Ramblings of Meteorologist Jim Sullivan

The Key Points Up Front:

1. Any views here are mine and are done for fun, and don’t represent anyone else’s
2. The pattern of “cold west, warm east” continues through most of this week. Cold and fairly active weather will continue across the western U.S.
3. Later this week into this weekend, cold spills into the central and eastern United States. The pattern then remains fairly chilly across the Lower 48 for a couple of weeks, with the coldest overall conditions across the northern Rockies and northern Plains
4. This colder pattern may be somewhat dry overall. With that said, some snow may occur from a lee cyclone late this week/early this weekend across the Upper Plains/Midwest…then, lake effect snow, a minor clipper or two, or perhaps a wave riding along the baroclinic zone (if there’s any sub-tropical jet activity) may bring some potential for snow farther south/east next weekend or beyond. Not a big winter storm pattern, but not hopeless for a bit of snow.
5. Several signs point towards the weather pattern in much of the continental United States trending mild yet again towards the end of November and early December
6. A few analogs and early signs in extended guidance suggest the potential for increased high-latitude blocking and colder weather may return into December. Given the ongoing La Nina and -PDO, confidence in this occurring isn’t high yet. However, the coldest December since at least 2016 is distinctly possible if things align properly.

The Details:

After a very chilly first half of October, which was enough to plunge most of the eastern U.S. to below normal values for the entirety of the month, mild weather has dominated the pattern over much of the central and eastern U.S…this pattern continues into this upcoming week, but the GFS and Euro ensembles both indicate below average temperatures over fairly decent stretch for almost the entire CONUS closer to mid-month:

It’s getting to be the time of year where colder anomalies have some more bite and can start bringing snow to more of southern Canada and the Lower 48. This cooldown has trended more and impressive as it moves up in time on the models, evidence by this GFS ensemble trend loop for the same time period:

After such a prolonged mild spell, is this cooldown shown in the medium to long range feasible, and how long may it last?

Looking at a few things, including the character of the Pacific jet stream and the MJO, suggests that the projected cooldown on the ensembles has legitimacy. Let’s start with the pattern over the last week leading up to Friday:

The pattern is undoubtedly a warm one for the central and eastern CONUS and into southeastern Canada, with a positive AO (PV generally tight and confined to the high latitudes), a very positive EPO (the trough over Alaska and western Canada), a negative PNA (Aleutian ridge and West Coast trough), and positive NAO (trough over the north Atlantic).

The above pattern correlates very nicely to the expected pattern when the Pacific jet is retracted, as it has been most of late October and very early November per the above chart:

The flat ridge south of the Aleutians, trough over Alaska and western Canada/the western US, and downstream ridge over the eastern US are quite characteristic. There are a couple of ways to analyze how the Pacific jet is progged to behave on current guidance over the next week or so…one is just to throw in a loop of ensemble mean forecast 250mb winds:

We start off in the middle of last week with a “jet retraction” look, with the jet barely extending off of Asia and a ridge over the Aleutians, directing the jet stream towards Alaska and the northwest coast of North America, helping facilitate the current very mild pattern across a good portion of the Lower 48. However, quite a bit of westerly momentum is getting added as I type this and into early this week. Initially, the ridge near the Aleutians just amplifies and disperses a lot of this added momentum around it, but eventually the jet extends and pushes the now-amplified ridge east, facilitating cross-polar flow into North America and really cutting off the Pacific influence by this upcoming weekend. Here is the GEFS forecast Pacific jet character:

A similar trend, from a pronounced jet retraction in late October towards neutral, with perhaps some modest lean towards an equatorward shift or jet extension over the next week or so. Here are what a Pacific jet extension or equatorward shift tend to look like:

Note the trend towards much higher heights into the northern Pacific and Alaska, with resultant troughing and colder weather over a good portion of North America…in particular, over western and central Canada into the northern Rockies and Plains with an equatorward shift, but bleeding towards the East Coast. The equatorward shift looks more like a -EPO but still somewhat negative PNA, while the jet extension looks more like a classic +PNA with a somewhat negative EPO. This generally matches the ensemble’s forecast pattern over the next week or so:

The impetus for this change seems to be a combination of increased east Asian mountain torque and tropical forcing working out of the eastern hemisphere and into the western hemisphere early this month. Initially, the pattern this weekend into the first half of the week simply looks like a significant amplification to the going pattern:

At first, we just see the ridge south of the Aleutians amplify, which amplifies the downstream pattern over North America (western trough and eastern U.S. ridge). However, this very mild look over the eastern U.S. doesn’t last too much longer as the ridge amplifies into Alaska, forcing cold south towards the CONUS, and then east as an Aleutian low develops. An Aleutian low is a trend towards a more positive PNA.

Note positive MSLP anomalies (high pressure) dropping into eastern Asia right around now, signifying a significant increase in east Asian mountain torque, which results in added westerly momentum to the east Asian-Pacific jet. There may be another push later this upcoming week, before a fairly well agreed upon trend towards a much more negative east Asian mountain torque into mid-November:

Positive East Asian mountain torque tends to result in an extension of the Pacific jet and subsequent rise in the PNA, as the extended jet and associated left-exit region towards the north-central Pacific tends to favor low pressure over the northwest Pacific or Aleutians and subsequent height rises to the east closer to the West Coast of North America.

At the same time, the MJO has come out of the Maritimes and into the western Pacific at a fairly high amplitude, and is now quickly working into the western hemisphere. The satellite/VP anomaly loops over the last two weeks show a similar trend as the above RMM plot, with strong uplift over the western Pacific until the end of October quickly shifting east. Note how areas that had the strongest upward motion (~120E) at the beginning of the loop have strong sinking motion now, and vice versa over portions of the central and eastern Pacific. Clearly, the tropical forcing is changing from the forcing that led into our current very mild pattern.

After a decent Phase 7 passage, an increase in blocking near Alaska in October-December is quite common:

We are not going to get the NAO help in the short term, but blocking located near Alaska/a -EPO is often common during and following a phase 7 MJO this time of year as seen above. The MJO is about to briefly get into Phase 8, which also often leads to a colder outcome into the southern and eastern U.S. than a Phase 6-7 in the fall:

And for what it’s worth, the same October-December lagged composites from the CPC for a phase 8 MJO, which is briefly occurring right now:

The issue with these is that the pattern right now into this week, which would be a “Lag=0” for the current phase 8 RMM indication, is that the NAO is quite positive with a deep trough out west. That said, a phase 8 MJO this time of year often corresponds to immediate and short term blocking, with an Aleutian low/+PNA and colder trend over most of North America. This is then followed by the PNA trending more negative and blocking generally breaking down within a few weeks of the phase 8 MJO occurring.

Despite the phase 8 MJO pointing to a -NAO, I am fairly confident it will remain positive for the next 10-15 days. This is because the stratospheric PV, while getting stretched a bit at times, is going to remain on the stronger side of normal, which argues towards a +NAO for now. In addition, recall back to this EPS mean sea level pressure/anomaly loop:

Note multiple high pressure systems dropping in just east of the Rockies over the next 10 days. The Rockies are not quite as tall as the Himalayas, but they’re still plenty tall enough to produce a mountain torque that requires increased momentum in the jet stream to conserve angular momentum in this equation (the positive mountain torque removes momentum by slowing Earth’s rotation slightly). This speeds up the jet over the eastern U.S. into the North Atlantic, encouraging positive NAO trends. Given that the NAO has been positive since late October, I don’t see it flipping negative in the face of this through at least the middle of November.

Looking Beyond Mid-November:

A dramatic shift to a colder pattern in the CONUS, first into the Pacific Northwest over the next few days and then expanding into the rest of the CONUS through next weekend and through mid-November is agreed upon by the major ensembles and seems very well-supported by various forcing mechanisms…driven by a plunging EPO and rising PNA. But how long will it last, and does it mean shit for the approaching winter season as a whole?

To answer the first part of this question, we will examine things such as mountain torque and tropical forcing yet again. Starting with mountain torque, and picking up closer to mid-November, after our upcoming increase in mountain torque and injection of westerly momentum into the Pacific jet:

Positive east Asian mountain torque is ongoing at the start of the loop, but as low pressure develops over much of southeast Asia, this becomes negative, reducing westerly momentum into the Asian-Pacific jet around mid-November. This suggests the jet will have a tendency to retract again, which would render the upcoming PNA spike brief and could eventually cause the EPO to start trending more positive again.

Remember, we are in a La Nina with a solid -PDO, evidenced by the cool tongue across the equatorial Pacific, the warm waters over the northwest Pacific, and the cool waters south of Alaska:

These both favor a -PNA and often favor a predominantly +EPO as well. We look to buck that over the next week due to tropical forcing progressing through the Pacific and Western Hemisphere and due to a positive east Asian mountain torque, but when those shorter term pattern drivers subside, the lower frequency forcing can often dominate yet again. So, the expected negative east Asian mountain torque towards mid-November is a signal that a more -PNA and more neutral EPO is a distinct risk into the latter portions of November given the background forcing.

Most guidance (the European ensemble shown above) suggests tropical forcing quickly progressing through the western hemisphere over the next week or so, with eastern hemispheric forcing (in particular, over the eastern Indian/western Pacific oceans) becoming dominant once again towards the second half of November. Given the background state, this seems very plausible after the current higher frequency forcing moves through the western hemisphere in the shorter term. RMM forecast plots also indicate similar, with the ongoing Phase 8 MJO not lasting long and then becoming “indeterminant” (which usually means lower frequency forcing can exert more influence again). Thereafter, the MJO may move into phases 5 or 6 after the 15th. Here is the European ensemble forecast MJO progression through November 19th:

Recalling back to the above October-December MJO phase temperature correlation plot above, phase 6 is quite mild in the CONUS. This general evolution (eastern Indian/western Pacific Ocean tropical forcing becoming dominant again into mid-November) is well agreed upon by various medium to long range forecast models/ensembles. In addition, the water temperatures support convection and tropical forcing wanting to favor this general region again, as does where enhanced convection/upward motion has been most persistent in recent months near the equator:

So, the background pattern suggests a risk for a retracted Pacific jet and -PNA/+EPO to return when other shorter-duration forcing weakens. In addition, there’s fairly consistent indication that the shorter-duration forcing (mountain torque, MJO) will weaken and may in fact positively feedback with the “background” La Nina pattern into late November.

There’s enough support for a significant cooldown that it will take some time for milder weather to return, we’ll probably have a full 1-2 weeks of generally cold conditions (for this early) across the CONUS. However, many indications are that milder weather gradually returns from west/southwest to northeast in the second half of November.

In terms of snow potential over the next few weeks, the higher elevations over much of the western U.S. will remain cold and active through this week with a -PNA and dropping EPO. As the cold ejects east late this week, a lee cyclone may bring snow potential to the northern Plains and Upper Midwest. Thereafter, it looks like a fairly dry EPO-induced cool/cold shot into the central and eastern CONUS. Not all hope is lost, as some snow may occur via lake effect, weak clippers, or perhaps any waves riding along the baroclinic zone if the sub-tropical jet isn’t totally quiet. Likely not a pattern that supports a widespread, major winter weather event (especially in November), but it’s possible some minor or localized snow events occur outside of the mountains starting this upcoming weekend and through roughly Thanksgiving.

Some Fun With ENSO/MJO Analogs:

The rest of the winter is not yet lost though. In fact, the findings discussed below are what prompted me to make this whole thing a blog post instead of a Tweet thread. I just thought “there’s a lot here, let’s try to tie it together better than intentionally cryptic Tweets would”.

This much MJO activity in the fall leading into a La Nina isn’t too common. Since 1975, only 4 La Nina Novembers saw the MJO move into both phases 7 & 8 at an amplitude of greater than 1. Those are 1996, 2000, 2010 and 2016. Oh, there’s that fabled 2010-11 analog. We also have a very negative IOD and solidly westerly QBO. Let’s see if we can leverage those high-confidence, long-duration pattern drivers. We have a negative IOD this fall and will have a westerly/positive QBO through this winter, full stop.

2010 and 2016 had a westerly QBO

1996, 2010 and 2016 had a negative IOD

How did those 4 years play out during December?

All but 2010 had a solid -EPO, and 2010 still had very pronounced north Pacific ridging along with a a huge -NAO. All but 2016 had a -NAO. All but 2000 had a -PNA. All but 2016 had a -AO. Here is a look at the October-January monthly temperature and precipitation anomalies for these years (weighted slightly based on QBO/IOD matches):

For reference, here are October’s temp/precip anomalies:

Overall, the October pattern this year actually resembles those analogs, though with the pattern seemingly shifted west slightly. The very warm start to November likely dooms the eastern U.S. to finish with a somewhat mild overall month, but the pattern after this week may resemble these years in November fairly well. It’s at least in the ballpark enough that these analogs aren’t immediately invalid.

I believe how the MJO and mountain torque play out during late November and into December will heavily influence if the potential for a very blocky, cold, active December comes to fruition given the risks for a milder pattern (at least temporarily) returning by the end of November. Just looking at some various products, the EPS (already shown), GEFS, and Australian BOM MJO forecasts:

The GFS/Euro both have the MJO strongly and definitively progressing through phase 6 and towards phase 7 into late November. This could be indicative of a similar angular momentum/global wind oscillation orbit to what we are currently seeing, as per this CFS forecast:

An orbit through phases 5-7 (as has recently occurred) supports convection moving east towards the Dateline and eventually the western hemisphere, along with positive mountain torque and troughiness into the mid-latitudes:

With that said, we are fighting the background state and there is some uncertainty to how all of this progresses 10-20+ days out. Getting the tropical forcing to propagate east again (as the GFS/Euro ensembles are suggesting towards late November) and getting another push of increased east Asian mountain torque is likely critical towards the door being open for a very block and cold December with elevated snow potential across a good portion of the CONUS. If not, the chances for a mild start to the winter quickly increase.

It’s still a little early on in the game, but we’re getting to the point where we can start narrowing down what various “pattern drivers” may look like this winter and examine analogs and even some seasonal forecast guidance for clues. For the most part, I’ll focus on cool ENSO winters / La Ninas and expand from there, with emphasis on tropical SSTs / forcing and the QBO.

For those who just want the maps, here are my current thoughts on temperature and precip through the winter. I’m lumping March into winter because it’s a month that can still prove to be productive for snow for many areas, and because many analogs are chilly in March. The detailed write-up with reaosning, analogs, and a look at longer range seasonal models follows. I hope to follow up with more focused thoughts in November, time permitting…

My guesses on teleconnections:

AO: Solidly negative.

NAO: Near neutral or slightly negative overall, possibly large swings

EPO: Near neutral or slightly negative overall.

PNA: Negative, especially later in winter

A stronger or more central-Pacific La Nina could pose warmer risks, especially in the central and eastern U.S., as would a strong stratospheric PV early in the winter that’s coupled to the troposphere. A quick increase in sunspots could also be a warmer risk. If forcing ends up more focused farther west into the Indian Ocean it’d likely point to a warmer winter.

The La Nina staying weak to moderate and more basin-wide would point to a chillier outcome with more blocking being more likely. A weaker stratospheric polar vortex early in the winter, and tropospheric blocking developing in November, would also point towards a chillier / blockier outcome. Scandinavian or Ural ridging in October or November would also point towards a chillier and blockier direction, as would sufficient tropical forcing occurring east into the Pacific at times.

I expect a very up and down winter, with very cold air frequently available over Canada that occasionally works south into much of the central and eastern CONUS this winter, along with the Northwest. The Southwest will generally be milder with a somewhat weak sub-tropical jet. The southern Plains, Gulf Coast, Southeast and lower mid-Atlantic will probably lean mild due to sharp warm-ups when blocking relaxes and / or the PNA shifts negative, though a couple of periods of EPO-induced cold all the way to the coast are possible. The most persistent cold will be in the northern Rockies and Upper Midwest. The analogs really like the first part of winter (December into January, possibly starting to turn colder in November) for blocking and cold into the central and eastern US (with cold focused more on the northwest in February). The analogs have a -EPO, -NAO, and -PNA pattern in March.

I’m optimistic about snow in the northern Rockies and Pacific Northwest, along with from the Great Lakes and Ohio / Tennessee Valleys into much of the Northeast and New England. Along I-95 from Baltimore to DC and Richmond, events will be rainy at times when blocking relaxes so snow will struggle to reach or exceed normal, though there should be opportunities for wintry weather when blocking is strong enough to suppress the Southeast Ridge.

The technical write-up / thoughts:

Starting with SSTs…a few features stand out:

1. A developing La Nina that appears likely to be weak to moderate and likely “hybrid / basin wide”, perhaps with a modest east lean
2. While the North Pacific is warm overall, the PDO is decidedly negative (very warm waters near Japan, cool waters south of Alaska)
3. The western Pacific warm pool is quite robust, as has been more common than not lately
4. The Indian Ocean Dipole is somewhat negative, similar to last fall
5. While not entirely classic, the AMO is positive

We’ll start with a La Nina focused approach and then expand from there, focusing on winters since 1950. To give the best sample size possible, I initially started with all cool ENSO years (everything from cool neutral to strong La Nina). Here is what the last 20 cool ENSO winters looked like from the December – March period:

Please note that through this exercise, I use the most representative average on the US Climate Division plots, as using the 91-20 averages will make everything with older years look colder than they should.

It’s overall what you’d expect in a La Nina, with a predominantly -PNA and cold focused over western Canada and the NW US. The inclusion of some cool neutral years probably helps skew the AO and NAO somewhat more negative than if I didn’t include them, as 5/7 cool neutral ENSO winters have a predominantly -AO in the December-March period, and the two that average positive both had one month that registered -1 or lower.

Some recent La Nina winters have gone huge…95-96 and 10-11 are the best examples, though even a year like 17-18 had some very legitimate cold and snow even if it also torched at times. Others, like 11-12 and 16-17 (even though the latter was still snowy in parts of the interior Northeast) were blow torches. La Nina is a nice starting point, but what “flavor” can we expect this year?

We’ll first start by looking at cool ENSO / La Nina winters by strength:

Cool neutral looks least “Nina like” with the North Pacific ridge edging farther east, a solidly -NAO, and unsurprisingly given the weak ENSO, a somewhat less defined temperature pattern (and is notably cooler in the Four Corners). Weak to moderate La Ninas begin to take on the most classic pattern, with the mean north Pac ridging shifting south / west, though it’s not totally disastrous until you get to the strong La Nina mean pattern. Note that only the strong La Nina composites have a blatantly +AO and NAO pattern, and unsurprisingly they’re by far the warmest, even with a winter like 2010-11 somehow happening with a strong La Nina.

Some stats to ponder, that will be referenced more throughout. Again, it’s still a small-ish sample for each individual ENSO intensity, so debatable how many of these are statistically significantly, but this isn’t a paper it’s a blog post so let’s not dwell:

Out of all cool ENSO winters since 1950, 15/37 (41%) had an overall -NAO for the December – March period, with 5 more (14%) averaging positive but still having 1 or 2 months with notable blocking. 22/37 (59%) had an overall -AO for the period, with 4 more (11%) averaging positive but still having 1 or 2 months with notable blocking.

“Notable blocking” is defined as one month with a -NAO or -AO (whichever one I’m referencing at the time) with a value of -1 or lower, or two months of -0.5 or lower, even if the overall December-March period averages positive.

The difference between La Nina winters that have a -AO vs a +AO is significant. Here are -AO La Nina winters in the CONUS (leaving out cool neutral years) vs the +AO La Ninas:

Of these 37 cool ENSO events, 32 occurred when reliable records for sudden stratospheric warming events (SSW) have been kept, and 16 of those 32 (50%) had a SSW event. Here is what any La Nina winter (this excludes cool neutral winters) with a SSW looked like, vs any La Nina winter that didn’t have a SSW:

Interestingly, there is little difference from the Upper Midwest into New England. However, the mean for La Ninas with a SSW is cooler across the southern U.S. across the board. This likely indicates that cold air punches well south more easily in a La Nina that has a SSW event vs one that doesn’t.

Some more stats:

Cool neutral ENSO winters: 3/7 had a -NAO overall with 1 more having notable blocking. 5/7 had a -AO overall, and the other two had notable blocking. 2 out of 6 possible winters (one being too far back in time for reliable SSW data) had a SSW event.

Weak La Ninas: 6/15 had a -NAO overall, with 4 more having notable blocking at some point. 9/16 had an overall -AO, with 1 more averaging positive for the D-M period but having notable blocking at some point. 6/13 possible had a SSW event.

Moderate La Ninas: 4/6 had an average -NAO, and 1 more had notable blocking. 5/6 had a -AO overall. 4/6 had a SSW.

Strong La Ninas: 2/9 had an overall -NAO, and 3/9 had a -AO, with 1 more having a notable negative month. Surprisingly, 4/7 possible had a SSW.

Overall, 8/16 cool ENSO winters that had a SSW went on to have an overall -NAO, with one more still having a month or two of notable negative NAO. 10/16 had a -AO, with two more still having a notable negative month or two. Overall, the SSW cool ENSO winters as a whole aren’t necessarily more prone to having a -AO than other cool ENSO winters but it does seem to at least somewhat help -NAO prospects.

While La Nina does tend to promote a stronger polar vortex, and hence can help lead to a +NAO, AO, and EPO (quite a warm pattern, just ask January and February of 2020 or much of the 2011-12 season), clearly cool neutral events, along with weak and moderate La Ninas, can be workable with regards to either an overall -NAO or -AO winter, or a winter that averages positive but still features notable blocking episodes.

La Nina Location

Beyond the intensity of La Nina, the location of where the cooling is focused is quite important. This is similar to the whole “El Nino Modoki” thing, except with La Nina, and can make a big difference. The idea of focusing not only on the strength of the La Nina, but also where the cooling is focused, has gained quite a bit of traction in recent years, and this is a commonly referenced paper:

https://link.springer.com/article/10.1007/s00382-014-2155-z

The following figure from that paper shows the differences they found between eastern Pacific and central Pacific-based La Nina events:

The EP events (top) have a more neutral AO (whereas the CP events have a very positive AO), along with slightly more ridging into Alaska and a somewhat negative NAO. This results in a much weaker (though, still present) Southeast Ridge compared to the CP events. Overall, the pattern shown for EP La Ninas is not nearly as warm (and futile for snow) in the central and eastern U.S. as CP La Nina events.

I binned the aforementioned 37 “cool” ENSO years into central Pacific, eastern Pacific, and “hybrid” (characteristics of both) based events. Note, a few of the cool neutral events were too undefined and left out. It roughly matches the paper…eastern Pacific-based events have a weak mean -AO and -NAO, with more ridging into Alaska, resulting in the coldest temperature composite across the CONUS. Central Pacific events are least hospitable for cold / snow in the central and eastern U.S. overall, with a +AO, +NAO, and +EPO mean, while the hybrid events are decidedly La Nina-like in the Pacific, with a -PNA, but have enough of a -NAO that they’re somewhat redeemable. In general, central Pacific-based events tend to be stronger than eastern Pacific-based ones.

Here are the VP anomalies sorted by “location” or “type” of the La Nina…the standing wave is focused near 120-130E in all of them, though CP La Ninas have the strongest sinking anomaly near the Dateline. Dateline convection is good for cold in the central and eastern US, whereas a lack of it isn’t. That seems like the most striking difference to me.

Out of the 12 CP events ID’d, 4/12 had an overall -NAO, though 6/12 managed an overall -AO, with one more having notable blocking at some point.

Out of the 9 hybrid events, 5 had a -NAO and one more had notable -NAO months at some point…interestingly, 5/9 had an overall -AO, but no other years have a +AO but notable blocking in 1-2 months (this is one of the few where there’s more -NAO than -AO).

Out of the 13 EP events ID’d, 5 had an overall -NAO, with 4 more having notable -NAO months worked in. 8/13 had an overall -AO, and 3 more had notable -AO months worked in.

Overall, the CP events seem somewhat less likely to have a -NAO and -AO than overall cool ENSO winters, though some winters have done it. Hybrid events seem a bit more prone to having a -NAO winter, while EP events don’t hurt -NAO chances and really help -AO chances.  

This Year’s La Nina

So far, the surface cooling has been gradual and modest, with Nino 3, 3.4 and 4 near a -0.5C anomaly…Nino 1+2 has been struggling to dip negative, though has cooled a bit recently. The sub-surface is more impressive, however:

There’s a fairly large area of greater than -2C anomalies beneath the central Pacific. So far, it’s been a struggle for this to upwell efficiently, especially the farther west you go, but the sub-surface anomalies are notable…greater than they ever got with last winter’s borderline moderate event:

The SOI averaged +16 in July, +4 in August, and +9 in September, likely coming in near 10 in October based on my very informal eyeballing of forecast MSLP patterns across the Pacific. As long as this keeps up, the surface should continue to at least gradually cool through the fall as the chilly sub-surface water upwells.

With the surface still consistently running warmer than 2020, but with a cooler subsurface and continued positive SOI, it seems like a solid weak La Nina is favored with moderate still possible if we see more surface cooling soon.

The CFS continues to advertise a borderline strong La Nina and has consistently been on the strong side of guidance with this event. SSTs over the last couple of months have not been cooling as quickly as the CFS has been insisting on. With a continued +SOI and enhanced trades, to go along with the chilly subsurface, cooling of the surface through the fall is quite likely, but the rapid cooling into October isn’t quite happening so the CFS still seems too strong with its forecast.

The ECM ENSO plumes are on the other side, with the mean being a borderline weak La Nina that weakens into the spring. The CFS also has the weakening into the spring. However, the ECM is already running slightly too warm with the forecast heading into the beginning of October (we’re right around -0.5C right now), and in my opinion doesn’t have enough cooling through the fall. A few members still touch moderate.

The September IRI ENSO plumes seem like a decent middle ground, as the dynamical average (red line) has a tri-monthly peak of around -0.8C. I feel like some strengthening is likely through the fall before weakening into spring, and that a solid weak La Nina is likely with some potential to reach moderate.

From the same aforementioned paper, here is the overall profile / evolution of east based, central based, and “hybrid” La Nina events (respectively):

In the eastern-based events, the cooling is initially centered off the South America coast and gradually propagates westward, whereas in the central Pacific-based events the cooling is centered west of 150W. Hybrid events initially have a more east-based cooling but tend to be stronger and eventually cool more into the central Pacific.

This time-longitude plot of the SST anomalies, courtesy of the CPC, shows that cooling was initially east based late spring into early summer, and is still firmly strongest east of 150W. However, Nino regions 1-2 have been somewhat warm until very recently, and there has been a bit more cooling near and west of the Dateline than is typical this early in an east based event. This event is NOT central-based, but it may be something between an east-based and hybrid event as 2020-21’s La Nina was.

First cut of the cool ENSO winters…winters that saw a weak to moderate La Nina that was either east-based or hybrid:

Certainly not a bad look at all, with ridging in the North Pacific amped enough to force the PV a little farther southeast over North America, which is quite a bit cooler for the central and eastern CONUS. There’s also a somewhat -AO and -NAO in the mean pattern. Somewhat typical La Nina precip pattern, wet in the Pacific NW and northern Rockies, along with the Ohio Valley, though somewhat drier into places like New England. As is often the case most of the south is fairly dry overall in these winters, though places like NM and western TX aren’t quite as dry.

7/14 of these winters had an overall negative December-March NAO, and 5 more were positive overall but had a month or two of notable negative. 10/14 had a -AO overall, with one more having some notable blocking worked in. Overall, a greater share of both -NAO and -AO winters than all cool ENSOs combined. 7/14 had a SSW event.

Second Year La Nina

Before moving on to some QBO stuff, just wanted to touch on 2^nd year Ninas, as that’s often a point of emphasis (first vs second year ENSO events) when searching for analogs. Overall, 2^nd year cool ENSO events (of all intensity and location) are ugly:

That’s a +EPO, -PNA, and at-best neutral NAO and AO. It’s a mild look for the eastern US with a Southeast Ridge. Taking 2^nd year cool ENSO events and limiting it to weak-moderate, east-based and mixed La Ninas, it improves a bit:

The biggest differences are the EPO being noticeably less positive (and perhaps a bit negative overall), and lower heights over southeastern Canada which would keep the Midwest, Great Lakes and New England colder. Differences at the surface, all 2^nd year cool ENSO events on the left, the better ENSO intensity and location matches on the right:

Overall, it’s encouraging that the “better” 2^nd year ENSO matches are colder than all 2^nd year events. Not many of these winters are great QBO matches (which I’ll get to next) and it’s not a huge sample, so even though 2^nd year La Ninas kind of get a bad rap (and you can see why) I don’t think it’s any kind of overwhelming factor here.

Quasi Biennial Oscillation (QBO)

We’re going to have a -QBO at 30mb and 50mb this winter. The September value at 30mb was -16, and there are stronger easterlies above 30mb that are still downwelling. The 30mb QBO will be quite negative through this winter, potentially starting to rise into spring. We will start first with cool ENSOs that have a negative / easterly QBO at 30mb:

The main differences to the overall cool ENSO winter composites is a more -NAO and lower heights over much of North America, which mutes the Southeast Ridge somewhat in the means.

9/18 of these east QBO 30mb winters had an overall -NAO, with 3 more having notable blocking for a month or two. 13/18 had an overall -AO. 9/14 possible had a SSW event. Overall, having a -QBO at 30mb in a cool ENSO winter seems to increase the potential for a -NAO, -AO, and a SSW event at least somewhat compared to normal in a cool ENSO winter, which isn’t unsurprising.

When you take just the weak-moderate, east-based or mixed La Ninas with an easterly QBO at 30mb, you get this:

The big change is a somewhat more -AO and much more -NAO, which results in lower heights across much of the northern and eastern CONUS compared to the full set and is much cooler overall over the eastern US. It’s quite a good look and includes some La Nina winters that went quite “big” in the snow department.

4/6 had a -NAO overall, and the other two had a month or two of notable negative values. 5/6 had a -AO overall. 4/6 had a SSW event at some point.

Closer Examination of this Year’s QBO

If you recall back to the current time-height chart of the QBO, the structure of this year’s QBO is easterlies descending through the stratosphere, with westerlies hanging on below 50mb into the winter. We actually have a decently large set of years to use for this exercise. We have 11 cool ENSO winters with a similar QBO structure / evolution, and 9 with a rather opposite structure / evolution. We will explore what this does now…here are the years with a similar QBO:

Along with a somewhat -AO and decently -NAO, the EPO is decidedly more negative than all cool ENSO events combined, and even more negative than just easterly 30mb QBO winters with better ENSO matches. The PNA is closer to neutral too (though I’m not sure I buy that given how negative the PDO looks heading into the winter). Overall, 6/10 of these winters had a -NAO overall, with one more with notable blocking for a month or two. 8/10 had a -AO overall, and 7 had a SSW event.

Here are the 9 cool ENSO winters with a relatively opposite QBO structure to this winter:

Definitely a much more -PNA, +EPO and +NAO, giving the cold west, warm east look that pops up and ruins some La Nina winters. Excitedly and probably not surprisingly, this difference holds up when looking at weak-moderate, east-based and hybrid La Nina events:

Here’s a look at how similar ENSO (weak-moderate, east based or mixed La Nina) winters act with similar QBO evolution and opposite QBO evolution. The sample size does start getting rather small for the oppose QBO winters, but overall, you can see the large difference in the EPO, PNA, and even NAO domains and how winters with a similar QBO to this winter tend to be colder overall in the CONUS…and tend to be colder in the east and warmer in the west.

The winters with a similar QBO and ENSO (the set on the right in the above image) all have blocking…4/5 have a -NAO, and the other (17-18) had a notable -NAO for one month. They all had an overall -AO. 4/5 of those winters had a SSW event.

Thus far, the anticipated strength and location of the La Nina, along with the QBO (both just having a -QBO at 30mb, but especially the evolution with the easterlies downwelling through the stratosphere during the winter) all point to increased potential for a -AO and -NAO compared the typical cool ENSO winter, along with optimism that the EPO can be neutral to perhaps somewhat negative this winter…most signs, so far, pointing to this being the type of La Nina winter that doesn’t have the stereotypical cold west / warm east look with a +NAO and Southeast Ridge.

Solar Cycle

We are coming out of a deep solar minimum between solar cycles 24 and 25. Many papers (such as this one: https://agupubs.onlinelibrary.wiley.com/doi/full/10.1029/2005JD006283) have attempted to establish a link between the solar cycle and northern hemisphere winter stratospheric / blocking patterns. In general, years in a solar minimum tend to have a weaker stratospheric polar vortex and are more prone to blocking, with the opposite true in solar maxima winters. There has been some theorizing that the signal lags a bit, i.e. the greatest potential for blocking is during and just after a solar minimum (and vice versa for maxes). Regardless, sunspots are beginning to increase as we head into cycle 25; barring a significant increase this fall this shouldn’t be a negative yet.

Out of the 37 cool ENSO winters examined since 1950, 11 occurred during or within a year or two of a solar minima. Here is what their overall 500mb and temperatures did when blended together:

It’s somewhat reassuring that there are several winters in here that don’t make the previous QBO or “better ENSO” sets, and we still get a signal for somewhat higher heights into AK, a better -NAO signal, and a cooler look overall in the CONUS compared to the “base” for cool ENSO winters. With small sample sizes and the same years showing up in certain sets, it’s tough to filter out what’s legit or not, but there are a few strong La Ninas in here, several winters that are poor QBO matches, and a few central Pacific La Ninas, so perhaps there is at least something there to the idea of being in or just past a solar minimum enhancing blocking potential in a cool ENSO winter.

Out of these 11 winters, 5 had an overall -NAO, though 3 more also had notable negative values for a month or two as well. 7/11 had an overall -AO, and one more had a strongly negative month mixed in. 5/10 possible had a SSW. Small-ish sample, but it does seem that overall this portion of a solar cycle is at least somewhat more conducive to a -AO (and somewhat a -NAO) compared to all cool ENSO winters.

Like everything else, focusing on weak-moderate, east-based or mixed La Ninas makes the Pacific a bit more favorable and doesn’t hurt the -NAO signal overall, which makes things a bit cooler over the central and eastern US:

Interestingly only 3/6 have an overall -NAO for December – March, though the 3 that averaged positive (85-86, 96-97, 17-18) still have a month or two of notable negatives. 5/6 had an overall -AO, and the one that didn’t (96-97) still clocked a -1.7 for December. Despite all having an overall -AO/NAO or at least some notable blocking at some point even if it was positive overall, “only” 3/6 had a SSW event. Small sample, but impressive that each winter in a similar point in the solar cycle with a reasonably close ENSO to this year had some notable -AO/NAO at some point (if not simply negative overall for the whole winter).

Indian Ocean and Western Pacific Warm Pool

Moving back towards SSTs and tropical forcing, we have a somewhat negative Indian Ocean Dipole this fall along with a very robust western Pacific warm pool. Out of the 37 cool ENSO years since 1950, only 11 have a somewhat similar or similar combination, likely because the western Pacific warm pool has been growing more robust in general with time. All of these years are 1983-84 or beyond. First, here’s the entire set, regardless of how flimsy the ENSO match is to this year:

At 500mb and the surface, it’s fairly similar to the last 20 cool ENSO winters overall, though is somewhat cooler overall, especially over the eastern U.S. The EPO is a bit more negative in these winters, as is the AO, compared to all recent cool ENSOs overall. Robust forcing moving through the western Pacific, especially in the late fall and early winter, can help pump ridging into AK and also has been known to lead to disruptions to the stratospheric polar vortex. However, in mid-winter, forcing in those areas does (at least at t=0) tend to correspond to warmer conditions over the eastern CONUS, which may be why the set isn’t especially cold overall in the south and east, despite some encouraging sings for blocking.

Of the 11, only 4 have an overall -NAO for December-March, though 2 more have some -NAO for a month or two as well. 6/11 have an overall -AO, with 3 more having notable -AO for a month or two. Notably, 8/11 have a SSW event.

It’s worth noting that only 4 of these winters are a decent ENSO match..the others are either strong La Ninas, cool neutral, and / or central Pacific-based events. To maintain a somewhat decent sample size while focusing the ENSO a bit more, I’ll go with all weak-moderate La Ninas with a negative-leaning IOD and relatively warm West Pacific. When letting 3 weak central-based Ninas in, that leaves us with 7 winters that are relatively decent matches from the Indian Ocean east across the Pacific. They look like this:

3/7 have an overall -NAO, 2 more have at least a month or two of notable negative values, and 1 more (08-09) was weakly positive but averaged quite close to neutral all winter. 5/7 have an overall -AO, 1 more had a strong negative month at least, and the last (08-09) was again very close to neutral but slightly positive. 5/7 have a SSW event.

It’s interesting that the signal for a weak to moderate La Nina with a warmer western Pacific and negative-leaning IOD, like many other analog sets for this year seem to point to, also contributes to a decent likelihood for a -AO and SSW event and doesn’t hurt -NAO chances. It also has a look that’s come up a few times, even with some different years thrown in, of cold focused over the northern Plains / Upper Midwest but still working all the way to the East Coast, with a somewhat muted SE Ridge.

This Year’s Tropical Forcing

Looking at a similar thing from the same way…over the last two months, tropical forcing has been most prominent over the eastern Indian Ocean / western Pacific, and somewhat over the Atlantic towards South America:

And the multi-model consensus favors that general pattern of standing waves to continue through the fall. Based on current SSTs, it’s not unreasonable:

Looking for years with similar feature (in particular, the standing wave near 120E), there are several cool ENSO winters that have similar tropical forcing in the fall. Again, sort of making a guess / assumption here, but I don’t think it’s unreasonable:

Strong forcing in the eastern Indian Ocean / western Pacific isn’t a short term “cold” teleconnection for the central and eastern US during the heart of meteorological winter (in fact it’s warm), but as discussed above can eventually encourage blocking and perhaps SSW events. So, the overall pattern in winters following falls that had similar tropical forcing to what’s expected this fall makes sense. Overall it’s fairly blocky with a -AO, -NAO and -EPO on the mean, but the PNA is negative too. The coldest air is focused over the northern Rockies and Plains, but in some of these winters it still consistently powered through all the way to the East Coast.

The Analogs

When factoring all of the above factors…ENSO, QBO, solar cycle, Indian Ocean / western Pacific waters, expected tropical forcing…here are my current analogs, with some weighting:

The October pattern has some similarity to what’s modeled for this year, with a +EPO and ridging over eastern Canada, however, it may not be negative enough with the PNA:

From there, the mean pattern in the analogs begins developing blocking in November, which peaks in December with a trough centered near the Great Lakes. The NAO is negative in November and December, positive in January and more neutral in February. While the AO remains negative on the mean in January and February, the EPO gradually rises and the PNA falls, which allows the southern and eastern US to warm up. In March, a -NAO and -EPO return with a -PNA persisting.

Based on the factors examined and factoring in some uncertainty, I think the mean analog pattern is a very decent starting point for this winter.

Among the analogs, I want to briefly examine which analogs finished with a more negative / positive AO, how they played out, and how they got there:

In terms of how the winters play out, it’s interesting. The two +AO winters in the set were blowtorches with little snow in the central and eastern U.S. Between the strong -AO winters and the more variable AO winters, the NAO is much more negative in the strongly -AO winters. The cold isn’t as intense in the northern Plains / Upper Midwest, but it’s much cooler in the Southeast. The more variable AO winters are colder in the northern Plains / Upper Midwest into New England but have a stronger Southeast Ridge…it’s much more of a gradient pattern.

5/7 of the stronger -AO analogs had an SSW event, 3/6 possible of the variable AO analogs had an SSW event, and though 88-89 had an SSW it was in late February…11-12 didn’t have one.

In terms of what to watch for to try to determine which way this winter will break, we’ll start with October – December SSTs, something that will be fairly easy to track over the next couple of months. Sort of ignore ENSO strength, as one of the two winters that went +AO was a strong La Nina and skews it. But outside of that, the -AO winters are notably warmer in the western Pacific than the other two sets. The +AO winters are notably warmer farther west in the Indian Ocean and have a -AMO look with a cool north Atlantic, whereas the other two sets have a more neutral to positive AMO look with a warmer north Atlantic (especially the -AO set). The -AO set has warmer waters nosing all the way towards the west coast at 30-40N, while the middle set is more mixed there and the +AO years have colder water working down the west coast in the fall. Here are current SSTs, and the multi-model consensus for this October-December:

The warm West Pac / eastern Indian Ocean, +AMO look / warm north Atlantic, and eastern extent of the stripe of warm waters in the Pacific all look more like either the -AO or more variable set as opposed to the +AO set…though, let’s see if this holds up through the fall.

Differences in the pattern leading up to these winters became evident in the fall. Looking at the October pattern, years with a -AO had blocking near Scandinavia and the Urals, whereas the other two sets didn’t. The +AO set has a +EPO/+PNA combination in October that neither other set has. The variable set has a flat ridge south of Alaska and the Aleutians, which neither other set has (though you can argue the -AO set has it but it’s displaced west). Over N America, the -AO set has a western US ridge and ridge over SE Canada. The +AO set has a similar look too, but with much more troughing over the eastern US. The variable set looks most similar to the modeled October pattern for this year in North America in my opinion:

The CFS and ECM Weeklies certainly can be quite wrong, but here are their last bids for the overall October 500mb pattern. They’re both remarkably similar overall. I see a mix of things that match the Octobers proceeding -AO analogs and those proceeding more variable AO analogs, though little that matches the +AO analogs.

Of note, the flat ridge south of the Aleutians and negative anomaly near the West Coast (a -PNA) seems closest too the -AO analogs, though isn’t quite perfect. The Ural / Scandinavian ridge is also closest to the -AO years, as is the modest positive height anomalies near Greenland (the other two sets have negative anomalies). The North American pattern looks most like the variable AO set. The subtle trough extending towards Japan on both models is there in both the variable and +AO set.

Overall, the modeled October pattern would seem to argue against going the way of the +AO analogs, and probably has  more pointing towards the -AO analogs outside of North America. But, it looks much more like the more variable / closer to neutral AO analogs over North America.

If there’s any doubt still after October, the analogs that went +AO really started to stand out in November compared to the rest:

Seasonal Model Predictions

Looking into some current model forecast guidance, as has been mentioned by many over the last couple of months, the August and September forecasts from two models that have recently done ok in terms of forecasting winter stratospheric polar vortex strength ahead of time, the UKMET and European seasons, are forecasting a relatively weak stratospheric PV in their most recent (September) runs, especially early in the winter:

Again, these forecasts from the models can certainly be wrong this far out, but they’ve shown some success in recent winters. At the least, it’s not a bad thing that models are trying to hint at a weaker stratospheric polar vortex to start this winter as several analogs corroborate that idea. The September multi-system forecast (Copernicus) mean SST anomalies featured a weak, basin-wide / hybrid / mixed La Nina for the winter:

I think this is overall reasonable, though again can’t rule out a run at moderate. Regardless, it seems to be in a decent spot ENSO wise to look at and compare to some of the analog-based thoughts. Here’s a look at the September runs of the previous-discussed European and UKMET seasonal models, along with the JMA:

The Euro is almost a dead ringer to the “variable” or closer to neutral AO analogs, with a definite -EPO and negative leaning AO overall, but a +NAO allowing for the Southeast ridge to flex more. The UKMET is a little farther west with the ridging over the Pacific and starts sagging the TPV into more dangerous territory over western Canada. It’s too neutral with the NAO for a strong SE ridge not to crop up. The JMA looks more like the -AO years, with a bit more of a negative near Japan and also decent ridging near the North Pole, extending towards Greenland. It has the trough the farthest southeast over North America overall.

The multi-system mean, including many others such as the CFS, Canadian, and French seasonal models, looks a lot like the analogs that had a more variable / closer to neutral overall AO.

The 2m temperature forecasts suggest more trouble for the cooler air to push into the CONUS than the analogs imply…perhaps there’s a gradual trend towards that due to Hadley Cell expansion, but likely not this quick. Based on what similar 500mb patterns have done in past winters, outside of the UKMET which is closer to being more dangerous I think the 500mb patterns on the ECM, JMA, and overall multi-system mean are encouraging for a cold winter in the northern U.S. with some ridging over the Southwest and Southeast, with a gradient pattern in between. The UK depiction looks similar to weak-moderate La Ninas that have a more opposite QBO downwelling to this winter, which is interesting.

For what it’s worth, the CFS also looks a lot like the analogs that had a more variable AO:

The 2m colder temperature anomalies, much like some of the other models, also struggle to push south into the CONUS compared to the analogs on the CFS.

Overall, the variable AO look among the analogs is the one I’m currently keying in on, though enough supports a more -NAO too that the analogs that featured a -NAO and were much cooler into the Gulf Coast, Southeast and Mid Atlantic aren’t off the table yet. It currently seems unlikely that we’re moving towards the +AO analogs.

Final Thoughts on Winter 2021-22, for Now

Things to watch for during the fall that could make this winter trend towards a more +AO, NAO and perhaps EPO would be a sharp increase in sunspot activity, a stronger or more central Pacific-based La Nina, more forcing over the Indian Ocean, or a stronger than normal stratospheric polar vortex in the fall that couples to the troposphere.

Things to watch for that could make this winter trend towards more blocking would be the La Nina staying weaker and not being centered too far west, along with a decent amount of western Pacific forcing and less farther west in the Indian Ocean. Sunspots remaining fairly low would be beneficial. Ural / Scandinavian blocking in the fall seems to proceed analogs that had a stronger -AO. The stratospheric polar vortex remaining weaker through the fall and not getting too coupled to the tropospheric PV would seemingly aid the odds of following the strong -AO analogs.

At some point, probably mid to late November, I hope to revisit and really hone in. At that point, I may give monthly temp maps a stab! Let’s see how things evolve through the fall.

The pattern coming up is going to feature plenty of cold / Arctic air over North America to tap, along with a continued tendency for blocking over the Atlantic. 

A Rossby wave making machine will be in place over the western Pacific, encouraging persistent north Pacific blocking and a cross polar flow into Canada. This is occurring while a tendency for high latitude blocking continues across the board for multiple reasons.

This loop of the GFS helps show the continued cross polar flow into Canada, and how the retrograding -NAO helps shove the tropospheric polar vortex towards the CONUS this weekend and next week, bringing a cold snap to much of the Lower 48. With the Pacific blocking likely continuing for the forseeable future, the model shows Arctic air reloading towards the end of the run in mid-late February. 

One of the reasons we will remain “blocky” for the forseeable future is the continued downwelling of the weakened stratospheric polar vortex into the troposphere, which encourages a -AO and blocking:

On top of that, there’s been a recent uptick in tropical forcing across the western hemisphere, and western Pacific forcing will continue until further notice:

This deposits momentum in the tropics and subtropics (enhances the sub-tropical jet). Because momentum is a conserved quantity, the increase in momentum in the tropics and subtropics results in a decrease in momentum in the higher-latitudes, which causes a tendency for blocking. The 12z GFS is shown as an example of this, note how the stronger sub-tropical jets over the eastern Pacific and eastern Atlantic occur beneath weaker higher-latitude flow, and how this encourages high latitude blocking:

The persistent western Pacific forcing will keep our Pacific blocking in place for most of February, and the Arctic air dropping into North America will encourage continued cyclogenesis near the east coast, which will help encourage Atlantic blocking as well. 

Basically, the US will lean cold for much of February. There will be a baroclinic zone across the southern / eastern US, though the cold will seep south and east at times given the -NAO and the quality of the cold that will be available. The Arctic air may lead to near to below average precipitation for a lot of the CONUS (save for perhaps the southeast/east coast), but the pattern will stay at least somewhat active given the baroclinic zone and hints of a subtropical jet. The GEFS and EPS weekly forecasts for the next 30 days speak for themselves: 

While suppression and lack of moisture may be a problem at times, we are already in a rather wintry stretch that will probably last through all of February, and perhaps into early March. 

It’s past November’s midpoint, so I’m clearly late to the party with this more technical write-up on what I think will happen this winter. My thoughts to date have been posted by me at times on weather forums and hinted at on Twitter, and my employer has also posted a couple of YouTube videos…but, this is the first time I’ve been able to fully sit down and write everything down. Given how late it is, my thoughts have evolved some from October.

This write-up will have 3 sections of discussion and then finish with the actual forecast. I’ll go through the drivers heading into the winter and what they may mean, some analogs based on said drivers, and also the recent / current pattern, how that may move forward and how that compares to some of the analogs.

Since I understand some people just want the maps, I will post them first, but the fun obviously comes from the write-up 😉

December temperatures:

Please note that I expect a week or two of normal to below normal temperatures in the Great Lakes, Northeast, and perhaps Ohio Valley and Mid Atlantic, along with windows where snow is possible, despite the overall mild temperatures in December.

January temperatures:

February temperatures:

March temperatures:

Total seasonal snowfall (including what’s already fallen, where applicable):

Potential Drivers This Winter:

Plot from: tropicaltidbits.com

We’ll start with the moderately strong La Nina event across the Pacific. This developed over the summer and has strengthened through the fall, and the atmosphere has clearly responded. The La Nina is currently moderately strong and will likely be at moderate to borderline strong strength for the winter.

Plot from: http://mikeventrice.weebly.com/

A recent weakening of the trade winds has allowed for both the surface and sub-surface to warm slightly, however, most  guidance suggests another period of strong trade winds developing the rest of this month and into early December, which likely will result in further cooling.

Plot from: https://www.cpc.ncep.noaa.gov/

Despite a little bit of recent warming, it is still quite cold below the surface across much of the Pacific, particularly east of the Dateline. Another period of enhanced trades would result in upwelling and notable surface cooling into December. Various models depict more strengthening into early-winter and suggest that a strong La Nina tri-monthly peak may still be in play, though the modest warming this month will probably make that difficult to pull off. Most guidance shows the La Nina peaking in intensity in December and then beginning to weaken:

Plot from: https://www.ecmwf.int/en/forecasts

Plot from: https://iri.columbia.edu/our-expertise/climate/forecasts/enso/current/

So, let’s start with the La Nina…here’s what the last 10 have gotten us when averaged together:

Gee thanks for nothing La Nina. Except for 10-11 and 17-18. Those are kind of OK. The last 10 Ninas, on average, fit the stereotype of starting colder than they end for the eastern U.S.:

The last 10 moderate-strong La Ninas only are similar for the winter as a whole, though are a little uglier in the NAO and EPO domains:

There were a couple of fairly good winters in there, and even some of the duds had brief but intense cold shots…though on the whole they’re not great.

In general, La Nina winters feature a stronger stratospheric polar vortex and limited high-latitude blocking, with the signal stronger in moderate to strong events with a +AO, NAO, and EPO on the above composite. There are a few exceptions, even with moderate to strong events, but with all else being equal they are not more supportive of high-latitude blocking than a neutral or warm ENSO.  

With that said, where the coolest water with the La Nina is associated can make a significant difference. This paper examines different response to East Pacific La Nina events and Central Pacific events:

https://link.springer.com/article/10.1007/s00382-014-2155-z

The tl;dr here is that “East Pacific” La Nina events tend to have a slightly more poleward north Pacific ridge, and more notably, a modest -NAO…as compared to “Central Pacific” events that have a flatter north Pacific ridge, and a +NAO.

The question is, is this event central, eastern, or more basin-wide?

From the same paper, above is the typical SST departures with time in East Pacific, Central Pacific, and hybrid La Nina events. The key difference between eastern and central events is that in eastern events, the cooling originates near South America and propagates west, while in central events the cooling is centered near 150W. Central events also tend to be stronger, though there are exceptions both ways.

Here is this year’s evolution. It’s clearly NOT central Pacific based…and the drift west in SST anomalies is typical of east based events. You can argue that for a time in late-summer, it appeared more centrally based, though it’s been hybrid or east-based this fall. Per the referenced paper, La Ninas don’t switch from EP to CP based (or vice versa) from fall to winter, so I don’t think this event is “becoming central Pacific based”…though you can argue there are some hybrid characteristics, which per the paper can lead to some tendencies of both with regards to the pattern response. This does suggest, at first blush, that we don’t have the “worse” flavor of La Nina and may maintain some hope for the event weakening during the second half of winter.

The mechanism for East Based La Ninas producing a more -NAO is not really explored in the paper. A couple of thoughts are that the low-frequency response is slightly closer to “neutral” conditions in the EP events, as the cooling isn’t as far west. Other research (that I’ll reference later) suggests that La Ninas in general don’t have as robust of a blocking response to the MJO as El Nino / neutral ENSOs do. Another thought is that the MJO may be more active into the western Pacific during EP events than during CP events. The paper did present rainfall rate anomalies in CP and EP events:

The CP events (B) have a stronger negative precipitation anomaly that is centered farther west, focused at or just west of the Dateline. The EP events have a weaker anomaly centered slightly farther east, closer to neutral years, perhaps muting the effect La Nina has on dampening the MJO into the western Pacific somewhat and allowing for more blocking to result. In a somewhat concerning trend, most of the seasonal models, that overwhelmingly have a +NAO for December-February and in some cases have a flatter north Pacific ridge, all have precip anomalies for the winter that look more like a CP event:

The JMA:

Euro:

Jamstec:

And the CFS:

This has been consistent on the majority of seasonal model runs for the last few months. This could be a red flag that despite the SST profile clearly not behaving like a CP event, we may still have a heightened risk for a +NAO (and EPO) for December-February as a whole, due in large part to the response to this La Nina. Or, it could mean the seasonal models are a little too +NAO and warm…hmm.

Other SST Areas of Interest:

1 is La Nina

2 is the Indian Ocean. Last year saw a strong positive Indian Ocean Dipole event, with cool waters near Australia and warm waters closer to Africa. This contributed to convection/forcing frequently sitting over the western Indian Ocean during the fall of 2019 and into the first half of winter, with a dearth of forcing over the Pacific, especially in the critical period when the polar vortex began wrapping up in December and early January. This fall, the IOD is neutral to perhaps slightly negative. This argues for the low frequency forcing centering farther east than last winter, but in tandem with the much stronger La Nina doesn’t necessarily promise that the MJO frequently gets far into the Pacific.

3 is the Western Pacific warm pool…as has often been the case in recent years, it’s quite warm and expansive. This, along with the neutral to weakly negative IOD, does argue for convection getting into the western Pacific and perhaps helping to amp the north Pacific ridge due to enhancing Rossby Wave Trains across the Pacific. As referenced above, if the convection doesn’t wander close to the Dateline, this still may not dump the cold air in where the East Coast wants it on a consistent basis.

4 is the North Pacific…it’s quite warm across the board, but this is a negative Pacific Decadal Oscillation. The warm waters near Japan and cooler waters near the southern Alaskan and western Canadian coast are characteristic of a -PDO, and the short-medium range pattern argues for these anomalies worsening. A –PDO tends to be associated with a weakened Aleutian low and a more negative PNA, which causes troughing to be biased towards the Pacific Northwest.

A note that the off-equator SSTs are quite warm over the central and eastern Pacific (a positive Pacific Meridional Mode, or PMM). This is one of the stronger +PMM events observed during autumn during a La Nina. This may add a bit of kick to the sub-tropical jet over the next couple months.

5 is the northern Atlantic. The SST pattern here has been used, especially more so in the past, to predict the winter NAO. The configuration this spring – now has been slightly more conducive to a -NAO than recent years, with cold waters south of Greenland/Iceland and somewhat “warmer” waters closer to Greenland and Iceland. While not as unfavorable to a -NAO as most recent years, it still is displaced north of where it is prior to -NAO winters, and doesn’t really move the needle in the face of other factors if they support a +NAO winter.

The Quasi-Biennial Oscillation:

Plot from: https://acd-ext.gsfc.nasa.gov/Data_services/met/qbo/qbo.html

The QBO is a typically nice downwelling of westerlies and easterlies in the stratosphere. This is shown on the image above. However, the QBO basically “skipped” the easterly (negative) phase in 2016. After one and a half normal oscillations from 2017-2019, we had an extremely anemic easterly QBO late 2019-early 2020, and will be in a westerly (positive) QBO for the winter.

An easterly QBO tends to favor a weaker stratospheric polar vortex, which can lead to more high latitude blocking in the troposphere, while a westerly QBO tends to favor the opposite. Yes, there are easterly/negative QBO years with limited high-latitude blocking, and westerly QBO years with blocking…and yes, sudden stratospheric warming events (SSWs) can still occur in a +QBO and lead to a weaker polar vortex and increased blocking potential. In fact, per this graphic posted by Dr. Amy Butler on Twitter (@DrAHButler), half of westerly QBO La Nina winters featured a sudden stratospheric warming event (filled circles had a SSW):

(1/n) What factors might influence the #polarvortex this winter? QBO and ENSO both can have an effect; however, for westerly QBO and La Nina conditions expected this winter (upper left quadrant), an equal number of years (7) have had SSWs (solid dots) as not (empty dots)…. pic.twitter.com/5NXoxZHS1R

— Dr. Amy H Butler (@DrAHButler) October 5, 2020

This may be a significant wild card this winter, as otherwise, the combination of a La Nina and positive QBO (along with our current, absolutely putrid pattern) favor a strong stratospheric polar vortex, which is less favorable for tropospheric high-latitude blocking. This was a major culprit in the extremely mild winter of 2019-20. I will circle back more to this topic when tying all the drivers into our current / short-term pattern later.

Aside from the typical PV and blocking implications discussed each year, a phenomenon (that I frankly did not notice on my own when I started dabbling in seasonal forecasts) that has gotten increased attention in recent years is the QBO’s apparent influence on the north Pacific ridge in La Nina winters. In general, +QBO winters have a more poleward ridge, while -QBO winters have a flatter ridge (interestingly, EP La Ninas have a more poleward ridge, while CP La Ninas have a flatter ridge).

Here is a look at La Nina winters since 1980, that per a combination of the CPC’s 50mb monthly QBO data the QBO diagram above had a +QBO at or below 50mb through the winter. The most debatable inclusions are 17-18 and 11-12, but both had a positive anomaly at 50mb through at least December per the CPC data and held onto it just below that level through at least February.

Here are the -QBO La Ninas:

And the difference:

The +QBO La Ninas DO have a more positive AO and NAO than the -QBO La Ninas, but the differences in the north Pacific are beneficial for the Northeast U.S. Keep in mind, La Nina strength, location, and other factors are ignored for this exercise and many of these anomalies may not be statistically significant given the somewhat small sample size, but it’s food for thought, at the very least, and has worked more times than not in La Ninas over the last 15 years.

Solar Cycle:

We are coming off of another deep minimum in the 11-year sunspot cycle. Many papers (here’s one I quickly pulled up, though there are more https://agupubs.onlinelibrary.wiley.com/doi/full/10.1029/2005JD006283 ) have found a relationship between the solar cycle and the stratospheric polar vortex strength and tropospheric high-latitude blocking in winter. In general, solar minimums correlate with a weaker SPV and increased blocking, while maximums correlate with the opposite. I’m not going to spend a ton of time on this as I think other factors are already in control, but it was a factor weighed in the analogs.

Low Frequency Forcing:

We’ve clearly had a La Nina response to the low frequency forcing this summer and fall, with rising motion over the Indian Ocean and sinking motion over the Pacific. It can be argued the forcing drifted east some over the course of the summer and fall. Another way to visualize this is on a map showing velocity potential from September – November 15:

This shows a similar idea, though do note that the low-frequency forcing is strongest near Indonesia (near 120E). This signal was focused closer to 40E (much farther west) in fall of 2019.

This does suggest that convection may be most amplified this winter in MJO phases (3-6) that are not cold over the eastern U.S. (phases 5-6 are OK in early-mid December, which I will get to later). However, keep in mind that the MJO getting into the western Pacific can set off a wave train that eventually leads to blocking, especially if it continues propagating. Here are lagged composites for phase 5 of the MJO in DJF, which are quite warm with a strong SE ridge when the phase 5 occurs and in the short-mid range, but turn much blockier a few weeks later:

This low frequency forcing IS a warm signal/risk for the bulk of winter, but likely isn’t as bad as last winter’s in terms of potentially encouraging MJO propagations that can lead to temporary breaks in the milder weather down the line.

The location of this forcing was a factor weighed when determining analogs.

Atmospheric Angular Momentum:

Atmospheric angular momentum is, extremely basically, how much westerly flow there is across the globe. Typically, El Ninos feature a stronger / extended Pacific jet due to increased forcing over the Pacific, and higher AAM. La Ninas typically feature the opposite.

Above is the seasonal correlation to AAM in DJF. The correlation is weak in much of the U.S., but in general higher AAM features a +PNA pattern with higher heights over Canada and lower heights over the southern U.S…basically a typical El Nino pattern. If you inverse the colors (use imagination) and emulate the correlation for a -AAM, it looks much more like a La Nina with higher heights off the West Coast and across the southern U.S. and lower heights over much of Canada.

The AAM over the last year or so has been interesting. It was quite low in the fall of 2019, likely due to increased Indian Ocean forcing and limited Pacific forcing, which contributed to a very ridgey pattern in the mid-latitudes (in particular, across the Pacific) and limited disruption to the stratospheric polar vortex, allowing it to get quite strong by early 2020. It rose later in winter, though was distributed poorly and didn’t shake-up the +AO pattern. The AAM fell again over the summer as La Nina developed and strengthened. It has risen this fall and isn’t dropping back off yet. This atypical AAM for a La Ninia may explain why we’ve had cold weather over western North America and warmth nearly everywhere else in November (and quite possibly December, more later), which is not how La Ninas typically play out (usually, La Ninas start colder than they finish in the eastern US).

The fact that the AAM is kind of high for a La Nina isn’t necessary bad, but how it’s currently distributed is poor for winter weather fans in the eastern U.S. and Europe to say the least:

Plot from: https://atlas.niu.edu/gwo/

Momentum in the tropics and especially sub-tropics is quite weak. The mid-latitudes are rather unremarkable, but momentum (westerly flow) is high north of 60N, indicative of a +AO and lack of blocking. This distribution was seen at times in January-March of 2020 when all attempts to develop blocking failed.

Plot from: stratobserve.com

Part of the reason we couldn’t shake the anomalous westerly momentum in the high latitudes in early 2020 was because the stratospheric polar vortex was very strong and coupled to the troposphere. Nothing was moving that, apparently…despite the increase in AAM and more active MJO during the second half of last winter, and despite a lot of model hints that we would eventually move that. Fortunately, the stratospheric polar vortex, while becoming strong again, is not yet coupled to the troposphere (note how zonal wind anomalies below the white line are still comparatively weaker than higher up). If this doesn’t change in the short-medium range, it may provide for some hope for an eventually blockier pattern if other factors can align.

Analogs:

Now that we’ve more than thoroughly discussed some of the potential drivers that are evident this summer-fall, let’s start diving into the analogs. Factors I weighed while searching for possible analogs included:

1. La Nina, with preference for first year events and moderate to strong hybrid or east-based events
2. The Indian Ocean, with preference given to years with a neutral or weakly negative IOD in the fall
3. The fall PMM, with preference given to years with a higher PMM in the fall
4. The PDO, with preference given to years with a somewhat negative PDO
5. The QBO, with preference given to years with a westerly QBO, and some tolerance for years with descending easterlies, but westerlies below 50mb through the winter
6. The October teleconnections (AO, NOA, PNA, and EPO)
7. The September-October low frequency forcing
8. The general hemispheric pattern and AAM in September and October
9. The speculative November pattern
10. The speculative November MJO when that data is available
11. Years with a large / strong West Pac warm pool
12. The solar cycle (minimum or coming off a min)

There weren’t many analogs that check all of those boxes to an acceptable degree, but if I had to give a few tiers:

Top tier: 1988-89

“Good” tier: 64-65, 99-00, 08-09, 10-11, 17-18

Potentially useful tier: 73-74, 75-76, 95-96, 96-97, 07-08, 11-12, 16-17

Other years evaluated: 54-55, 55-56, 67-68, 70-71, 71-72, 74-75, 83-84, 84-85, 85-86, 98-99, 00-01, 05-06

Here are some composites when using the top 3 tiers with weights:

It fits the general La Nina theme, and looks more like the +QBO La Ninas with more poleward ridging in the north-central Pacific, but also a more +AO overall. December is generally the coldest month, with -EPO and some -NAO potential, with January likely quite up and down but with the AO, EPO, and NAO all trending more positive and the cold backing west. The February look has a robust SE ridge, though the deep vortex near Hudson Bay would probably be deceptively chilly /  snowy from the Great Lakes into New England.

Unfortunately, most of these years were horrible matches to the pattern we’re seeing this November. The few years that made an analog tier higher than “also ran” that had a November pattern that was even somewhat palatable compared to this November were: 64-65, 73-74, 88-89, 99-00, and 11-12. Here was their composite November pattern:

This is not by any stretch perfectly like what this November will look like after the second half of the month, but is somewhat tolerable overall, especially compared to the whole analog package’s November pattern (not shown). Arguably, 1988 and 2011 are the best matches to this year. For reference, here is what this November has done through mid-month:

For some idea of how the second half of the month may change this, the next 10 days on the most recent EPS run as of this writing on Wednesday:

Not perfect by any stretch, but still somewhat OK overall. Here is what those years did in December, and then the rest of the winter:

Overall, this is the shittier end of the larger analog pool. However, without giving the rest away yet (dramatic effect!), I do not think this December’s pattern will end up like the composite pattern in the above Decembers.

With that said, a select couple of those years did end up turning much colder in December and / or January:

64-65 is actually one of the higher scoring analogs…the only thing that really held it back from being a tier higher is a lack of MJO data, distance back in time, and a marginal November pattern match.

88-89 was one of my highest scoring analogs.

Not the greatest analog, but was similar in enough areas to be rated somewhat high.

Most of the winters with similar November patterns to this year were not good winters overall, though more than half featured a legitimately cold period at some point in December or January, and January of 2000 did pull off a pretty good East Coast snowstorm. It is worth noting that 64-65, 88-89, and 99-00 all had at least some similarities to this year WRT the September-October hemispheric pattern/AAM, low frequency forcing, QBO, and in 88-89 and 99-00, some similarities to the November MJO (MJO data not available back to 1964). Worth noting that these analogs have several similarities to this year, are a decent match to this November’s pattern, and have notable cold and/or snow over the eastern U.S. in either December or January. It is possible…

Out of all of the analogs, the only ones that had some similarity to this month’s MJO, the QBO, the low frequency forcing in fall, and the hemispheric pattern / AAM in September-October are 88-89, 95-96, 99-00, 08-09, and 17-18. I will throw 64-65 in there, as despite a lack of MJO data it is a good match in the other areas. However, the November pattern in 1995, 2008, and 2017 was not like this year. WRT the 95 and 08 MJO, they’re not perfect matches, but may have a similar move towards phases 4-6 as some models have over the next two weeks, but a couple of weeks earlier in those years. Given everything, a final blend of analogs includes those 6 years with weighting…an interesting note that 4/6 of these winters featured a SSW.

My final maps don’t look exactly like these composites, mainly because we’re far enough into this fall to speculate about the December pattern and its possible ramifications without analogs, but some takeaways are that December into January may offer the best shot at blocking and colder weather into the eastern U.S. February is the greatest risk to see a strong SE ridge. While March is mild on the mean, it’s somewhat muted compared to February, and the pattern February into March would keep Canada cold, so if the SE ridge isn’t dominant it could allow for an opportunity to turn chilly and try snowing in March in the Midwest and Northeast.

Ongoing – Medium Range Pattern:

As has been the case much of November, the pattern is not conducive for cold / snow in much of the central and eastern U.S. The poor distribution of AAM is evident, with ridging in the mid-latitudes pushing anomalous westerly flow to 50-60N, resulting in a +AO. There is a small block near Alaska, that will dump enough cold into Canada that a couple of systems may produce some thread-the-needle snow from the Midwest/Great Lakes into parts of the Northeast through the weekend after Thanksgiving, though otherwise for now a southeast ridge will keep mild air dominant over the rest of the central and eastern U.S.

Convection is very active across the Indian Ocean right now; a combination of the low-frequency base state, a Kelvin Wave, and the MJO appear to be causing this. The MJO is progressing through phase 2 at a somewhat respectable amplitude, contrary to some prior model predictions. This Twitter thread by @griteater features a nice loop showing the increase in convection and upper-level divergence as the MJO arrived in the Indian Ocean (definitely a good Twitter follow for those interested in mid-long range forecasting, he also posts on some of the common forums such as AmericanWx and 33andrain).

Accordingly, notice how the convection and upper level divergence have 'exploded' there on this running 5-day average loop of OLR (blue) / Velocity Potential (green contour) / Divergence (arrows). Loop captured from the JMA website – https://t.co/qZeTeU95Wz pic.twitter.com/xaHU21p0Rz

— griteater (@griteater) November 19, 2020

This analysis on Dr. Mike Ventrice’s website shows similar, with the MJO and Kelvin wave constructively interfering right now. There may be brief subsidence behind the Kelvin wave, though I think the robust nature of the MJO and low-frequency forcing will offset that and allow the convection to continue propagating east. Yet another plot, this from Carl Schreck, shows the current constructive interference, and the potential for the low-frequency forcing centered near 120E (as discussed earlier) and warm SSTs to keep convection going to at least 120-150E into early-mid December:

Plot from: https://ncics.org/portfolio/monitor/mjo/

In general, the GEFS has consistently been more bullish on getting the MJO into phases 6-7 than the Euro and CFS. Again, given the low-frequency forcing centered near where the MJO enhances convection in phases 5-6 and the warm waters in those areas, I do think convection becomes active between 120-150E (and perhaps a bit farther east) later November and into early December. It is possible that the MJO itself doesn’t make it out into the Pacific, as the low frequency forcing will dampen the MJO’s signal in the short term, and that same LF forcing suggests suppressed convection once farther east into the Pacific. However, convection between 120-150E is not bad through early December, given wavelengths still shorter than mid-winter, and a compelling argument can be made that the MJO is a significant driver at the moment. A few various tools illustrate this I think:

The lagged October-December phase 2 composites on the CPC’s website (each lag is 5 days) shows a similar pattern in the first few pentads to what most ensembles suggest occurring over the next 1-2 weeks, with troughing over the northern Pacific, Alaska and near the West Coast, along with ridging over the central and eastern CONUS. That then gives way to a more +PNA pattern. The anomalies in the later pentads are generally not statistically significantly, so although that’s still not a great pattern it’s not particularly concerning if other factors suggest differently. An interesting note is that the strong low across the northern Pacific that floods much of North America with warmth in the shorter term does not necessarily persist in these composites.

Another tool, from Paul Roundy (http://www.atmos.albany.edu/facstaff/roundy/waves/), shows the typical MJO response at much finer scales than three-month periods. In transition seasons, when the wavelengths typically change quicker than a 3-month mean will capture, these regressions add quite a bit of additional value.

The phase 2 regression for the week of November 17^th shows a number of similarities to the pattern we’re actually seeing…and, moving the regressions forward through phases 3-6 with time shows a fair amount of similarity to the pattern showing up on many of the medium range ensembles in late November, followed by an increase in blocking near the West Coast and in the NAO domain, should we get increased convection to near 120-150E by early December.

The CPC October-December Phase 5 lagged composites show similar as well:

Both Paul Roundy’s tool, and the CPC composites, show a much warmer response to Phase 5-6 with a Southeast ridge during winter, so as the wavelengths get longer, this teleconnects to what people traditionally think it does in winter…warmth for the eastern U.S. But since wavelengths are still shorter, that’s not yet the case, and it can actually be helpful for winter weather fans in the central and eastern U.S.

There should be an increase in AAM in the tropics and sub-tropics in late-November as the MJO propagates east towards the far western Pacific, and as a strong East Asian Mountain Torque event occurs…

There has been a fairly persistent signal among the GEFS and EPS for a block/ridge near and east of the Ural Mountains to dump cold air and high pressure into eastern Europe for a prolonged period of time in late-November into the beginning of December, leading to a prolonged, significant +EAMT event:

This will lead to an acceleration of the east Asian-Pacific jet (the tropical forcing moving east towards 120-150E will also contribute to that). There’s also some minor signal on the EPS for an increase in West Pac tropical cyclone activity around the beginning of December (the GFS and GEM have had a number of false alarms while the EPS has remained much more bearish thus far).

This will add momentum in the sub-tropics, which may eventually improve the distribution of momentum across the mid-high latitudes and lead to a regime more favorable for high-latitude blocking, and also will encourage a change in the Pacific pattern in late-November and early December. With convection most active well west of the Dateline, this strong jet will want to break somewhere over the central or eastern Pacific (the +PMM and active STJ may initially cause this to occur farther east than it otherwise would in a La Nina). This will cause a significant attempt at raising heights over the western U.S. and into Alaska, which may increase the PNA and drop the EPO.

You can see this on Thursday’s 12z EPS, which is more bearish than the GEFS in getting convection into the western Pacific (also note the sub-tropical jet enhancement from the positive PMM east of Hawaii):

The wave breaks want to happen somewhere over the central or eastern Pacific, though initially the TPV over Alaska may be stubborn. The stratospheric PV will also briefly swing towards Alaska around Thanksgiving, which doesn’t help this cause:

So, we likely won’t see an immediate rise rise in heights over Alaska, but if the SPV moves away from Alaska and the TPV backs west…which the EPS and GEFS consistently show both at the moment…the wave breaks would eventually have a much easier time trying to raise heights on the West Coast and into Alaska. Initially in late November and early December, the lack of height rises over Alaska will result in the PNA rising, but the EPO won’t drop immediately. The op models consistently have a very progressive, wavy pattern with sub-tropical support to end November and start December with a +PNA look. The ensembles don’t look very stormy yet in this regard, but their mean pattern by the end of November / early December is too smooth:

(A note that this is a couple runs old now as of this posting, and it has trended a bit sharper with its ridge and trough axis)

The wavelengths in early December aren’t long enough to support a trough from 160E to 100W, and a ridge from the West Coast of the U.S. to eastern Canada. While it’s likely it’s generally troughy and generally ridgey in those areas, the pattern will probably be quite a bit wavier than an extended ensemble mean implies, which will allow for a bit more cold air to be available than modeled, and along with the active sub-tropical jet may bring snow potential when / if there’s enough polar jet influence. Given the STJ involvement, if marginally cold air can get displaced well into the CONUS any snow threats, in a more Nino-like fashion, may not favor the northern tier.

December will start quite mild across most of the CONUS with a progressive, but active pattern with a somewhat enhanced sub-tropical jet. That’s pretty baked in based on with the very +EPO to end November. But it very well may be stormier than the smoothed ensemble mean implies, with at least some transient cooler shots too. And, once the TPV starts sliding west of Alaska, it’s quite possible heights rise along the West Coast and towards Alaska quicker than the smoothed means suggests too. A day 12-13 51-member ensemble mean will not resolve something like a wave break at all.

While I think there’s legitimate potential for at least a temporary -EPO developing in early-mid December, I’m not as confident in Atlantic side blocking following suite. It is possible, as both the Roundy and CPC plots show a -NAO response to phase 5-6 forcing in the late-fall, and there have been hints of it on longer range guidance (the ECM monthlies have consistently had a weak -NAO for December. Let’s point out what, other than the Euro monthly, may point to a negative NAO:

An MJO propagation into the western Pacific (in particular, phases 5-7) in late November or early December, as mentioned, can favor the NAO dropping into early-mid December. In addition, Rocky Mountain Torque will remain weak if not somewhat negative into the mid-range (it shouldn’t increase until we start seeing EPO-induced cold shots, early December at the earliest). This favors a weaker jet from eastern North America into the western Atlantic.

With an active STJ late November / early December and a +PNA, the potential is there for strong cyclogenesis in the central or eastern CONUS, which may help raise heights / perhaps cause a wave break over the northwestern Atlantic. This, along with some (but not consistent) hints on guidance of a dropping NAO into early December, and the tropical forcing, does give multiple ways to get the NAO to drop. A couple of sources for pause:

From this paper: https://journals.ametsoc.org/jcli/article/31/13/5293/92625/The-Impact-of-the-Madden-Julian-Oscillation-on

In general, the MJO does not produce as robust of an increase in blocking during a cold ENSO as it does during a warm ENSO in the winter. Note how phase 7 with a warm ENSO has a strong, statistically significant correlation to a -NAO. There are hints of a lowering in the NAO following a phase 7 with a cold ENSO, but it’s weaker and for the most part not statistically significant. Note than there is a temporary increase in blocking along the West Coast and into Alaska shortly following a phase 7 during a cold ENSO.

This suggests that while a -NAO isn’t impossible if we get the MJO to propagate into the western Pacific over the coming couple of weeks, the cold ENSO may otherwise lower the chances of it delivering the goods compared to otherwise. This does, however, further point to height rises along the West Coast being a possibility.

My other concern:

After the PV gets shunted away from AK and allows for a better opportunity for the wave breaks to raise heights there, it moves towards the eastern North America – northeastern Europe sector. While it’s not strongly coupled to the troposphere right now, this may discourage prolonged blocking in the North Atlantic, or at least make it so everything needs to go just right for it to occur.

My overall thought on the NAO over the next few weeks is that while I think we will see attempts for it to go negative…and at times, the op models will do it in a big way…I’m not confident we pull it off. A legitimate -NAO ups the ante along the East Coast in December, though one may not be needed for something if the Pacific side improves as I think it may. Basically, I lean narrowly towards a +NAO for December, with perhaps a couple of short dips, with a lower potential for a wave break in just the right spot to pull it off. Not impossible, but not really predictable at this stage…perhaps I’m just bitter from watching most -NAO attempts fail the last 7 or 8 winters, with the notable exception of March of 2018.

Thoughts on tropical forcing deeper into December:

There is a divergence among guidance, with the GFS in particular keeping forcing active between 120-150E for much of December, while the EPS and CFS, along with the JMA, generally appear to focus it farther west in the Indian Ocean by mid-December, causing the attempt to raise heights on the West Coast into Alaska to be brief and be quickly replaced by a pattern similar to what we have now with the TPV frequently settling near AK. Even if convection does remain active near Indonesia and spark a change to a pattern we’d more typically expect in a La Nina December, as the wavelengths look more like winter later in the month that does teleconnect to a Southeast Ridge. Although I wish other guidance showed it, the low frequency forcing has frequently flared up around 120E this fall, which may point to the GFS solution having some validity. This would somewhat prolong cold risks, probably focused around the second week of the month, before trending milder late in the month.

The CFS weeklies have consistently appeared to develop a robust, eastward propagating MJO for mid-late December that may get into the western Pacific in early January. While the EPS mean VP anomalies focus mostly over the Indian Ocean on the most recent weeklies, a majority of members have a coherent MJO signal at some point beginning mid-late December. The low frequency forcing amplifies this signal over the Indian Ocean and there’s quite a bit of disagreement on timing at this point, so the ensemble mean looks like a standing wave over the Indian Ocean with no MJO activity, but that likely isn’t realistic. An amplified MJO moving across the Indian Ocean would bring a warmer signal to the eastern U.S. as it occurred, but could lead to a cool-down at some point (likely towards mid January) for the eastern U.S. as it attempts to emerge over the Pacific. This would have ramifications on the January forecast.

Thoughts on the polar vortex into December / beyond:

As referenced above, the polar vortex is quite strong, but not well-coupled with the troposphere yet, and is occasionally getting knocked around and stretched a bit. So, while it is currently a signal for mild weather, it’s not an entirely hopeless proposition for it to weaken at some point as it was last January – March. The pattern showing up somewhat consistently on guidance the rest of November into early December is likely a more favorable pattern for disrupting the polar vortex than what we saw early this month.

The CFS is shown, though the EPS and GEFS have occasionally had the look of an Aleutian low and Scandinavian-Ural high as well. These are both features that stand out on composites of the precursor tropospheric pattern to SSW events:

A is the proceeding pattern in the 45 days prior to a vortex split, C is the pattern in the 45 days prior to a displacement. Both similar, though the Aleutian low and Scandinavian ridge are more amplified proceeding the splits. Another set of maps from the same paper shows similar

Here is a link to the paper: http://web.mit.edu/jlcohen/www/papers/CohenandJones_JC12.pdf

The polar vortex will again be quite strong in the short-medium term, but runs of some longer range guidance over the last week or so has slowly opened the door to some weakening mid-late December…shown are the CFS, GEFS and EPS:

CFS and GEFS PV plots from https://simonleewx.com/

My thought is the PV will weaken at least some in mid-late December, though I don’t think a SSW is likely before the New Year unless we set off more tropospheric blocking in the shorter term. I do think that the ridge across the Urals supporting continued opportunities for +EAMT in December, and the possibility of another more robust MJO between mid-December and mid-January, may give a window for a SSW to occur in January. I think after that, typical Nina forcing would favor the PV strengthening into February if it isn’t significantly weakened by mid-late January. A 2018 outcome where the La Nina weakens late in the winter and the MJO becomes active would then be the only hope for a PV disruption later in the season.

So, I see two clear opportunities to weaken the PV in December into the first half of January. If this successfully occurs, an MJO propagation into the Pacific in January could provide for a temporary colder and blocky pattern to start 2021, with the +PMM leading to some continued STJ activity. After that, the PV may very well get strong and coincide with a return to Indian Ocean forcing to provide for significantly milder weather heading into February. This is followed by an outside shot at a prayer to end the season if we can get one more coherent MJO as the La Nina begins to fade.

Putting it all together…the forecast:

December:

A +PNA to end November and start December leads to a very mild / warm start to the month across the CONUS (slightly cooler over the Southeast / East Coast)…however, an active STJ and a wavier pattern than the ensemble means will show at this distance may allow for enough polar influence for an opportunity or two at a winter storm across the central or eastern U.S. Given the +PNA and active STJ, this may occur in areas such as the Ohio / Tennessee Valley, Appalachians, Southeast or Mid-Atlantic, farther south than what you may expect in a La Nina and more typical of an El Nino. A +PNA, split flow, and active STJ are features that are frequently seen in an El Nino!

This window is brief, the first week or so of the month and with a questionable amount of polar air to work with, so it may not work, but there should be a bit to track on the models as this comes into range. Thereafter, the SPV moving away from Alaska, persistent +EAMT, and convection getting into the 120-150E area (and likely persisting a bit more as the GEFS has due to the lower frequency forcing in that region this fall) likely allows for a window of -EPO, and a pattern more typical of a December La Nina. I think there’s still enough momentum in the Pacific jet that this dumps into the Midwest and Northeast as opposed to the Rockies and Plains as typically occurs with EPO shots.

This trends towards cold edging towards the Northwest / northern Rockies into late December, with a Southeast Ridge cropping back up due to forcing becoming more focused on the Indian Ocean, and due to Indonesian convection beginning to teleconnect to a Southeast Ridge in later December.

With the start of end of December quite possibly warm for a large portion of the country, the month as a whole will be mild across the CONUS.

While I think the NAO likely ends up positive for the month, if we take advantage of the potential to develop a -NAO in early or mid-December, it would up the ante for snow in the Northeast and Mid Atlantic in that window, and may slow the warm-up later in the month. Even without a -NAO, I think there are opportunities for some snow in December in the central and eastern U.S. (even the southwest may have a window with the STJ, the NW and Rockies will be active and trend colder later in the month as well).

January:

January may start cold in the Northwest U.S. and warm across the southern and eastern U.S. However, I think the MJO has another opportunity to propagate east during the first half of the month. I believe the PV will be weaker and more receptive to the MJO producing a blocking response than it is in early December. This may cause another period of -EPO, -NAO, and a somewhat active sub-tropical jet for a couple of weeks starting the first or second week of January. This will bring another window of winter farther southeast towards the Ohio Valley, Mid-Atlantic, and perhaps even the Southeast.

I think we start trending towards a Southeast ridge later in January as the MJO fades and Indian Ocean forcing crops back up. How much blocking develops in the first half of January will of course influence how quickly this warm-up occurs.

February:

This is generally the warmest month in the analogs, with an amped central Pacific ridge dumping cold into Alaska, western Canada, and the northwest / north-central CONUS. If we have persistent Indian Ocean forcing and a strengthening PV to open the month, that will probably be the case again this year. If there’s cold available in Canada, the +NAO may lead to confluence east of New England that leads to high pressure over eastern Canada that can cause cold air to seep down into New England and perhaps the Great Lakes in February, causing some wintry threats in these areas…however, the pattern I envision for February is not snow-friendly in the Ohio Valley, and especially the Southeast and Mid-Atlantic.  

March:

Cold air likely remains available in Canada, so if the MJO becomes active again in late-February or March it may shake-up the pattern enough to bring it south and bring one last shot of winter in March. If this does not occur early enough, the Southeast Ridge dominates much of March and keeps the cold over the Northwest, Plains, and Midwest / northern Great Lakes / northern New England.

Teleconnection Guesses:

AO: Generally positive, but may go neutral or negative briefly in early-mid December, and perhaps a bit more legitimately in January

NAO: Generally positive, may briefly go negative in early-mid December, with greater potential for a couple-few weeks of a variable / negative at times NAO in January

PNA: Negative

EPO: Generally somewhat positive, but will dip at times. Mid December and perhaps mid-late January may offer windows.