerque unfortunately (although the record for September is 35 since 1931). All it took to go from 90-100 degree readings to snow was six hours of 40-70 mph winds in most places. Who knew? Believe it or not, some of the years with similar ENSO setups that I like do actually have fierce cold shots as a feature of the pattern in the SW in the second week of September. One of the more interesting years to look at is 1936, although I'm not going to use it as analog. 1995 actually had fairly intense cold dump into the SW in September. I'm getting a little concerned that there may be some very intense (although 1-3) dumps of cold into the West this winter. It's increasingly looking to me like the PDO is going to be the most negative it's been in a long time for one thing, I think below -1.0 is possible for Nov-Apr. Red River at 8,600 feet had it's earliest snow on a July-June basis since at least 1906, they have reliable weather records there for 1906-2014 in that town. It's not uncommon to snow there in September, but getting more than four inches is rare in September, and the prior earliest snow had been 9/17. Outside Santa Fe - Los Alamos -

This is Red River around 8 pm - would you say you guys had more or less by then?

I saw gusts to 71 mph at Albuquerque on sustained winds of 47 mph out of the East at one point. But to be honest...that's actually pretty common here. The east wind is the main reason Albuquerque gets 10 inches of snow long term rather than 15 or 20, the wind just destroys the snow. I was looking at my method for calculating what the PDO will do in Nov-Apr (in aggregate), and the best match I got was 1961-62. PDO was around -0.4 for March-August on the Nate Mantua method (I think it will be this year too when the Mantua PDO comes in sometime in the next week), and I think Nino 1.2 will be around 20.0C on the monthlies for October (they tend to run much warmer than the weeklies). Anyway...a lot of the closest years for Nino 1.2/March-Aug PDO blends have essentially a super -PDO. I wasn't really expecting that, but it'd be interesting if it verified. Still have to see the August PDO value though, and see if there are any last second changes with Nino 1.2. I mention this because you can get unusual cold/dry air dumps into the West if the PDO goes really negative. It's not a cold signal at all for the East.

Red River, New Mexico, at 8,600 feet above sea level has reliable weather records for 1906-2014. The local NWS is calling for six inches of snow with this storm in that town. That would be the third highest September snowfall on record there - and only behind 1936 and 1971. Getting up to 4 inches of snow in September is pretty common at that elevation (~once a decade) but more than that is pretty rare. For the 1906-2014 period, there is also no record of accumulating snow earlier than September 17th.

Red River, at 8,600 feet above sea level in Northern New Mexico has reliable snow data for 1906-2014. It's never snowed in September before 9/17 in that time frame. The forecast for six inches would be third most snow in September for the 1906-2014 period too.

This stuff is surprisingly hard to find, but the trend in 2007 at the subsurface was pretty clearly to rapid cooling. There was a brief plateau, and then the bottom really fell out in September In any event, 2007 is the much better match, not just in the tropics, but certainly in the local weather. It makes more sense too, it's a ~27.3C El Nino transitioning to a ~25.0C La Nina in 2007, while 2010 was like a 28.1C El Nino to a 25.2C La Nina - much bigger change for only a year. To stay on pace with 2007, I think you'd have to see the subsurface fall to -1.0 at least by the end of the month.

My primary issue with 2007-08 is that Nino 4 eventually cooled off a lot too - to around 27.0C. It took an enormous amount of subsurface cooling to do it though. https://www.cpc.ncep.noaa.gov/products/analysis_monitoring/ocean/index/heat_content_index.txt YR MON 130E-80W 160E-80W 180W-100W 2007 8 -.16 -.42 -.68 2007 9 -.35 -.69 -1.03 2007 10 -.52 -.87 -1.19 2007 11 -.54 -.97 -1.19 2007 12 -.49 -.87 -1.08 In the current setup, the subsurface cold seems to be slowly thinning, not deepening like in 2007.