raindancewx

The 30-day SOI has run up to +14. If we hold at +10 to +18 like in December 2007, you should see pretty strong warmth develop in the South either in or by February. None of the years in the +10 to +18 SOI range are cold in the South in February. It's arguably the strongest correlation for the SOI of the entire year. The coldest major +SOI December years are still 'slightly warm' in the South the following February.

Those four years above as a blend are very warm for January. I've been looking at the January cold-ENSO years following two El Nino winters. Those are pretty cold in the West pretty reliably. 1929-30/1930-31 El Ninos ---> Jan 1932 1957-58/1958-59 El Ninos --> Jan 1960 1968-69/1969-70 El Ninos --> Jan 1971 (may look warm in the West, but featured a record cold shot and record warmth) 1976-77/1977-78 El Ninos --> Jan 1979 1986-87/1987-88 El Ninos --> Jan 1989 2014-15/2015-16 El Ninos --> Jan 2017 I think I'd add 3.0F to this in the Plains for lack of snow cover and +1.5 everywhere else since the analogs are centered on the 1970s.

When I did my forecast for winter, I knew 2007-08 had a +WPO just from looking at temperature patterns in Asia and the US, but I didn't realize there was a formalized index for it. It's a weird time frame to do a forecast, but the WPO is super useful for trying to figure out if Feb-Apr will be cold or warm out here. In the meantime, CPC still has the Western US cold for a while, especially the interior. I was doing an analysis of my analogs earlier today for 12/1-12/21 v. US observations. I'm pretty close in a lot of the eastern Midwest, Northeast (outside Maine), interior West, and the edges of the South (VA, Arkansas, Oklahoma). I'm not close at all Northern Plains and Deep South, but the deeper we get into the month, the more the month has looked like my blend. One of the ideas in my forecast was at least some cold periods in the interior West this winter, based on low sea ice in the Summer + cold ENSO. That's at least partially right. If you use Albuquerque as a proxy, we haven't hit 60 this month. Even on a 'warm' day like today when we'll be mid-50s, it was 22 in the morning, which prevents the highs from getting out out control. I'm curious to see if the areas outside the Northern Plains and Deep South can trend toward what I had more by the end of the month.

I'm starting to see a lot of speculation about an active tornado season. One thing that is against that is the current WPO / Nino 4 blend. The WPO is a very strong signal for a cold SW / warm Central US in April if it is positive, and the opposite if negative. A blend of a +WPO December and a cold Nino 4 December (27.6C or so) is unusual. The blend I get is something like this: That favors a pretty -WPO in April, which is actually a warm Southwest/cold Central pattern, i.e. not real tornadic. The WPO correlations change dramatically in May. But if you want a big tornado season, you should be rooting for the WPO to hang on positive through April. WPO Feb Mar Apr 1953 1.38 0.58 -1.86 1972 -2.26 -1.87 -0.01 1985 -1.12 1.81 -0.53 2008 0.01 -1.35 0.28 2009 1.71 0.22 0.02 Blend -0.06 -0.12 -0.42

The Euro still has a system in the time frame you'd expect after the SOI crash. I would expect something to come through. I realized yesterday there is monthly data for the West Pacific Oscillation. We flipped from an extremely negative WPO in October to a positive WPO in November-December. That's useful for preventing winter warmth out here. I hope it continues in Feb-Apr, because it's a very strong cold signal for the Southwest in that time frame. The WPO correlation to the US in October and what actually happened in October is basically uncanny, it's a powerful little index at certain times.

I realized yesterday that there is monthly data for the WPO, the West Pacific Oscillation, which is actually super powerful as a predictor in some months. Check out October (-2.44 was actually the WPO September reading, should say -1.18 on the graphic for Oct): Since October, the index has been positive. Check out December, v. the rolled forward +WPO, +PNA, +AO, +NAO blend for November: I don't think the +WPO, +PNA, -AO, -NAO December look is right for January. It looks pretty off to me. But we can see. The look is essentially January 2000, which didn't have the +WPO or +PNA, and I'd bet on both continuing at least a bit in January. Arctic is probably wrong too. But just thought I'd show the composite. The WPO is actually very powerful at out here for Feb-Apr, so I am hoping it stays positive.

Nino1+2 Nino3 Nino34 Nino4 Week SST SSTA SST SSTA SST SSTA SST SSTA 04NOV2020 20.3-1.0 23.6-1.3 25.1-1.5 27.8-0.8 11NOV2020 21.2-0.3 24.1-0.9 25.7-1.0 27.9-0.7 18NOV2020 21.0-0.7 23.6-1.4 25.2-1.5 27.9-0.7 25NOV2020 21.0-1.0 23.7-1.3 25.3-1.3 27.9-0.7 02DEC2020 21.8-0.5 24.4-0.7 25.5-1.1 27.6-1.0 09DEC2020 22.4-0.1 24.3-0.8 25.4-1.2 27.7-0.8 16DEC2020 22.0-0.8 24.3-0.8 25.6-0.9 27.8-0.7 I don't know that Nino 1.2 averages really change as rapidly as depicted, but that's the only region that cooled off by actual temperatures this week. Nino 3.4 is at 25.5C for the month, which is about -1.0C for 1951-2010. Nino 4 is still pretty cold. Nino 1.2 averages 22.58C for 1951-2010 in December, so that's still the laggard in terms of coolness. The 20.86C last month was about -0.6C against 1951-2010.

The +NAO Nov, -NAO Dec thing is pretty unusual if it verifies, regardless of ENSO - but there are no La Ninas with the combo. These are the only years to do it since 1950. You can see it is a mixed bag for January. Of the years in the chart 1953 is closest to US conditions for Nov-Dec. Also very curious to see the weekly data tomorrow. Should give a good idea if December is really going to finish warmer than November in Nino 3.4. The SOI is around +13 for the past 30 days despite the big crash recently. Has been similar to 2007-08 since October on that indicator. https://www.longpaddock.qld.gov.au/soi/

Will be interesting to see how the winter transitions after December. Currently looks like Dec 2017 and Dec 2018 as a blend.

My winter forecast included the idea that the NAO would go negative in December and then revert positive for Jan-Feb based on 'NAO analogs' which match the 2020 NAO in predictive periods to similar behaving predictive periods. Will be curious to see how that goes. The NAO is a much more important as a national feature for Jan-Apr than at other times.

Date Tahiti (hPa) Darwin (hPa) Daily Contribution 30 day Av. SOI 90 day Av. SOI 20 Dec 2020 1011.45 1005.80 10.02 12.80 9.49 19 Dec 2020 1012.11 1002.50 30.57 13.03 9.54 18 Dec 2020 1012.34 1002.05 34.10 12.68 9.35 17 Dec 2020 1013.64 1003.35 34.10 12.22 9.08 16 Dec 2020 1013.54 1004.30 28.65 11.67 8.85 Keep your eyes on the end of the month. That's a pretty massive drop for one day. The SOI is still positive, so it's probably a northern stream system with some pop if something is going to come through. The SOI crashed pretty hard in a very positive month about this time (12/18-12/19) in 2018 ten days before the end of the month system. 2018 352 1014.64 1006.70 21.90 2018 353 1013.25 1008.05 7.68 This is late 2016 - 2016 355 1010.63 1002.35 23.66 2016 356 1010.94 1003.25 20.60 2016 357 1009.60 1005.95 -0.36 2016 358 1007.23 1006.85 -17.33 Pretty hard to find 20 point drops in a day at this time of year.

Date Tahiti (hPa) Darwin (hPa) Daily Contribution 30 day Av. SOI 90 day Av. SOI 20 Dec 2020 1011.45 1005.80 10.02 12.80 9.49 19 Dec 2020 1012.11 1002.50 30.57 13.03 9.54 18 Dec 2020 1012.34 1002.05 34.10 12.68 9.35 17 Dec 2020 1013.64 1003.35 34.10 12.22 9.08 Those 20 point crashes are rare...but fun usually. No idea if this will work, but if you run objective "closest high" matches for 22 US cities and then pull out the most common top-10 close matches as analogs, you get a pretty good map for December, followed by an extremely warm US winter. I would say, very generally, either 2011 or 2012 or both were common western matches, while 1999, 2011, 2017 were more likely in the East. Overall, 2001 was the year that most commonly appeared as a top 10 match, which surprised me a bit.

Nino 4 in the prior winter is fairly predictive of what the NAO does in February. Bodes better for February 2022 in the East with Nino 4 likely to finish in the low 27s. But Nino 4 was 29.19C last winter. Not a good sign for this year. Above 28.7C in Nino 4 the prior winter, only 1 in 20 winters is below -0.2 for the Feb NAO. February NAO averages +0.6 for years following Nino 4 over 28.7C in winter, but averages -0.1 in the remaining colder years. This is your +NAO La Nina composite - although I'd expect a somewhat different result.

I use this for snowfall data. Locally it is imperfect and has some small data issues, but it's 99% right. https://xmacis.rcc-acis.org/

Let's say the -NAO holds on for December. The transition from a +NAO November to a -NAO December is extremely rare: only seven years do it from 1950 to 2019. Those years are 1953, 1958, 1967, 1976, 1978, 1989, 2001. None of these years are La Ninas. But, 1967 is close, and 1978 and 1989 and 2001 are cold-Neutral. The closest match to US temperatures in Nov-Dec is actually 1953 though and it isn't close. But: 1989, 2001 are next, and then 1958. I would weight the years like this for January: 1954 (x3) - closest temperature match Nov-Dec, and low solar 1959 (x1) - follows an El Nino 1968 (x1) - closest to a La Nina 1979 (x1) - cold-Neutral following two El Ninos 1990 (x2) - OK temp match, neutral 2002 (x2) - OK temp match, neutral

The Weather Channel says winter is over. Don't really buy it personally. I wouldn't expect the South or Southwest to be cold overall, but I think January won't be super warm out here, and I doubt March will be real warm in the Southeast or Northeast. Southwest should roast in March, but I don't buy it getting into Wyoming or Nebraska either. It may be more of a +3 look than the +5 or +8 they seem to be implying.

Looks like 6.6" total for Philadelphia, 10.0" total for Central Park, and 12.9" for Boston (Logan). Those figures are all within range of my five winter analogs.

As of Midnight eastern: Central Park had 6.5" Philadelphia had 6.3". The difference is, Philadelphia has had rain/freezing rain in recent hours, while NYC (Central Park) is still reporting snow at this hour. From 1931-2019, Philadelphia has only had at least six inches of snow on a December calendar day 12 times - so this was pretty solid. The bigger snows that can happen are almost always later in the season. Philadelphia went over to plain cold rain for about an hour around Midnight, so I'd imagine some to most of the snow is gone already.

Snow totals for Philadelphia, New York City and Boston are not final yet, but for reference: In Boston, from 1931-32 to 2019-20, from October-May, 84/89 years see four or fewer days with at least six inches of snow. It's not official yet, but I'd expect Boston to top 6 inches on 12/17. The median from 12/18 to 5/31 is one day with six inches of snow. I would go above the median though for Boston, with 3-5 days over six inches, since my analogs had four for Boston. In New York City, from 1931-32 to 2019-20, from October-May, 82/89 years see two or fewer days with at least six inches of snow. It's not official yet, but I can't see NYC not topping 6 inches today. The median from 12/17 to 5/31 is one day with six inches of snow. So I would expect two days overall with 6"+ this winter, instead of the median of one. In Philadelphia, from 1931-32 to 2019-20, from October-May, 85/89 years see two or fewer days with at least six inches of snow. It's not official yet, but I think Philly will (just) top 6 inches today. The median from 12/17 to 5/31 is zero days with six inches of snow. So I would expect just this one day overall with 6"+ this winter, especially with the snowstorm today and in October following my winter outlook idea, with rapid deterioration in totals from Boston to NYC and NYC to Philly, i.e. more extreme cut off than the normal 41"/25"/22" seasonal totals.

Jamstec has an El Nino trying to develop in Nino 1.2 in Summer. Sort of like the 1996-97 to 1997-98 transition but weaker/slower.

Not bad for Western Colorado yesterday.

Where is it then? That snow in the north here has been there since September.

It looks like a lot of the Northeast has soil temperatures in the low 40s still. That's not super warm, but it doesn't help either. Usually shave off 20% from the model depictions if the ground is above 40 out here. I'd have to look but you guys were probably in the 60s in the 2009 storm before it happened, but this isn't as ideal as that setup either.

The weeklies imply December is around 25.45C in Nino 3.4 so far. It was 25.31C officially for November. The models show Nino 3.4 warming throughout the winter. So if December finishes around 25.5C, the winter should be above that. The last winter to finish below 25.5C in Nino 3.4 was 2010-11, so this is going to be one of the longest stretches in the data without a major La Nina. For my purposes, a La Nina warmer than 25.5C in Nino 3.4 in winter isn't moderate. On the CPC ONI standard, there will probably be three trimesters below -1.0C: SON, OND, NDJ. I don't think December to February will qualify. On the Jan Null standard, it will just barely qualify as a Moderate La Nina. Against 1951-2010 averages, it's pretty likely that only October & November are 1.0C below average or colder, with only OND likely to qualify as -1.0C or colder. I mention all this because with the current readings and forecast warming, this event may not end up materially stronger in winter than years like 1995-96, 2011-12, or 2017-18, which all had Nino 3.4 at about 25.75C in winter. The Fall peaked more impressively than those years though, which is interesting.

Actually think Uncle Sam has the right idea here in most spots. I'd go a bit lower for NYC/NJ/Philly/Western CT. The short range models are good at detecting sleet and small warm profiles above the ground from what I can remember in college when we evaluated model performance in Nor'easters.