Winter 2023/24 - A Post Mortem

[raz.org.rain]

 Bristawl Si it's a likelihood that we dismiss at our peril. I've discussed it at some length with @CryoraptorA303 and so far there's a very strong possibility. But, as we know, the climate has a habit of going in unexpected directions.

[Catacol]

On 11/02/2024 at 12:13, jules216 said:

you have just said what I echoed too,broascale changes in polar field, SSW etc. How come I didnt need to be expert in GSDM field yet I managed to see in advance a SSW that would re-shuffle the good synoptics. I even mentioned it in models thread way in advance. Very similar f.ucked up SSW happened in January 2007,2015 etc. How come All the tools/people/mets/UKMo hot it wrong big time even though they scrutinize GSDM or incorporate it in their algoritm? Magic

Jules - you are way off. Pattern matching is a mugs game because no 2 patterns are identical, and analogues have become near useless. And the SSW didn’t reshufffle anything because it scarcely occurred in the first place. Go read Amy Butler’s blog. Far from reshuffling as a result of a SSW, the reflective nature of the failed SSW contributed to the collapse of the cold pattern because we didn’t get that SSW in any real sense. Didn’t reverse early January, climbed above climatological average within a few days.

CC is our greatest enemy in terms of landing cold. Claiming seer-like qualities in spotting a warmer than average winter based upon pattern matching is the equivalent of grabbing glory for betting on Bayern to win the Bundesliga because it happens most of the time. No skill whatsoever.

[CryoraptorA303]

So I've been away doing some background research over the last couple of days.

I decided to try and fuse the sciences of climate and cladistics. Using a temperature-precipitation axis, courtesy of Starling's Roost, I was able to group January and February of this year with some "relatives".

January 2024 exists at the base of the "2003oidea" family, with its closest relatives being the more derived January 2003, and the even more derived Januaries 1971 and 1934, which form their own sister clade. I haven't plotted a full Order-scale cladogram of the January 2003oidea's more distant relations yet, however I do have a very rudimental basic graph:

January-2024id-cladogram hosted at ImgBB

IBB.CO

---

Image January-2024id-cladogram hosted in ImgBB

 

February was a lot more interesting to me and I constructed a much more detailed cladogram for February 2024's relatives.

Wetwarmia hosted at ImgBB

IBB.CO

---

Image Wetwarmia hosted in ImgBB

Here we can see that February 2024 is part of what I have dubbed the 1989 superfamily, the 1989oidea. Within this it diverges from the true 1989 lineage and forms its own family with none other than... February 2022. The two Februaries are, according to provisional figures, very close relatives. This offshoot of the 1989 superfamily is the 2022idae. Going slightly more distant, February 1923 is also a member of the 1989 linage, forming its own more basal, monotypic family within the 1989iforma infraorder. Going further out, we can see that the 1989 lineage is a close relative of the 1997 superfamily. This branch, the 1997oids, contains Februaries 1990, 2002, 2014 and 1997. Even further out, we reach the base of the greater 1989 lineage, the 1989imorpha, which is sister to the 2020idae, a monotypic family containing 2020, which is a more distant relative of the whole group and represents a slightly different lineage to the 1989-1997 one. This forms the Wetwarmia order, named as such because all of these Februaries are wet and mild.

1989's closest relatives are 1995 and 2000, with 1995 being the closest.

Out of the entire family, those with either a hot summer or an extremely notable heatwave are 1923, 1989, 1990, 1995, 1997, 2020 and 2022. Perhaps 1923, 1997 and 2020 are the only two that didn't necessarily have a hot summer overall and only the notable heatwave. 2000, 2002 and 2014 are the only three that had neither, although correct me if I'm wrong, all three had a notable dry spell, and at least one warm summer month.

Out of all of these, 2022, 1995, 1989 and 2000 are February 2024's closest relatives overall. 2000 is the only one without either a hot summer or a notable heatwave, and even then it was still a dry and warm one. 1997, 2002 and 2014 are all more distant relatives and more closely related to each other. Perhaps most noteworthy is 2022 is the closest relative of all, and furthermore 2024 is also 2022's closest relative.

Considering as well as this, one of January 2024's closest three relatives is 2003, and to my knowledge 1934 was not a particularly cool summer and 1971 looks okay-ly warm for the decade, the indication here seems to overwhelmingly be a dry, hot summer ahead, with a high risk of extremely scorching temperatures.

These cladistic groupings also have some experimental evidence, as the spring months of 2023 are all closely related to either 1926 or 1959, or both, and August is a close-ish relative of 1906 and 2016, and all of these had notable September heatwaves. May 2023 is also a very close relative of 1940, which was the previous June record.

Too many people are getting caught up with the Nino collapse talk, so I thought I'd take a different angle to the great question of "what will this summer be like?".

[jules216]

 Catacol Well SSW in a sense is these days also a meaning of a substantial event of polar stratospheric warming. The definition of 5 days in a row of easterlies ať 10hPa and split like 2012,2018 of course didnt happen. But what happened was a reflective event similar to January 2007 and based on that I did warn of similar outcome in February. Do you think that I didnt read A.Butler papers and posts? I do follow all in detail. Now I ask you, did you follow 17.1.1987 displacement/reflective event that brought/induced the end of very favorable pattern to date only to bring 6 weeks of no winter fór it to return in March. The best analog matches for second half of winter are 1998,2007 and 1987 in general winter European circullation. You Can disagree but if you disagree and say I am wrong then pull out this February so far vs the blend of those years. Or you want to say that I am just guessing again and just being in luck with the years? What was your forecast anyway? How did your thoughts pan out in detail? 

[Penrith Snow]

 sundog indeed,we can only look back with envy at our parents time when weather charts looked like this.

What a shame Netweather wasn't around to appreciate it.

Andy

[sundog]

 Penrith Snow indeed. Growing up in the 80s for example it was cold spells that got me interested in the weather in the first place. The way things have gone with our winters I feel cheated lol.

[Met4Cast]

 Blessed Weather

On the other hand, this will be the third SSW of the winter that has utterly failed to produce. A lot of hope is placed upon SSW's to produce but the truth is roughly 70% of split events produce colder weather across NW europe, that % is significantly lower for displacement events (as you know, I'm sure!)

SSW's can produce holy grail synoptics, 2018 for example, but that was one of the most extreme synoptic responses to a SSW on record, the reality is usually (and unfortunately) more muted.

[Blessed Weather]

3 hours ago, Met4Cast said:

On the other hand, this will be the third SSW of the winter that has utterly failed to produce. A lot of hope is placed upon SSW's to produce but the truth is roughly 70% of split events produce colder weather across NW europe, that % is significantly lower for displacement events (as you know, I'm sure!)

I wonder if the initial Canadian Warming was a fail tho? It might actually have played a role in giving us the favourable (for cold) NH synoptics in the form of a Greenland High that could well have delivered a notable cold spell. But agree that unfortunately the second SSW that occurred around the 17th January seems to have been the factor that on this occasion messed up the favourable synoptics. Win some, lose some.
Here's an extract from a NOAA blog dated 16th Jan discussing the Canadian event and imminent second event:

The polar vortex is acting up
.....But what makes the stratosphere’s current behavior unexpected and somewhat rare is that the polar vortex seems to be more disrupted at the lower levels, closer to the stratosphere-troposphere boundary. For more than a week, high pressure has been sitting in the troposphere over Greenland. It’s possible that the recent minor stratosphere warming reinforced this Greenland high pressure, which then drove a wedge into the stretched-out polar vortex in the lowest part of the stratosphere, splitting it into two lobes.

Full NOAA blog: https://www.climate.gov/news-features/blogs/polar-vortex/polar-vortex-acting

Whatever, for stratosphere watchers it's been a very interesting winter - three events noteworthy in itself (although strictly speaking we await to see if the third SSW happens or not!).

[stodge]

 Met4Cast Yes, I am firmly of the view Wave 1 warmings from Siberia (far more common than Canadian warmings) have regularly affected British winters for decades if not centuries.

The PV is sent back to Canada/Greenland and we all know what happens then.

This process of PV development, displacement via Wave 1 and eventual split either via an SSW or final warming has, I suspect, been the pattern for hundreds of years.

As to why there are nuances - some winters see much less Wave 1 activity than others, I don't know. 

I don't know if there has been any study into how the polar vortex behaved 100, 500 or 1000 years ago.

As to how it will behave in a warming world, I don't know either. A warmer atmosphere suggests more energy and more volatility so you'd expect more warmings but would it be weaker longer term?

[Weather-history]

Has this been a bad winter for the teleconnections pundits and medium/long models?

 

 

[jules216]

My spring analogs. I Can already detect height rises to NE to be catalyst of ,2 week cold snap in central Europe. Apríl mild and May black sea low and changeable. Low drought risk in Europe. There you go

[rwtwm]

Ok, so the winter has been a let down. That on its own isn't worth much investigation in a warming climate. That said, I'm interested by the NWPs persistent forecasting of HL blocking. Particularly in the last 3/4 weeks. 

Remembering that the NWP models are just the physical interactions between packets of air, there was something leading the suggestion that a mid-atlantic ridge or even a north Atlantic block would set up at day 12. 

For me it's not sufficient to say there is a bias towards blocking. That's the symptom. It's not on every run all the time, so something in the configuration would have led to blocking but for an unmodelled interaction. Or there's something about the initial configuration that allows the bias to take hold. 

I assume much smarter people than I are already looking at it to improve modelling. If anyone is aware of any blogs or papers on the topic I'd be interested in seeing them.

[Met4Cast]

13 hours ago, stodge said:

As to how it will behave in a warming world, I don't know either. A warmer atmosphere suggests more energy and more volatility so you'd expect more warmings but would it be weaker longer term?

Very likely the opposite based on a few different studies. 

The Stratosphere is actually cooling with climate change rather than warming, increasing amounts of CO2 trapped in the atmosphere means less heat radiating out & as such heat gets trapped primarily in the troposphere allowing the stratosphere to cool, a cooler stratosphere = larger temperature gradient between the Arctic and Equator which would lead to a stronger polar vortex rather than a weaker one. 

You’re right though, more energy in the troposphere would suggest an increased likelihood of stronger “attacks” up into the strat so it’s quite a difficult one. 

[Earthshine]

 Met4Cast This is another huge nail in the coffin for those few still banging on about solar activity causing warming.  If the sun was outputting more energy the entire atmosphere would be warming.  The fact we can observe the stratosphere in fact cooling is a sure sign of increasing carbon dioxide in the lower atmosphere.

[lassie23]

This winter was a redux or last winter which was a redux of a redux of a redux of the past 4 winters

[TEITS]

One thing I would like to know more about (I may contact Met O) is when it comes to computer modelling there is one aspect many seem to forget. We assume they are fed the data and off they run working out the equations. However all hardware requires software and it is this software in my opinion why different models differ in performance. 

I would like to know more about how the software is programmed. Some say historical weather data is fed into the system whilst others disagree. How often do they update this software? Do they even know the cause of what has changed in our winters these past few years as something has obviously changed. If they don't know the cause then how can they remedy this into the software program?

This winter we would of been better off using my Commodore Vic 20 that is currently sat in my loft.

[Penrith Snow]

 CryoraptorA303 Good a analysis.

Pattern matching isn't perfect at all but IMO it's more useful than the mumbo jumbo of the MJO etc, and yes I have read the papers.

Put it this way, pattern matching clearly pointed to a very mild February, I said as much in  late January but the new so called signals pointed to extensive northern blocking and a cold month, even the MetO were taken in.

It was never going to happen! December and January were very similar to other strong El Nino winters so the outcome for February was inevitable, add to that the fact that mild December's and January's are nearly always followed by a mild February, 1983 is a rare exception.

The fact that other 'signals' got it so badly wrong goes to prove how unreliable they are as a forecasting tool, yet we still hear evangelist going on about MJO phases, AAM and the dreaded mountain torque, give it a rest guys or at least explain why it went so wrong while good old fashion pattern matching had February spot on.

I respect people who research new areas in meteorology but some humility would be welcome.

Andy

[rwtwm]

1 hour ago, TEITS said:

Some say historical weather data is fed into the system whilst others disagree. How often do they update this software?

It's a matter of fact that (with the recent of the exception of the AI models) the models work from observations and apply physics from there. Agreement is irrelevant. 

Updates seem to be every 18 months or so, but it varies by model. Whenever people post parallel outputs, that's because the met organisations are testing updates to their models.

17 minutes ago, Penrith Snow said:

yet we still hear evangelist going on about MJO phases, AAM and the dreaded mountain torque, give it a rest guys or at least explain why it went so wrong while good old fashion pattern matching had February spot on.

At lot of the MJO predictions have been based on composites. That's good old fashioned pattern matching to my mind. Just looking at different patterns!

[CryoraptorA303]

 Penrith Snow I'm just going to say it, the mid-term models absolutely suck at mapping winter in Europe. So much so that this year the long-term seasonal actually predicted the winter better than them. We were repeatedly coldbaited when it was obvious to everyone on the ground that we were just going to have mild, mild and more mild all winter. We've been struggling to shake off the low-mid teens all winter, even up north without the UK high we had in January. So far this February has been so abnormally mild that if this continues, by the end of the third week this will probably already be the warmest on record no matter what. We have not seen a daily high below 12.1°C all month, remotely normal daily minima have been reached nationwide on maybe three nights of the whole month, and today will mark something like the sixth or seventh day with a high above 14°C. A warm wave akin to the ones in 1998 or 2019 occuring later in the month will just push this February to another level and we'll end up being the warmest on record by a degree or more.

This has been the most Iberian winter England has ever experienced, with persistent low-mid teens, gross dullness and wetness. If it wasn't for the UK high in January that broke it for a couple of weeks then I have no doubt in my mind that this would've been the mildest winter on record. Quite telling that February 1989 is one of this February's closest relatives.

 Djdazzle

15 minutes ago, Djdazzle said:

I do not want cold in March. It's a spring month, so I'm now looking for the first 21C of the year.

You might not have to wait until March at this rate. A severe warm wave in the latter half of the month would not shock me by breaking the 21.2°C record at this point. Even if that doesn't happen, March will almost certainly see a warm wave that easily gets us to the 20s, quite possibly for several days, by the looks of things. While it wouldn't surprise me, I wouldn't quite bet on a March record yet. I'm betting on a severe May heatwave this year.

[Met4Cast]

8 hours ago, Weather-history said:

Has this been a bad winter for the teleconnections pundits and medium/long models?

In a word, no, as my post 1 page back attempted to diagnose & show. 

 

2 hours ago, Penrith Snow said:

The fact that other 'signals' got it so badly wrong goes to prove how unreliable they are as a forecasting tool, yet we still hear evangelist going on about MJO phases, AAM and the dreaded mountain torque, give it a rest guys or at least explain why it went so wrong while good old fashion pattern matching had February spot on

Perhaps if you took the time to read my post 1 page back on this thread you’d see that i’ve done exactly that. You’d also see that broadly teleconnections have gone largely as anticipated throughout this winter but shorter term changes within the stratosphere (that cannot be predicted) have caused a lot of issues. 

[raz.org.rain]

 Met4Cast subjects like these make you realise just how complex the atmospheric system is. You really have to question if human industrialisation and subsequent heat trapping GHGs are trumping the cooling characteristics of our atmosphere. It's why I'm somewhat sceptical of any climatic computer modelling, there's seemingly no ideal way to appropriately account for GHGs so they invariably underestimate the potency of anthropogenic warming. There was a study into how the jet stream may respond to climate change in the near future that suggests it could end up shifting polewards. I'd imagine the effects on Europe's climate would be significant.

 

Storm-Steering Jet Stream Could Shift Poleward in 40 Years

WWW.SCIENTIFICAMERICAN.COM

---

Changes in the position of the fast-moving air current could disrupt weather patterns

 

[Methuselah]

 raz.org.rain But Deniers still accuse climate models of overestimating GW (remember the 17-year-long 'pause'?) but, anyway, empirical data suggest that they are within the IPCC's error bars?

I think it's the havoc AGW is causing locally that's the biggest problem, as some parts of the world have experienced super-cold conditions but many more are having repeated droughts and heatwaves. Take the UK for example: 40C, one year, and ridiculous amounts rainfall, the next. . .

Methinks we're all up Sheet Creek without a paddle. 

[North-Easterly Blast]

 Penrith Snow Although not the same drop as 1983 I can think of a few other years that had a February that was considerably colder than a mild December and January (winter 1931-32, Dec 5.3, Jan 6.3, Feb 2.9, winter 1929-30, Dec 5.8, Jan 5.6, Feb 2.5, winter 1872-73, Dec 5.3, Jan 5.2, Feb 1.8).  A more recent example to a some extent of a winter with a colder February compared to December and January could be 1993-94, with Dec 5.5, Jan 5.3, Feb 3.2).  Although not fitted neatly into calendar months, the warmest January on record in 1916 (CET 7.5), was followed by a mid-February to mid-March period with a CET of 2.5 (still a 5*C drop from January to the mid Feb to mid Mar period, like the calendar month of Feb 1983 was from Jan 1983.

I have had a look at really cold Februarys (CET sub 2*C), and most of them have happened in winters that did have a significant cold spell at some point in the earlier part of the winter, though not all of the sub 2*C CET Februarys have been in winters where the earlier part of the winter has been continuously cold.  Therefore I conclude that there is certainly good evidence to suggest that if there has not been any significant cold spells in December or January, then it has always been very unlikely for seriously cold conditions to develop in February that are certainly enough to bring the CET under 2*C.  On top of this, I also looked at the overall temperatures for the period from the Septembers prior to the cold sub 2*C February CETs.  My findings concluded that the majority of our really cold Februarys (sub 2*C CET) were preceded by a September to January (five month period) that was close to average, not warm like the most recent September to January period has been. 

Though that said, in the past after other winters with a very mild December and January, not all Februarys have been super mild like this year; some of them have been closer to average. 

Given the fact that the earlier part of this winter had not brought any significant cold spells (other than very short lived cold snaps), and especially given the overall temperature profile since last September as described above, I thought that it would never bode well for any serious cold to develop in February.  Although I never on the other hand thought that it was a guarantee that it would be super mild like it has been and since January 21st as well.  It has basically been one short cold snap last month abruptly falling apart on January 20th and then we have faced a month's worth of weather close to rivalling the warmest 28 or 31 day CET period on record in January and February.

[Weather-history]

 jules216

This is where I have problem with analogs. 

Look at the years given and try apply them at a more microlevel, in this case the UK

Take April

Years given 1987, 1998, 2007 and 2016

Now look at the CETs for those years.  How can anyone extract what possible the weather maybe given the difference in those CETS? 

Edited February 16 by Weather-history

[jules216]

 Weather-history i Can apply most pretty Well in general to Slovakia for March however. About end of first decade in March came cold spell. The combined anomally vs what clusters start seing in the begining of March. I am quite satisfied with this as starting point.