raindancewx

You guys aren't doing half bad for winter so far relative to the rest of the country. I think going forward, El Nino should be viewed more in the context of elevation than it is now. The mountains of New England & the Southwest are probably the only areas of the US with a real shot of a cold winter still. Maybe the Northern Plains too, Billings was similarly warm through last January and still finished cold, because their weather is stupid amazing in term of variation, it was 22F below or something in February after something like a +7 Dec-Jan. For most of the US, a month 20F below average is lottery odds to physically impossible. These are 12/1-1/3 temps v. 1981-2010 averages nationally. Roger Smith had mentioned the mid-January torch in his mid-October forecast - pretty good. Nashville: (8.0 x 31) + (3 x 11.3) / 34 --> +8.3F Billings: (6.6 x 31) + (3 x 12.3) / 34 --> +7.1F Chicago: (6.3 x 31) + (3 x 13.0) / 34 --> +6.9F Kansas City: (6.4 x 31) + (3 x 12.3) / 34 --> 6.9F Jacksonville: (5.4 x 31) + (3 x 9.7) / 34 --> +6.3F Atlanta: (5.9 x 31) + (3 x 7.3) /34 --> +6.0F Detroit: (5.0 x 31) + (3 x 13.0) / 34 --> +5.7F Amarillo: (5.0 x 31) + (3 x 5.0) / 34 --> +5.0F New Orleans: (4.0 x 31) + (3 x 10.3) / 34 --> +4.6F Seattle: (3.9 x 31) + (3 x 6.2) / 34 --> +4.1F Salt Lake City: (3.6 x 31) + (3 x 5.5) / 34 --> +3.8F Denver: (3.7 x 31) + (3 x 3.3) / 34 --> +3.7F Pittsburgh: (3.0 x 31) + (3 x 10.7) / 34 --> +3.7F Richmond: (3.0 x 31) + (3 x 9.5) / 34 --> +3.6F Houston: (3.2 x 31) + (3 x 6.0) / 34 --> +3.4F Boston: (2.5 x 31) + (3 x 13) / 34 --> +3.4F San Francisco: (2.9 x 31) + (3 x 3.0) / 34 --> +2.9F Oklahoma City: (2.3 x 31) + (3 x 7.0) / 34 --> +2.7F San Diego:(2.2 x 31) + (3 x 2.4) / 34 --> +2.2F Philadelphia: (1.2 x 31) + (3 x 5.5) / 34 --> +1.6F Albuquerque: (1.3 x 31) + (3 x -2.2) / 34 --> +1.0F Flagstaff: (-0.2 x 31) + (3 x -4.0) / 34 --> -0.5F

I wanted to bump this - Roger's idea of an eastern blowtorch mid-January from mid-October looks quite good. The MJO going into phase five at high amplitude should be good for NM/CO snow pack at least in mid-January. Here is what the US looks like to date..."before" the warm part of the pattern arrives. As I've noted elsewhere, the model pattern (CFS/Canadian/CPC) for January 2020 in the US looks a lot like mid-Sept to mid-Oct...as I hinted at 10/16. Nashville: (8.0 x 31) + (3 x 11.3) / 34 --> +8.3F Billings: (6.6 x 31) + (3 x 12.3) / 34 --> +7.1F Chicago: (6.3 x 31) + (3 x 13.0) / 34 --> +6.9F Kansas City: (6.4 x 31) + (3 x 12.3) / 34 --> 6.9F Jacksonville: (5.4 x 31) + (3 x 9.7) / 34 --> +6.3F Atlanta: (5.9 x 31) + (3 x 7.3) /34 --> +6.0F Detroit: (5.0 x 31) + (3 x 13.0) / 34 --> +5.7F Amarillo: (5.0 x 31) + (3 x 5.0) / 34 --> +5.0F New Orleans: (4.0 x 31) + (3 x 10.3) / 34 --> +4.6F Seattle: (3.9 x 31) + (3 x 6.2) / 34 --> +4.1F Salt Lake City: (3.6 x 31) + (3 x 5.5) / 34 --> +3.8F Denver: (3.7 x 31) + (3 x 3.3) / 34 --> +3.7F Pittsburgh: (3.0 x 31) + (3 x 10.7) / 34 --> +3.7F Richmond: (3.0 x 31) + (3 x 9.5) / 34 --> +3.6F Houston: (3.2 x 31) + (3 x 6.0) / 34 --> +3.4F Boston: (2.5 x 31) + (3 x 13) / 34 --> +3.4F San Francisco: (2.9 x 31) + (3 x 3.0) / 34 --> +2.9F Oklahoma City: (2.3 x 31) + (3 x 7.0) / 34 --> +2.7F San Diego:(2.2 x 31) + (3 x 2.4) / 34 --> +2.2F Philadelphia: (1.2 x 31) + (3 x 5.5) / 34 --> +1.6F Albuquerque: (1.3 x 31) + (3 x -2.2) / 34 --> +1.0F Flagstaff: (-0.2 x 31) + (3 x -4.0) / 34 --> -0.5F

Year DJF JFM FMA MAM AMJ MJJ JJA JAS ASO SON OND NDJ 2010 1.5 1.3 0.9 0.4 -0.1 -0.6 -1.0 -1.4 -1.6 -1.7 -1.7 -1.6 2011 -1.4 -1.1 -0.8 -0.6 -0.5 -0.4 -0.5 -0.7 -0.9 -1.1 -1.1 -1.0 2012 -0.8 -0.6 -0.5 -0.4 -0.2 0.1 0.3 0.3 0.3 0.2 0.0 -0.2 2013 -0.4 -0.3 -0.2 -0.2 -0.3 -0.3 -0.4 -0.4 -0.3 -0.2 -0.2 -0.3 2014 -0.4 -0.4 -0.2 0.1 0.3 0.2 0.1 0.0 0.2 0.4 0.6 0.7 2015 0.6 0.6 0.6 0.8 1.0 1.2 1.5 1.8 2.1 2.4 2.5 2.6 2016 2.5 2.2 1.7 1.0 0.5 0.0 -0.3 -0.6 -0.7 -0.7 -0.7 -0.6 2017 -0.3 -0.1 0.1 0.3 0.4 0.4 0.2 -0.1 -0.4 -0.7 -0.9 -1.0 2018 -0.9 -0.8 -0.6 -0.4 -0.1 0.1 0.1 0.2 0.4 0.7 0.9 0.8 2019 0.8 0.8 0.8 0.7 0.6 0.5 0.3 0.1 0.1 0.3 0.5 CPC has +0.5C for OND (+0.46C v. their baseline). 27.11C in December. Using my older 1951-2010 base base, that's a +0.66C reading in Nino 3.4 - still real weak. Until September, 1992 was easily the top match in Nino 3.4 since 1950. Starting January, 1993 should become a pretty good match once again going by surface and subsurface readings in Nino 3.4. QBO is still positive, for those of you who consider that important. 2004-05 looks like a solid match now. 2004 -4.84 2.61 5.45 10.46 12.97 11.75 9.96 8.74 7.29 8.00 4.35 2.45 2005 -0.69 -0.96 -0.33 -6.64 -15.09 -20.59 -24.20 -25.87 -27.80 -28.76 -29.55 -25.04 2019 9.02 9.25 11.82 13.36 14.59 14.36 10.96 9.97 8.25 7.27 5.07 1.66

Just for fun, here is the list of years since 1931-32, where Albuquerque has had more snow than DC, Baltimore, NYC, and Philly combined through January 2. The four NE cities average 5.1, 3.7, 3.4, 2.8 respectively for total snow through the end of December...or 15 inches in total in an average year. Albuquerque averages 1/4 of that - 3.7 inches through December in an average year. Looks a lot like what the models have for January. Completely different for February. Almost all are good Oct-May periods for snow out here, except for 1954-55, 1949-50, 2001-02.

MJO phase five at high amplitude mid-month should be good for NM & CO for moisture. My original analogs for winter, in my October forecast had 1/12-1/16 as a storm time frame that showed up in multiple years. That looks pretty good. GFS has had something coming through in that range for a few runs now, and the big SOI crash last night supports it....and obviously the MJO should be in phase five by then.

If we go by the prior min, the solar cycle is about to bottom out. 2019 had 3.6 sunspots/month, down from 4.2 in 2008. Absolute min was 2.2 in the year ending May 2009 in the prior cycle. I find that certain types of unusual events are far more likely to happen in the US below 55-sunspots/month annualized, at statistically significant levels over long-periods. I don't think it changes total heat on Earth much, just the distribution of it, mostly via effects on clouds. Albuquerque just had its coldest year since 2008. All the years at the min are actually pretty cold here. We had the 28th coldest annual high since 1931. Lows are not really impacted by solar at all, and were still real warm. City had 10.6 inches of snow in 2019 - most since 2015. I'm pretty sure Albuquerque has had more snow so far that NYC, Philly, Baltimore and DC combined this cold season. Outside the NW, I think Western Snow Pack and Reservoir numbers overall are pretty strong heading into 2020. The precipitation portion of my outlook has been pretty good so far, so I'd expect a more active January than December. Probably will be when/if the big Eastern heat wave comes since MJO phase five is pretty wet in the Southwest in January. MJO getting to phase four around 1/6 is similar to 2005, 2007, 2013.

Subsurface heat came in warmer in December than November, as expected. https://www.cpc.ncep.noaa.gov/products/analysis_monitoring/ocean/index/heat_content_index.txt This is the top match I could find, with each month within 0.1. 100W-180W Oct Nov Dec 1981 0.25 0.02 -0.22 1981 0.25 0.02 -0.22 1986 0.95 0.52 0.97 1986 0.95 0.52 0.97 1986 0.95 0.52 0.97 2005 -0.14 -0.57 -0.74 2018 1.59 1.36 1.06 Blend 0.69 0.34 0.40 2019 0.70 0.26 0.34 Blend is less warm nationally than the Canadian/CFS.

The SOI objective top matches for Oct-Dec correctly flagged the February pattern last year two months in advance. This year, the blend shows a nearly opposite setup. The SOI blends never get the magnitude of the heat or cold right though. For whatever reason, the December SOI in particular is a much better indicator for February than January. The CFS/Canadian/CPC outlook for January looks like Sept 16-Oct 15 temp anomalies in the US, after Aug 16-Sept 15 looked like December temp anomalies. I know the matches aren't exact, the idea is you had a very warm December nationally, but the same areas as Sept 16-Oct 15 were ''protected" relatively, and the same areas that were cold Oct 16-Nov 15 look cold in January, with heat concentrated somewhere in the East. If that is right for January, it means there is a pretty strong chance Oct 16-Nov 15 will be about right nationally for February, and then you guys cash in on cold/snow in March. 2016-17 has been a close NAO match in the predictive time frames, it "seems" right. Basically, I think March may be the best month for snow in the NE - the subtropical waves of moisture in the SW that energized your systems in late Nov-early Dec should return in March going by the timing of the MJO and the pattern overall. February I could see being frigid in Central areas of the US. Blending NAO monthly changes for May-Apr, and Sept-Mar works pretty well. In those time frames, 2016-17 was a good match...and March 2017 had the -NAO. We'll see. None of this will work if the models are completely wrong foe January. Continuing the pattern forward, April would probably be very cold in the Rockies and very hot elsewhere.

That 8.5" December for Boston my analogs had held pretty well it looks like, despite over 7" the first couple of days. My precipitation analogs have been pretty awesome so far for a blend created in Sept/Oct. Two distinct dry areas and two distinct wet areas both basically in the right spots to date.

Canadian has a much more extreme pattern than CPC & the CFS. Looks like a blend of January 2013/2017 to me. The old run (Nov 30) on the right actually is pretty similar. The simplest SOI pattern composite for Oct-Nov-Dec (1957/1951) also agrees with the Jan outlooks from the CFS/Canadian. MJO years with phase four around 1/6, like the current forecast in include 2013, 2007, and 2005 since the MJO data began, with 2013 as the best match. Canadian pretty much nailed December nationally.

The Canadian has a powerful La Nina developing Apr-Jun 2020, which it didn't show before. Long-term the Canadian has the PDO going pretty strongly negative too. It's also more on board with a dynamic cold West / hot East pattern than the CFS. It's closer to the 2013/2017 blend than the El Nino blend. Outside the Northern Plains where it went super warm, the Canadian forecast was basically solid for December - I'd give it a B+.

The Canadian has El Nino conditions through January, and then La Nina developing in April/May 2020 It is on board with the CFS for the cold West / hot East January pattern. Interestingly it sort of had that look figured out Nov 30. So far, my precipitation forecast for winter from October 12th, 2019 has been going better than my temperature forecast (I had Warm/Cold/Warm for West,Central,East respectively in December - the cold didn't really verify Central). Will be interesting to see if that continues.

The SOI was very different this year in Oct-Dec. Implies radically different Jan/Feb than in 2019. The match composite was basically correct last year too. The left figures are for 2019, the right from 2018. Amplification was wrong, but the SOI blends had the right spatial ideas for Jan-Feb 2019.

The CFS has trended to a cold January for WY/CO/NM. I'm not sure I buy it, but it is nice to see. If you wanted a recent set of years that look like its forecast, you can do January 2013/2017 as a blend. I prefer a more El Nino blend, so I did 1906, 1952, 1958, 1993, 2019. The CFS has a dry West for January, which I don't really buy at all, at least for the WY/CO/NM zone. I have January 1952/1958 in there, as 1951/1957 are top SOI matches to Oct-Dec in the past 90 years, and then 1992/2018 are in there because they look like the January forecast as a blend and were in my winter forecast as analogs. I threw in 1905 for some other reasons. I'm expecting the Southwest get more active when the cold begins to move out.

Nino1+2 Nino3 Nino34 Nino4 Week SST SSTA SST SSTA SST SSTA SST SSTA 27NOV2019 22.5 0.4 25.4 0.4 27.0 0.4 29.3 0.8 04DEC2019 22.5 0.1 25.3 0.2 26.9 0.3 29.4 0.9 11DEC2019 23.1 0.5 25.5 0.4 27.1 0.5 29.5 1.0 18DEC2019 23.3 0.4 25.5 0.3 27.2 0.6 29.5 1.1 25DEC2019 23.5 0.3 25.5 0.2 27.0 0.4 29.4 1.0 The weekly site doesn't have the new numbers, but they are here - https://www.cpc.ncep.noaa.gov/products/analysis_monitoring/lanina/enso_evolution-status-fcsts-web.pdf Here is the simplest blend I could find using El Nino-ish years that look like the CFS for January. 1951, 1957 are the top SOI matches Oct-Dec. 1992, 2018 are the top matches from my winter forecast. I threw in 1905 for some other reasons. If you want a super recent blend that matches the CFS, January 2013/2017 is close too. I suspect the colder blend is more realistic but we'll see. Also think the CFS might be pretty wrong. The top SOI match for Oct-Nov-Dec (pending 12/31) is likely 1951/1957. The January 1952/1958 blend looks a lot like what the CFS has. I mentioned this in one of the regional threads, but the Dec temperature map will bear a lot of resemblance to 8/16-9/15. The CFS January forecast looks a lot like 9/16-10/15, Cold West / Warm East. February would then be the 10/16-11/15 period i.e. it would be very cold in the middle of the US. This is all assuming the CFS is right for January. These aren't exact matches, but it shows a cycling pattern if the CFS is right. The -SOI in December is also a much colder signal for February 2020 compared to last year.

In an SOI sense, December has gone from a La Nina state in 2018 (+9.2) to an El Nino state in 2019 (-7.0 or so). The December SOI is highly correlated to temperatures in the following February for the NW & SE US. The +9.2 SOI was the only historical early indicator of the hot East / frigid West February last year. Even though the correlation map doesn't show it, the SW tends to get very cold when there is a flip from a La Nina to an El Nino (<=26.0C DJF in Nino 3.4 to >=27.0C DF in Nino 3.4 in one year). The biggest Nino 3.4 warm ups in a single year, from cold to warm, include, in recent years, 2006, 2009, 2018 - all quite cold here. I'll be very curious to see if there is a cold February out here in light of the "SOI base state" going from La Nina to El Nino. The top SOI flips Dec/Dec that match 2018/2019 well in the past 100 years include 1946, 1957, 1989, 1997, 2001, 2004, 2009. In a timing sense, Aug 16-Sept 15 has matched the Dec 2019 pattern fairly well for US temps, the CFS January forecast looks a lot like Sept 16-Oct 15. So if the CFS is right for January and the timing pattern holds, I'd expect Oct 16-Nov 15 to look like February 2020 - a frigid month for the central US. As a composite, the years I listed for the SOI Dec/Dec change favor the entire South cold and the entire North cold in February. I'd expect a blend of the two for February, maybe one part of the month is very cold across the entire South, VA to CA, and the other part of the month is very cold between the Rockies & the Mississippi. CPC uses 26.65C as the basis for Nino 3.4 average in December. So ~27.15C or so on the weeklies should count as a +0.5C month. ONI for OND will likely be right around +0.5C, despite CPC insisting on a 70% odds of Neutral conditions in Fall. CPC also tends to warm up the most recent old month or two by 0.05C or so when the new data comes in, which helps with the likelihood of a +0.5C reading.

My crazy theory at this point is that 8/16-9/15 resembled December nationally for temps (cold N. Plains/avg NE), 9/16-10/15 resembles what the CFS has for January (cold West/warm East). That would make 10/16-11/15 the likely pattern for February if the CFS isn't completely wrong for January. I'm assuming the 12/28-12/29 panels as a blend are right. So the whole middle of the US would be frigid, including the interior South, in February before a major warm up that last just most of March & April. The months are never going to be exact matches, but the patterns do repeat. All of this will depend on the CFS being right and nothing huge changing in the tropics. But the month to watch is February going by the progression, especially in the absence of the +SOI in December like last year. Dec had coolness north, warmth elsewhere. January - coolness is shown West. Warmth elsewhere. Now look at February.

Cold water is gone from Nino 3.4 once again. Ties in well with the big SOI response after the positive readings Dec 1-10.

I have four methods for predicting Spring precipitation here: NDJ + Solar conditions Closest Precip Patterns July-Dec Replicate Precip Patterns July-Dec Statistical Regression The middle two methods imply a pretty wet Spring. I have a lot more faith in the third method this year compared to the second, as it was very difficult to mimic July-Dec 2019 with historical data this year. Heavily reliant on 1905 & 2004 to do it, with other years in there as 'fix its'. Essentially, you try to match the data in black for July-Dec with the blend. The green is then the forecast from the rolled forward blend. You can see how well the blend did last year, even though I gave up on November.

Forecasts for this storm were kind of a mess. Simultaneously more moisture advection than forecast - much higher dew-points - and then less actual precipitation at the same time. The line of showers that came with the cold front went literally due east until the Rio Grande River, and then went Northeast, missing most of Albuquerque, so I got literally no rain or snow. My winter forecast was premised on near average to wet conditions in the SW, despite few storms with a lot of precipitation. That still looks about right. Fall was well below average for precipitation and had fewer than average wet days, until November saved it. December has had two days with 0.1" or more precipitation in Albuquerque. I have a statistical regression for winter precipitation here, it said 2.10", +/-0.8" for winter in the city, at 95% certainty based on hindcasts. We're at 0.27" through 12/27. The signal is pretty strong for a wet January, around 0.75" for Albuquerque (double normal) based on how wet November was. Following the 10 wettest Novembers, January was wet nine times. Following the 20 wettest Novembers, January was wet sixteen times. November 2019 was the wettest November in over 100 years here. Somewhat weaker signal for a wet February too. It's very early, but if the wet January verifies, I'd also expect May to see the usual weather insanity, but amped up to 11 this year, with large hail, blizzards, tornadoes, cold snaps, heat waves, rain and thunderstorms for NM & CO and the High Plains. Nov-Jan wetness in Albuquerque is a strong predictor of an active May. The 2.26" from Nov-Dec 27 is already a top 20 Nov-Jan period for wetness locally in the last 100-years, with five weeks to go. A top five Nov-Jan is pretty likely here. Those Springs would be 1941, 1979, 1992, 1993, 2005. The CFS seems to want to roast the middle of the country for January based on the entrenched dryness. I don't really buy the dryness it shows for the West in January though. Pretty big area of the US +5F or more, if taken literally. That little blue area in Mexico makes me think the CFS isn't really "seeing" past the first 5-10 days of the month. The snow storm for Mexico will probably not be enough for that one area to be cold if everything around it is warm.

CFS continues to trend to a very warm January for the East. Warm ~entire East January is not a common outcome in El Nino years. Since 1930, it has happened in January 1931, 1946, 1952, 1964, 1992, 1995, 1998. 2012-13 isn't an El Nino, but it had the +IOD and a similar (colder) subsurface trajectory Oct-Dec. 1954, 1964, 1993, 1995, 1998, 2013 as a January blend may be what it is trying to do. Most of these years already had a lot of Eastern warmth, more than this year, in December. So I remain skeptical of the CFS solution. If you shift the warmth it shows to the NW somewhat and weaken the warmth everywhere, it looks like the El Nino 'warm East' January composite, which may be where the CFS is going. Not what I forecast - but we'll see what happens. If you throw out January 1946 and 1964, the weakest "warm East January" analogs, the composite shifts east. The 1930-31 El Nino was an El Nino following an El Nino. 1994-95 followed a neutral/warm Neutral. So did 1991-92. 1997-98 & 1994-95 had very strong positive IOD fall patterns, like 2019. So it 'makes sense' as a blend. Except 1991-92, all low solar years - 60 sunspots/year or less. Warm AMO years too. Canadian will be out 12/31 - will be interesting to see what it shows. Looks like the weeklies on the CFS are pretty cold in the SW as my methods and the composites above show.

The 100-180W waters below the surface in the Tropical Pacific look like they'll come in around +0.3 for December 2019. Using the data, a blend of 1986, 1991, 2000, 2012, 2018 looks pretty close to 2019 for Oct-Dec. 100W-180W Oct Nov Dec 1986 0.95 0.52 0.97 1991 1.41 1.22 1.71 2000 -0.37 -0.67 -0.96 2000 -0.37 -0.67 -0.96 2012 0.40 0.34 -0.27 2018 1.59 1.36 1.06 Blend 0.60 0.35 0.26 2019 0.70 0.26 0.30 The blend looks somewhat like what the CFS is showing for January 2020, although it will change its mind 20+ times by 1/1/2020. Want to see the final number next week for the subsurface, but the blend is warm in the middle of the US, somewhat cooler east/west. The SOI has gone ballistic in recent days. A blend of Oct-Dec 1951 & 1957 is currently close, but should continue to change through 12/31. Those values of -20 for consecutive days are typically trouble for someone in the US. Year Oct Nov Dec 1957 -0.3 -11.0 -4.3 1951 -12.3 -8.5 -8.3 Blend -6.3 -9.8 -6.3 2019 -5.2 -9.5 -6.7 Date Tahiti (hPa) Darwin (hPa) Daily Contribution 30 day Av. SOI 90 day Av. SOI 26 Dec 2019 1007.10 1008.85 -28.39 -6.48 -6.79 25 Dec 2019 1006.63 1008.30 -27.97 -5.81 -6.46 24 Dec 2019 1008.40 1008.85 -21.64 -5.33 -6.12 23 Dec 2019 1009.75 1007.95 -9.97 -4.97 -5.81 22 Dec 2019 1009.29 1007.80 -11.57 -4.77 -5.73 21 Dec 2019 1009.10 1008.15 -14.38 -4.47 -5.65 20 Dec 2019 1010.74 1008.65 -8.46 -4.38 -5.62 19 Dec 2019 1010.39 1008.00 -6.90 -4.76 -5.74 18 Dec 2019 1009.79 1007.60 -7.94 -5.44 -5.98 17 Dec 2019 1008.73 1007.60 -13.44 -6.09 -6.28 16 Dec 2019 1008.58 1007.75 -15.00 -5.94 -6.36 15 Dec 2019 1009.50 1008.65 -14.90 -5.38 -6.30 14 Dec 2019 1010.24 1008.70 -11.31 -4.87 -6.20 13 Dec 2019 1010.74 1007.75 -3.79 -4.71 -6.20 12 Dec 2019 1011.38 1007.85 -0.99 -4.92 -6.46 11 Dec 2019 1012.71 1008.00 5.14 -5.19 -6.77 10 Dec 2019 1012.35 1009.10 -2.44 -5.68 -7.10 9 Dec 2019 1012.01 1009.20 -4.72 -6.09 -7.20 8 Dec 2019 1013.30 1008.65 4.83 -6.30 -7.27 7 Dec 2019 1014.25 1009.15 7.16 -6.46 -7.56 6 Dec 2019 1014.24 1009.75 4.00 -6.17 -7.88 5 Dec 2019 1012.70 1010.55 -8.15 -6.53 -8.09 4 Dec 2019 1012.06 1009.10 -3.95 -7.41 -8.12 3 Dec 2019 1012.67 1007.20 9.08 -8.42 -8.21 2 Dec 2019 1012.38 1007.35 6.80 -9.07 -8.48 1 Dec 2019 1012.31 1007.80 4.10 -9.40 -8.74

The GFS has been sending a snow storm pretty deep into Mexico early in 2020. Will be interesting to see if that verifies. The huge recent SOI drop suggests it is possible.

December looks fairly similar to anti-1997, with 1940/1976 working fairly well as a blend so far. I doubt January will be as cold as the 1941/1977 blend or anti-1998 would imply. Long term, a positive NAO in December is a pretty strong warm signal for the mid-south in January. NAO is around +1 for December so far. I do think there are some good cold shots in January, but we'll see. A negative SOI December is a weak warm signal for the Northern half of the US in January too, not much of a cold signal anywhere though.

GFS still has storm timing faster later this week. The big SOI drop 12/22-12/24 ties in fairly well with the storm shown around 1/3 on the GFS, despite it being super far out.