raindancewx

The group of years you picked is pretty terrible for the last full three months. Its completely wrong for the jet stream pattern over Asia for three months, and that determines what will happen for the build up of cold air in the Fall for us in the USA from both Russian and Canadian sources. You can already see the issues with it - because Asia is opposite the correspondence of the 'sin wave' so to speak puts the wrong signature over Greenland. My simple blend had it warm in October, yours is cold. The warmth I had over Northern Canada/Alaska is what will matter for the winter, and I showed it very warm for the period shown. You don't, you have the half degree warm / cold muddled crap that prevents any chance of your outlook getting US temps right later on when cold air is sourced from the North.

The cold wave in the West we've seen this week did show up at this time in 2013 for what its worth. I blend 2013 & 2024 for a reason - its not as cold/widespread as 2013 but definitely not as warm as 2024. The cold has displaced the warmth further east than last year.

From what I can see, no relevance for Dec, Feb, Mar for East Coast temps where it snows / where nor'easters are important. For the alarmists out there you can make the case whatever importance that was there in January is also rapidly diminishing - almost all of the East Coast in the 10+ inch avg snowfall climo zone is now in the borderline irrelevant correlation zone. The deeper blue zone is way south of where it is on the longer term map...

I've never understood the fixation with the PNA for you guys - its completely irrelevant for temps for the parts of the East where snow is heavily dependent on exact temps in big events. Storms and precip and patterns - sure. But who cares if you get storms if you can't determine the temp pattern from it? Its a pattern that matters for the NW and SE US, for temps in winter not the NE and SW US. These maps by the way show r, not r-squared. So anything below the 0.3-0.5 even for a long period of time is inherently unreliable. So I agree on the ENSO --> PNA relationship, but the point still stands - its not a big deal for the East where you guys live for temps in winter. Using Enso to predict PNA state is fine I guess but it just doesn't tell you much in the East in winter and hasn't ever really.

I've been telling you for literally a decade that ENSO has no direct correlation now, nor in the past to what happens in the East Coast for winter. So El Ninos are cold. Some aren't. Same for La Nina. Same for snowy/unsnowy. What is the point of these posts?

This is what Claude told me when I gave it this prompt - the newer versions of Claude are a lot better on pattern recognition than the older versions. What kind of winter do you expect in the USA following a September that looks like 2025? Looking at this 500mb geopotential height anomaly map for September 2025, I can identify several key features that would influence the upcoming winter: Key Pattern Features: Strong negative anomaly over Alaska/North Pacific (deep purple/blue) - indicating troughing Positive anomaly over western North America (yellow/green) - ridging pattern Negative anomaly over the North Atlantic - suggestive of trough there as well Strong positive anomalies in the high latitudes (around 60-90N over northern Canada/Arctic) Expected Winter Implications for the USA: General Pattern: This setup suggests a ridge-dominated western US and trough-favored eastern US pattern could persist into winter The strong Alaskan trough with western CONUS ridge is consistent with a negative Pacific-North American (PNA) transitioning toward positive or a split flow pattern Regional Expectations: Western US: Likely warmer and drier than average, especially in the Southwest and Great Basin Reduced storm activity in California Possible drought concerns Central/Eastern US: Greater likelihood of colder outbreaks reaching the Plains and East Increased storminess along the East Coast More variable temperatures with periodic Arctic intrusions Pattern Analogs: This resembles aspects of weak La Niña or neutral ENSO winters with a tendency toward cold East/warm West splits. The strong high-latitude blocking hints at potential for polar vortex disruptions, which could mean occasional severe cold snaps in the eastern 2/3 of the country.

I've been quite happy seeing the extra moisture this October. A lot of big systems (Lows) are forecast by NE Asia for a while, that's typically consistent with big systems moving through the SW at a 17-21 lead time this time of year. So I do see some support already for an active November. That's implied to some extent with the ACE to ABQ cold day correlation as well. I could easily be completely wrong about the winter, but its hard to see a warm winter nationally - I could see it regionally (SE/SW, or both). Serious question - I haven't looked - if you did a difference in proportions test for seasons without a US hurricane hit v. with a US hurricane hit, aren't the seasons without a hurricane hit far colder, or more frequently colder at least? I can't think of many winters following no US hurricane landfall in the US that are very warm, and consistently so.

Just for reference - NAO was different. But globally its pretty similar. Blocking should show up at times, its just when? I'm assuming the super persistent warmth/high pressure by Japan means on balance the NAO is not predominantly favorable for Nor'easters in the dead of winter int he Northeast. But during the transitional periods (Nov/late Feb-late Mar) I am more optimistic for big systems by the NE than in recent years.

Forgot this. I have the counts for cold waves in the 80-120 ACE La Ninas locally. For whatever reason, the tendency shows the low-solar, low ACE La Ninas seeing severe cold locally late (February) while the higher solar, low ACE, La Nina years see it early (Nov-Dec). So low solar Feb 1955/1985 have the most days 5F or more below average locally. But Dec 1971, Dec 1988, Nov 2000, Nov 2022 have far more solar activity. The years that go to El Ninos, 1971 and 2022, also see higher numbers of cold days in Nov-Dec locally in these La Ninas. My best guest is one big cold outbreak mid Nov-mid Dec for cold, and then smaller outbreaks later. La Nina, 80-120 ACE # Of Days by Month, AVG Temp 5F or colder than daily mean in ABQ Year Nov Dec Jan Feb Nov-Feb 1954 1 9 10 15 35 1971 6 15 8 5 34 1984 6 7 8 11 32 1988 9 11 9 5 34 2000 17 4 12 6 39 2022 14 5 5 7 31 Blend 9 9 8 8 34 # Of Days by Month, AVG Temp 10F or colder than daily mean in ABQ Year Nov Dec Jan Feb Nov-Feb 1954 0 3 3 12 18 1971 0 7 3 3 13 1984 2 3 4 7 16 1988 2 5 5 4 16 2000 10 0 3 0 13 2022 5 0 0 2 7 Blend 3 3 3 5 14 For July-June - 1954-55 is around 19 sunspots, 1984-85 is around 26. 1971-72 is about 101. 1988-89 is around 188. 2000-01 is around 163. 2022-23 is around 110. 2025-26 is probably lower than 2024-2025 (149) but still in the 1971, 2022, 2000, 1988 range.

I don't necessarily expect the cold to show up as severely as it does on those maps. It probably will come in warmer by 1-3F in two of the three months in Jan-Mar, and then as severe as depicted in the other(s). Taken verbatim the weighting would make Jan-Feb as cold almost as both 2014 and 2025 nationally, which is pretty unlikely. I also don't expect the WPO to flood Canada/Alaska with warm air all winter. If models are right the PDO/ENSO should move toward their warmer cycles and you should see a corresponding change in the Northern Hemisphere patterns. The 100-180W heat content at the equator down to 300m got to -1.33 last in Jan 2025 but is trending warmer/weaker this year, so you should see an easier breakdown of whatever the dominant pattern is late winter when the ocean heat content warms rapidly. It went all the way to -0.06 by March 2025. If we top out at -0.8 to -1.0 this year, the flip to 'warmth' that coincides with the surface pattern ENSO flips should happen a lot faster. https://www.cpc.ncep.noaa.gov/products/analysis_monitoring/ocean/index/heat_content_index.txt

By the way, I settled on this for analogs for temperatures for the cold season: Oct, Nov, Dec, Jan, Feb: 2013-14 (x4), 2024-25 (x4), 2018-19, 2022-23 Mar, Apr: 2019 (x4), 2023 (x4), 2014, 2025 Couple pretty cold months in there, but its not really long-lasting or consistent where it shows. Snow anomaly looks like a smiley face, going from Montana-Idaho border to NM-CO border in the West, Ozarks shooting NE to interior New England/Maine for heavy snow. In between (OK/NE New Mexico, North TX, etc) will have some nasty ice storms. I went below average for snow northern plains. I expect very little snow here for most of the actual winter, but there are windows in November, late Feb-late Apr that should be productive. For winter precipitation I used the Mar/Apr weighting for wetness/dryness. Jan-Mar has retrograding cold shots as the La Nina rapidly collapses while cold air is still around. So by March its Plains/West, but for January Plains/East. Edit: Conceptually, I settled on these blends for a couple reasons - the upper air pattern in July-Sept does look like a simple blend of 2013, 2018, 2022, 2024 globally. SSTs are close too. That's not a bad match for the US in Sept or Oct either. But the straight 2013/2024 dominated blend is much closer for US temps than the equal weight blend. In 2024-25, the cold was highly concentrated in January nationally and then to the east in February. My La Nina Atlantic ACE to Albuquerque Nov-Feb cold day count index implied 23 days would be at least 5F below the daily average. We had 24. This year with ACE still only 96, the implication is 31 days at least 5F below the daily average. I still expect the La Nina warm signal to show strongly, but the cold wave signal is pretty strong. I'm expecting either one-two major cold waves or several smaller cold snaps. The lower ACE also implies locally "severely cold days" - 10F or more below daily average - should be around 12, instead of the projected 7 last year (we actually had 8 or 9 I believe). The precipitation blend is still predominantly dry in the southern US, but much wetter than 2024-25. Additionally, a lot of the hurricane seasons with near no to no activity in the Gulf end up pretty cold late - 2013-14, 2010-11, 1981-82, 1962-63, 2014-15, 2022-23 - if going into an El Nino or trending warmer in Nino 3.4 y/y.

Relatively canonical +WPO look for temps in Oct so far - if displaced slightly. November is very different though if it holds.

Cold patch this month in the West should expand. Not really that similar to last October when the West was blazing. The cold season still looks much better than last year to me too. You could already see what we were heading into by the end of October last year - Lot of drought nuking rain coming the next five days. That near record setting storm in the North Pacific is a feature of the winter by the way, not a bug. You will have very powerful storms in the winter.

Still looks fairly close. Both the blend with and without 2007 are decent US temp profile matches for Sept 2025.

This article is pretty consistent with my thinking on how to model winter precipitation patterns in the US. It's strongly supportive of a wetter winter nationally than last year, despite what the models show. https://opensnow.com/news/post/how-atlantic-ocean-temps-could-impact-la-nina-2024-2025-winter-forecast Recent seasons on the Canadian have been depicted in an MJO pattern of 4-5 for winter. The models are showing 6-7 for this winter. That's very different from recent winters, look at the depicted east African precip pattern at the equator. Look at Brazil shown wet v dry in phase five. It's not 5 like in recent years. It looks like 6-7 to me, not perfect but better than the others. We had 4-5-6 for precip nationally last winter, but made drier by the Atlantic features. Again - this match holds up for a lot of my methods, although it isn't quite what I expect. But it did have 106 ACE as an example which still feels about right.