LRF winter 2021-2022 _ severe cold at times central, snow and cold eastern US

[Roger Smith]

Using a technique of weighted analogues by natural variability factors, I have determined the following prospects for the winter of 2021-2022 in North America. 

Cold may play a significant role compared to some recent winters. Severe cold is likely to show up fairly often in central regions and will sometimes spread into the northeastern US and mid-Atlantic states, even the southeast a few times. A coastal storm track may develop and snowfall could be generous compared to recent winters in the larger urban areas of the eastern U.S. 

Mid-January to early March is favored for the best winter synoptics in the east but December is not ruled out either. 

Following a sustained warmth this month lasting well into November, would look for a highly variable pattern to develop by late November and some cold outbreaks in December mixed with much milder spells. This could yield some cold spells of a few days' duration and possibly one significant snow event. A milder spell seems likely around Christmas to New Years and into the first ten days of January, followed by a series of colder outbreaks some of which could be quite severe. This will set up a better than average east coast snowfall potential. 

One of several years that showed up in the analogue set was 1898 (into 1899) which featured an epic arctic outbreak in mid-February. But in general, the index values tend to show equal chances of severe cold in late January and early to mid February, so perhaps one good spell in each month. The March index values tend to show a sharp reversal from severe cold at first to record warmth later. 

Anyway, I for one do not expect this winter to be as frustrating an experience as the last few have been for east coast snowfall fans, and the Midwest and Great Lakes could have quite a severe winter. Expecting the inland west to be quite variable with a persistent mild signal for the Pac NW and coastal BC into the western prairies due to frequent chinook activity, as the main track for severe cold will be through eastern SK and MB into the Dakotas and Minnesota. 

[raindancewx]

I'm pretty similar to what you have. I've actually finished my outlook but even though I sent it to my home email from work it hasn't arrived yet. Have to try again Monday if it doesn't arrive later today.

I arrived at a severe cold snap in February in four independent ways in my outlook:

- Objective most similar matches for La Ninas in Albuquerque for high temps and precipitation Jan-Sept (Feb 1939, 1951, 1955, 1972, 2011, 2021 blend)

- Similar solar/MJO/ENSO matches

- Februaries following severe Plains centered cold Februaries. I used Feb 1936, 1978, 1994, 2018 as a proxy blend for last year and then rolled forward a year to see if Feb 1937, 1979, 1995, 2019 looked like my analogs or had a cold February.

- Actual analogs I like overall are pretty cold February.

For me, warmth in November is a key feature ahead of the November cold snaps. Since at least the late 70s, the super-duper cold Februaries in the Plains have tended to feature very Novembers with a +WPO, +AO, +NAO look.

I do think November turns pretty cold late in the month. But should be pretty warm overall. Even last November, warm as it was, started to get cold in the West at the end of the month.

 

[Roger Smith]

Details of course did not match up perfectly, for example, there were alternating mild chinook spells and times where arctic air either preferred to stay inland west, or back-doored in from the expected strong arctic outflows into central Canada (which has had a severe winter, don't think the Midwest would qualify as severe, with a few limited very cold spells). The east coast remarks are closer to being right than wrong so far, waiting to see if March does what I think it will do and flip from cold to very warm after the equinox like 1948 as a prime example. What do you think raindancewx? You said you were leaning similar but I wasn't sure what your final call was, or how you think it may have done? 

For the three winter months, I think the local average temperature will be almost bang on normal, and snowfall here has been below long-term averages (which are quite high, it is not unusual to see 1.5 to 2.5 metres of snow on the ground at times here, the maximum snow depth in town has been 35-40 cm and out on the slopes around one metre possibly). We have had a lot more sunshine than is normal here, and almost no rain mixing in which has made it quite an easy winter compared to some (nothing worse than a heavy rain on top of a heavy snowpack). 

[raindancewx]

Your forecast was pretty good. My research is that the Northeast struggles with widespread snow in high-solar La Ninas (55+ sunspots annualized July-June) and also in lower ACE (<150) hurricane seasons. So I was skeptical of your big snowstorm/snow event idea, even though the frequent/potent Nor'easter idea I thought was pretty good, and it was of course. Lots of pretty powerful systems. I think most of the Northeast is below average still for snow, but I haven't really looked yet, since it's early. I know Boston and Atlantic City at least are above average.

My idea for (DJF) winter was for the Northwest to be cold, with the Southeast quite warm, up to +5F locally. Essentially a classic -PDO pattern. I had a chilly dip down the spine of the east side of the Rockies down to about Pueblo which didn't verify. If you take a raw comparison of my analogs, 1961-62, 1974-75 (x2), 2001-02, 2017-18 (x2), 2020-21, I was within 2F of observed temperatures with that blend for 2/3 or more of the continental US, which I consider an OK forecast on my system.

If you look on the CPC 6-10 and 8-14 periods, some Marches I've liked since Sept-Oct are still showing up. These include 2001-02 and 1961-62 with severe cold. I had a 60-page forecast in a thread I made in October if want to see, in the weather discussion area of this site.

I thought March would be severely cold at times in the Plains and North of the US, with some pretty potent cold shots in the West too, but there were some signals for incredible warmth as well. I do think when Feb-Mar is over, the US will be much snowier than average, like I forecast, in that period. Conceptually I thought something as severe as March 2019 was possible - we'll see though.

[Roger Smith]

Yes, I think that parts of the inland northeast US are still running below their normal winter snowfalls and only around ACY into the Delmarva can we find any significant above normal amounts to date. ACY is about twice its normal. The places doing worst relative to normal seem to be around eastern PA and central NY state. This has not been a particularly strong lake effect winter but the Toronto-Buffalo region has had considerable synoptic snowfall anyway. Back towards Detroit and Chicago the snowfalls are relatively less impressive and end up below average in parts of the Midwest too. There is a heavy snow pack in the border regions and the Canadian prairies. This trends to a normal snowpack in many parts of the Pac NW mountains and BC. Snowmelt has been gradual but almost complete now in southern BC valleys, at my mid-mountain elevation we never had as much snow as most winters and the ground cover now is around 8-10" in town and 12-18" in nearby lower alpine. 

In western Europe, a lot of people (including me) had forecasts that included some severe wintry conditions in January but all that happened was that a blocking high set up on the coast instead of over central Europe. This led to a fairly normal January that was unusually dry. February was about like December, mild and rather wet. Now there are signs of eastern European blocking and severe cold into the conflict zone in Ukraine, deepening next week. But that cold pattern is currently associated with less significant blocking and cold near the British Isles It has been an odd winter for that region, some record high temperatures on two occasions, long stretches of bland weather, and a few very strong storms. There has not been much snow away from higher ground. Last April was record cold in parts of west-central Europe so perhaps that pattern will show up again at some point this spring season. 

I think we come up with similar forecasts simply because our two methods are not that different, I derive index values from all data and my timing is from external sources of energy, yours is to identify similar teleconnections which are probably buried in the cause and effect of my index values. So I have (in the past few winters) always looked at your forecasts and the others posted here to see what I can learn about those methods, especially if there were any other index value weighting parameters that might emerge. My theory is that the bulk of the challenge is buried in the past numbers but only after they are modified by considerations of AGW and grid shifting (which means you could come up with an accurate prediction by taking grid points that were not stationary but shifting across a grid). In your own case, your analogues might have been more in south Texas 50-100 years ago as I am finding that the grid I use is shifting northwest with the magnetic field. The northward component of shift aligns with warming. The westward component may align more with different responses to storm track assumptions. This may be part of the reason why the past never repeats very precisely, because any situation where it might try to do so would involve the same signal over different terrain.