raindancewx

1973 with a warmer Atlantic would probably be a 2-3 month window of very powerful, wet storms, with marginal cold for a lot of the Eastern US. You'd force a lot of the fluke super cold/strong storms to the north with southern heat already building up from the south. Atlantic was frigid in the early 1970s. 1972-73 is the snowiest winter/cold season on record in a lot of interior West. I don't expect that this year, but it's a pretty special winter for a lot of the country, including the southern US.

Every forecast I've seen for this month, including my own, looks way too cold now. We're talking about +10-15 anomalies already guys. I wasn't fucking around in late November when I said the heat would be similar in magnitude to Dec 2015. I never meant that it was going to sit exclusively over DC to Boston as you all took it. The northern third of the US is averaging +5 or something, and we're losing the "cold" aspect of early Dec with the AO/NAO flip. The WPO still looks super positive, i.e. "low north / high south" oriented for the coming weeks, so lows should continue to dump into Los Angeles. The CFS is garbage, but it does this coming the next week. Its hard to imagine it completely screwing up something that close to now. You're talking a pretty substantial expansion of +5 to +15 type warmth in the northern US if it's remotely close. In week two, when it has no skill, the +5 to +15 zone moves east. Just a quick reminder - this is how Dec 2009 looked. That's why I've been so cranky about that year. I don't have an issue with it for the Fall/and roughly Feb 20-Apr 20, but it's not a good match for the core of winter. The ND/Montana border looks like it could be literally 25 degrees warmer than December 2009, and that's the "path" cold takes to reach a lot of you.

MJO looks like it is going to rapidly weaken now. Not much support for a run through 8-1 at high amplitude. I think it dies in 7/8, maybe even 6. The upcoming massive dump of moisture into Eastern NM in phase 6 reminds me of Goliath in 2015, when the MJO was in phase six as well. Someone in the High Plains of eastern NM or TX could luck into 1.5-3.0 inches of liquid equivalent, with most of it falling as snow. I enjoy snow, but I would prefer to not see the cattle ranchers lose tens of thousands of their livestock in this event. I'm a bit concerned someone is going to get a major freezing rain event out of this as well. This system looks like it has much less cold than Goliath, so the snow/rain issue is going to be more of an issue than in 2015. Track will be key too, and the models never know what to do with cutoff lows over the high terrain. I've always enjoyed the observations from Roswell in late Dec 2015 - temps range from +4 to +71, with 16 inches of snow reported. 2015-12-25 71 29 50.0 9.0 15 0 0.00 0.0 0 2015-12-26 51 25 38.0 -3.0 27 0 0.16 3.2 0 2015-12-27 30 24 27.0 -14.0 38 0 0.28 12.4 3 2015-12-28 32 10 21.0 -19.9 44 0 T T 15 2015-12-29 31 4 17.5 -23.4 47 0 0.14 1.2 13 2015-12-30 28 8 18.0 -22.9 47 0 0.00 0.0 14 2015-12-31 37 10 23.5 -17.5 41 0 0.00 0.0 12 https://en.wikipedia.org/wiki/December_2015_North_American_storm_complex

The East-West timing component of the Bering Sea Rule is nearly infallible to me - just wish it was more reliable on the north/south axis. I'm expecting the MJO to pick up amplitude in 5-6-7, slow down a bit, and then crash pretty quickly in 8. We're in phase 4 at high amplitude on 12/5 in an El Nino. So the comps are not many. In terms of the rotation, we're about half way between 2014 & 2018 timing. CPC has been showing 1991 a lot in the 6-10 and 8-14 periods as an analog, that was the year in my group that I liked most for December in my analog group. Dec 1994 is basically dead on to MJO timing at the moment. ~ish. 6-7 transition is often pretty stormy here in December, but not sure if it that will work this year with the -PDO.

Here is current snow across the US - Going to take a little while to fill in with the warmth coming this month. November was remarkably dry actually, on a national basis, although you can sort of see where the storm track was. Fairly consistent track of storms from Reno to Brownsville and then east to Florida. I've mostly been skeptical of a lot of snow in the Northeast on a precipitation basis for winter, rather than temps. That weird island of deep red right over Boston makes no sense to me. But it was there when I did my analogs for the winter too if you go back and look at my outlook. The CPC outlook for mid-month today is essentially the warm WPO composite. In a lot of ways, I think the WPO almost never going negative is the real culprit for a lot of the "lock in" of patterns in recent years. It doesn't look like it responded as directly to the MJO 4/5/6 looks in the older days. CPC has the EPO correlated most strongly to the TNI in December (trans nino index, essentially 1.2 v. 4), so the warm look coming is a good reminder we're not in a Modoki setup. Or even close to it really. Just not enough cold around to match December 2009. Never been a fan of that year for the core winter. It'll get better as an analog again in March, maybe late February when this event begins to collapse from east to west.

100-180W Subsurface Heat Year Sept / Oct / Nov 2018 +0.98 / +1.47 / +1.25 2009 +0.76 / +1.04 / +1.75 2006 +1.13 / +0.80 / +1.35 Blnd +0.96 / +1.10 / +1.45 2023 +1.03 / +1.13 / +1.50 Subsurface heat for November warmed up less than in Nov 2009. The blend is consistent with my 'dueling cold pocket' idea for the winter, with one in the South and one in the West. https://www.cpc.ncep.noaa.gov/products/analysis_monitoring/ocean/index/heat_content_index.txt My winter outlook had a brief monthly peak around +2.0C at this time before rapid decay late winter. I still think that's the right idea. Canadian had La Nina conditions by late Spring on the new run.

Ladies and Gentleman, MJO phase five. Is it right? Who knows? But it makes sense conceptually. November and October were pretty shitty forecasts though in retrospect from the Canadian.

Where are these Decembers that are cold in the East with MJO in phase 4-5 at the start of the month? I'm sure you could find some ancient year when the Earth is vastly colder that did it. But the recent years with any resemblance to recent patterns are 2017 which is a horrible analog, and then 2014, 2015, and 2011. Also 1999 which is very warm. I guess you could throw in 2000 and 2017 but they just don't match recent looks in any way shape or form. I don't know why you guys care what I think. I sure as hell don't value the inputs in here. You've all been posting the same stupid maps of tropical forcing that have no direct relation to our actual observable for months. Find methods that predict the weather and use them.

When was the last time the East was cold when the MJO went into phase five in December? It's not rocket science guys. We're going into phases 4-5. The Pacific is record warm in Nino 4 and the Indian Ocean is dis-favorable for the MJO weakening in 4-5. The MJO is predictable. You just have to master elaborate concepts in mathematics...like counting. I don't know why you guys can't learn this stuff. Gee I wonder what will happen.

The models have been pretty bad so far. Although the CFS does have a warm December. It can at least see the tendency for systems to undercut the northern heat. My unweighted analog blend from August beat the Canadian (currently +3 to +6 in the Northern Plains, not cold as the Canadian shows). I did get cute in November by changing the weighting of the years in the forecast. But overall, the raw blend +1 for its age did just fine. Local snow, precipitation, and high profile for November is nearly identical to the blend. November is already at an inch of snow here (let's see if my troll gets annoyed), with a bit more possible by 11/30. We've had double our normal precipitation. First month to do that since October 2022 locally. Month will finish warm, but not much. It's been cold lately, and yes the cold lately is centered in the Plains/West. I would invite all of you to look at precipitation patterns - Nov was an interesting month for precip. Observed:

Subtropical jet looks like it is placed pretty differently to 2009 so far, despite that November being a pretty good temperature match nationally. After this month, I don't think the pattern will look much like 2009 for a while, likely until March. Northwest and Southeast moisture paths are completely different for November. Through 11/19, we're at about 0.5 inches in the city in 2023. All of Nov 2009 only had 0.04". The wetter El Ninos in my outlook group included 1972, 1982, 1997, with 1991 near-record wet, and then 1951 and 2009 quite dry. I've always liked the 1972, 1982, 1991, 1997 group more for precip than temps, so it's nice to see that working out so far. I think you can see on the precip pattern why I've been skeptical of a 2009-10 / Modoki outcome. I think the precip patterns will be off from the ideal Nor'easter track later on. A lot of times the Oct-Nov precip patterns that setup more or less repeat, shifting north/south with the day length.

My personal view is the models that have the Northeast / Midwest pretty warm are looking at the very warm waters consistently by Japan. They're seeing ridging there, and they have it continuing through winter. That's a +WPO look, and it's hard to get real cold with that in the East. I've been pleased with the progression of the Fall as an anti-log to 1993. That year anti-1993 look has been very useful so far. It's super cold, coldest north, while 2023 is very warm, warmest south. Fall overall should finish like a blend of severe -PDO years with some warm Nino 1.2 years blended in there. To me, Fall looks a lot like 1947. Lot of other years with similar patterns to 2023, that are a step or two below 1947 including: 1940 (El Nino), 1956 (La Nina), 1960 (Neutral), and then arguably 2004, 2007, 2011, 2016, 2021. Good luck I guess Easterners? 1983 isn't too bad either. These were not my analogs, but if you had to pick the most similar Fall temperature profiles for the US, I think its maybe 1940, 1947, 1956, 1960, 1983, 2004, 2007, 2011, 2015, 2016, 2021 - very similar to the blend shown below.

Nice to see more rain and snow coming for me. Driest 12-month period for any time frame, i.e. Nov-Oct, Jan-Dec, May-Apr, etc back to 1892 is around 3.5-4.0 inches here. We were flirting with those record dry periods in the absence of rain in July. Starting to reverse now. It's almost like the subtropical jet is responding to an El Nino or something.

Bering Sea Rule implies a major system pulling through the Southwest US in the Dec 5-9 time frame. Broadly consistent with my analogs. Almost all of the six analogs had a snow event between Dec 8-12 for the SW. Major lows south of Kamchatka tend to pass over New Mexico in 17-21 days. Unweighted analog blend from me from October has represented November fairly well, with more heat coming in the short term.

The anchovy supply collapse this year is another feather in the cap of the Peruvian fishermen from day of yore. Inflation/Deflation in certain types of foods directly impacted by ENSO is another metric you can use if Nino 3.4 SSTs, MEI, ONI, SOI, RONI, PDO, NOI, forcing, MJO precipitation patterns and the PNA indicators aren't enough. https://weather.com/news/climate/news/2023-11-06-el-nino-anchovy-shortage-peru Niño was born in the Pacific Ocean off Peru. Hundreds of years ago, fishermen in the region noticed that the water warmed every few years in their winter, and their catch declined. They called it “El Niño de Navidad,” or "The Christmas Child." Peru has the largest anchovy fishery in the world. “Most of the anchovies in the world in terms of tonnage come from Peru,” John Phillips, procurement director at Roland Foods, told us. The country produced about 4.6 million tons of the tiny savory fish in 2022, according to industry reports. Anchovies feed on plankton forced to the surface by upwelling, which happens when winds push water across the ocean surface and colder water rises up from beneath. El Niño disrupts that process.

The time from the solar min to the solar peak actually looks pretty consistent with the rule posted above. The cycle min was the year ending February 2019, at 2/sunspots per month, and the cycle peak looks like it will be the year ending November 2023, so that's 4 years and 8 months, for a 123 sunspot/month peak, v. 5 years for 116/month in the chart above.

I'm increasingly open to the idea that we're actually passing the solar peak for this cycle. If you look at the data, we hit ~122 sunspots/month for the year ending September 2023. That's up from ~2 sunspots/month for the year ending Feb 2019. Oct & Nov 2023 both look like they could finish at 115 or less. We've not had two months in a row that low in a year or so. But they'll both beat Oct/Nov of 2022. But is December going to bounce back to over 110 after a few months below? Is January going to bounce back to 150 or something? I have my doubts. 2022 09 2022.705 96.0 16.3 1264 2022 10 2022.790 95.5 16.0 1215 2022 11 2022.873 80.5 16.2 1047 2022 12 2022.958 112.8 16.6 860 2023 01 2023.042 144.4 29.4 968 2023 02 2023.122 111.3 20.7 1014 2023 03 2023.204 123.3 17.9 1081 2023 04 2023.286 97.6 18.0 1132 2023 05 2023.371 137.4 19.6 1240 2023 06 2023.453 160.5 20.0 1248 2023 07 2023.538 159.1 17.3 1039 * 2023 08 2023.623 114.8 15.4 1095 * 2023 09 2023.705 133.6 17.6 1140 * 2023 10 2023.790 99.4 16.0 958 * I mention this because the 'ascending solar favors blocking' thing is likely moot if we're already starting to wind down. The last solar cycle peaked in October 2013-September 2014 at ~117 sunspots/month, so conceptually the timing is feasible. Especially since the cycles since 1762 have been observed to range from 9-13 years. It's not an automatic 11 year cycle. 2013 10 2013.790 114.4 8.2 421 2013 11 2013.874 113.9 8.2 321 2013 12 2013.958 124.2 9.1 402 2014 01 2014.042 117.0 8.2 398 2014 02 2014.123 146.1 10.7 384 2014 03 2014.204 128.7 8.6 493 2014 04 2014.288 112.5 6.9 486 2014 05 2014.371 112.5 7.5 493 2014 06 2014.455 102.9 7.7 469 2014 07 2014.538 100.2 7.4 477 2014 08 2014.623 106.9 7.6 486 2014 09 2014.707 130.0 8.7 419

Same PDO influence as what you'd expect over the past week. Should continue on/off for a while. The subtropical jet has been active though.

We're on the board here for snow. Parts of the city had a dusting to an inch, with 0.3" at the official site. Timing remains pretty close to my analog "snow signal" composite. These were periods when measurable snow showed up in multiple analogs. We had no rain or snow in town from 9/19-11/09, which is unusual in an El Nino. But I had two of the 10 lowest precipitation totals for that period as analogs with 1991 (0.26") and 1951 (0.42"). I've mentioned since 2015 that there is a direct negative correlation between Albuquerque and Philadelphia snow in El Nino. That relationship has been getting stronger with subsequent El Ninos (2015-16, 2018-19, 2019-20).

I'll take it.

Webb's datasets are pretty flawed. You'll see him talk about 30 blends of data washing out errors. At the end of the day though, the ship reports / ICOADS / sparse official observations pre 1950 all have pretty systematic flaws that can't really be teased out well enough to detect the older weak events for El Nino or La Nina. Beyond that you just don't have many SST observations in the older decades south of 50S until pretty recently, and you'd want to be able to look for the "PDO" ring in the Southern Hemisphere as well, as an example of an ENSO response that did or did not happen. Call it what you will, but CPC has the El Nino / -PDO result I've been expecting for months. You have storms showing up on the models undercutting a very warm northern US, with the West turning wet despite the supposed El Nino = +PNA look. For the moment, 2009 is a good analog to the pattern. There are real reasons to believe it won't continue. As we get closer to December, we should start to see the more canonical blazing East / cool West look that shows up with heat in Nino 4 / MJO zones 5 / +WPO years. Here is another correlation map - you can see we're still moving to the -PDO looks.

These measurements are like splitting hairs over the facial features in identical twins. The whole point of these seasonal forecasts is that you're supposed to look at everything, not just different versions of the same thing.

Which do you guys like now? RONI went up, MEI went down for the recent updates. For me, the issue with the MEI is timeliness as much as anything. Takes forever to update. If you map out correlations of Nino 3.4 SSTs, v. the MEI, v. the SOI, they're all basically the same long-term for sensible US weather, at about the same correlation rate. The RONI thing on the other hand is just overkill. CPC already normalizes Nino 3.4 against a running 30-year average in the Pacific. No need to subtract out the other oceans. The Indian is warming faster than the Pacific but not dramatically so. In any given ENSO event you can also have a temporary out of trend blip in the Atlantic or Indian v. the linear tendency in warming. As far as the PDO, the version of it I use is still extremely negative, at -1.7 in October. Link is down at the moment. https://oceanview.pfeg.noaa.gov/erddap/tabledap/cciea_OC_PDO.htmlTable?time,PDO In general, you have a fairly strong winter negative correlation between the +AMO/-PDO. -PDO and warm Atlantic conditions are associated with La Nina. This year we had an active and very warm Atlantic, so it's not really surprising to see the PDO hanging on. https://psl.noaa.gov/data/correlation/table/corr.table_dec.txt

Here is a list of El Ninos with a -WPO look in October. It's actually pretty common even in the stronger El Ninos. But it tends to flip back hard. The November look on the Canadian is relatively canonical for the transitional start to flip back, with a low pressure weakness by Kamchatka, but not quite over it, with that low north of high pressure by SE Asia. You can see how the WPO performed in my analogs at the weightings I used in my outlook below. -WPO El Nino October 1951, 1976, 1986, 1991, 1997, 2002, 2009, 2014, 2015, 2018, 2019 Off the top of my head, the +WPO / -EPO / +PNA / -NAO / -AO look for October is probably fairly similar to 1988 and 2003 along with the obvious match of 2009 that you all want.

Looking back, the biggest difference in the El Ninos that are cold in the Southwest v. the Southeast seems to be how persistent the WPO look is. The persistently mixed or negative WPO looks show like this - Persistently positive are more like this - The Canadian has the big area of enhanced sinking air by Japan and south, like the +WPO composite. So I think the second map is more likely. The WPO and EPO have some tendency to move with the PDO, but the WPO seems more tied to the warmth of Nino 4 / the Indonesian warm pool. There just aren't many -WPO years now, with thunderstorms usually enhanced in MJO phases 4-6. The reason I like the Canadian is the precipitation patterns it shows usually match what happens in the MJO zones. So then you can test to see if the rest of it makes sense. The look of MJO 8 usually suppresses the enhanced rain running east just north of the equator in the composite - which we don't see on the modelling.