Model Output Discussion - Into 2024

[bluearmy]

Just now, Allseasons-Si said:

Well somethings got to give because the CPC remains resolute.

 

Well that’s raw naefs centred  day 11 was very strong on the pattern 
no eps included at the weekend but day 11 eps mean was the same in any case 

1 minute ago, Mike Poole said:

It is run with 0z data.  But the 46 dayer is 100 separate ensemble runs.

And basically noting to do with the eps suite any more 

It’s a different model 

[northwestsnow]

Just now, Mcconnor8 said:

Yes all the weeklies run off the 0z data runs.

Well I didn't know that thanks..

We need upgrades tomorrow...

[snowray]

1 hour ago, Lukesluckybunch said:

You just have too laugh at the way it diverts the cold away from the UK at day 10..hmm as others said hope it's an outlier

I noticed that the ECM likes to tease us with cold charts that come to nothing, but when a real outbreak of cold weather is on the cards it's as miserly as the BBC!

[Uncertainty]

The mjo moving into P4 more strongly than expected and for longer, as well as the the bigger push of cold into the states is likely responsible for the flattening across the ecm suite today. Need to see a few more 46 runs to really lose faith. Nevertheless, no ssw and the mjo in a poor position is not ideal. Hopefully we can get a gfs 12z style couple of snowstorms in first before the relaxation. After that, plenty of evidence for the blocking to return with continued strat disruption and the mjo moving towards the WP. 
 

 

 

The 46 has backed off once or twice in the last few weeks and quickly got back in track. But the sooner the mjo gets a move on out of p4 the better.

 

[Mcconnor8]

Just now, northwestsnow said:

Well I didn't know that thanks..

We need upgrades tomorrow...

This is why the EC46 run a couple of weeks ago was showing a SSW despite the 12z EPS from that day having backed away strongly from it while the 0z EPS that day was for the SSW, and then the next day the EC46 also did as it was using the 0z data from the next day.

[Jeremy Shockey]

ECM 12z ensembles .  Op near the top. Mean very good indeed ( in the crucial period 14th to 16th )

[Weathizard]

2 minutes ago, Allseasons-Si said:

Well somethings got to give because the CPC remains resolute.

 

Encouraging, Quite important given the GFS/EC difference in that key area of low pressure tracking up the ESB, that is where my eyes are drawing to at 168, if that blows up and flattens out the pattern as the ECM main run suggests then it could end up in a dogs dinner of a pattern where we somehow draw the short straw again.

[Mcconnor8]

Will be watching closely on the 18z runs whether the shortwave present on the ECM at around T+120 moving North from the Atlantic Low is there or not, if not then we should be good.

[WYorksWeather]

12z ECM mean is a slight upgrade on the 00z, lowest point post 15th is around -6C for London, vs. -4C this morning.

Edited January 6 by WYorksWeather

[Midlands Ice Age]

On 04/01/2024 at 12:39, Midlands Ice Age said:

Whilst quiet in here(?), I did some work back in 2018 on most of the severe cold spells since 1860 showing where/what  they came from and what was the temps (850's) in the region the outbreaks occurred (generally Iceland for northern based and 'Moscow' for eastern based) ln and what was the CET as a result.

I did this by examination of the reanalysis charts after consulting people on the web (and in here) about what cold outbreaks to include.

I tried to publish the data, but couldn't do it properly, I might try again soon now if I can find the data. 

Anyway I still have the hardcopy and I am checking the data, and I give the relevant data in summary below.......

 I organised it in month order -  

During November, until Nov 2010 (Greenland based) there were very few other cold spells (3  - in 1978, 1904 and 1890.  (temps of 850 in 2010 was -14C, but fell to -20C at source,     CET -1.9C)

In December - 25 occurrences  mostly from straight northerlies. Greenland 'forced'   - a few of them, but mostly easterly based. when -28C  (in Dec 1892) was the coldest source and the CET was --3.12C).

Most Greenies started as a UK high and the coldest 850''s was in 1927 and 1892 when -24C was the 850.  These produced CET's of 

--2.6 and -3.37C respectively.

 

So onto January - when I recorded 31 cold spells.

Most based upon Siberian highs after Northerlies and a UK high..

Greenland based were in 1987 (-3.0C) , 1982(-4.03C),  1963(-3.56C), nearly all the early 1940's,    coldest (-3.49 in 1940),   1894(_4,4C) ,  1881 (-4.38C) .

These were produced by the lowest Greeny based 850's of between -20C and -28C) from the air sourced above/over Iceland. 

850's over the UK were (associated with the colder outbreaks) between -14C and -18C, otherwise -10C to -14C.

Gives you an idea of what may be possible if things go according to way they are being shown today.

I also have data for February, March and April - but I will leave that for later!!!, 

This is not to say something  will happen - as there are many 'risks' to overcome first.

I also have some research into that looking at how many cold spells followed blizzards, but I will leave that to another post.

Hope the mods are happy me posting this on the model thread, but many people would like to know what is the 'worst/best ' we can expect.

MIA

 

After the good response to the above post, I decided that (as opposed to the best values above), I would produce an average expectation as to what the CET might be from both the UK 850's and the 'Source' 850's. IF IT WAS POSSIBLE...

 Before I present any results I think people need to understand the restrictions -

1) The cold weather investigated had to be at least a 5 day spell of cold temperatures from the CET record and not to average above 0C.

2) as stated above data for the 850's was taken from the Meteorciel Renalysis charts either north of Iceland (for Greenies or a straight northerly) and Moscow for either Scandies or Russian  cold.

3) To match up roughly with the CET, I used 850's from 20 miles within the UK coastline, hence eliminating extreme Scandy highs in the North Sea, etc.

4) The CET  was the average of the 5 days or more of the event, and  if extended or repeated in the same month/year would show as separate events. Thus 1962 - 63 has 3 events (one for Dec, one for Jan and one for February).

 

It looked a reasonable simple task with 4 events for Nov,  31 for December, 33 for January, 30 events for February and 4 for March, in the database.

 

I discarded the November and March data as they were few in number  and had occurred mainly in the last 50 years . Why has this started to happen?

Also it was clear that the heat of the sun was a factor.

 

So the total number of events in my database was 94 - a reasonable sized data repository.

So fairly obvious (I thought) was to plot the 850 against CET  -          but this produced spaghetti.

So I repeated by month and it gave some results which showed that cold in December was better from the 850;s , than January, however  they were both  in  the main much colder CET wise  than February.   HOWEVER,  there were 2 events in early February which were exceptional (vis 1/2/56 CET 4.08C, and 5/2/1895 CET 4.81C) which were brutal. They both had severe Arctic  blizzard outbreaks before clear skies moved in from the East allowed the night time temps to plummet for several days.. 

 

So turned to compare the 'Source 850's' against the CET.

Immediately it was obvious that a better linear pattern was present for 85 of the events, with some sticking out above  (the Feb ones above), and some below - which all turned out to near UK based highs -  where the cold temps had built up 'in situ'  with high/medium  850's, which hadn't changed much.

However,  for the vast majority of cases the linear fit was -4.5C for -30C for Source 850's -  at the top of the range;   to +0C  CET for an -4.C Source 850 at the lower end of the range.  The above straight line graph also includes values  of about +/- 0.75C CET  on either side for each 850 of Source cold for the distribution.

To help a bit I have produced the below table of the straight line graph.

Source 850        -30C       -25C       -20C         -15C         -10C            -5C              0C

CET                    -4.6C      -3.8C      -2.9C         -2.0C        -1.2C         -0.4C            +0.5C

Please remember that values can be increased or decreased by about 0.75C from these valued depending upon factors like cloud and wind, over the UK, which can change the average position.

The most extreme position I have heard of was in the 17th century when a 50 mph blizzard directly from Siberia kept the temps around -10C for 3 days solid in SE England, whilst the rest of us 'boiled'  in -6C CET's..  Must have been some watch!!! Pity we didn't have N.W.  in those days.

The above are only intended to be a guide,     possibly for the less experienced.

!I hope you will find these useful many times in the future, and lets hope for top quality weather watching  this year.

MIA                              

 

Edited January 6 by Midlands Ice Age

[Frosty.]

I remain optimistic for my fellow coldies regarding week 2 regarding some meaningful arctic cold looking at the longer term ECMWF 12z ensemble mean, hopefully something is afoot following a benign anticyclonic spell.. hopefully NOT an actual foot.. …being serious though, I want January to produce a real taste of winter..for a change! Fingers crossed  

[sheikhy]

14 minutes ago, Allseasons-Si said:

Well somethings got to give because the CPC remains resolute.

 

Yeh and thats gona be the ecm giving in tomorrow morning!!!!!!

14 minutes ago, Allseasons-Si said:

Well somethings got to give because the CPC remains resolute.

 

Yeh and thats gona be the ecm giving in tomorrow morning!!!!!!

[Lukesluckybunch]

What's the general feeling then looking at the models..is this cold spell mid month very much on..or are these ecm runs on to something,surely we need cross model agreement sooner or later.

[Mcconnor8]

1 minute ago, Lukesluckybunch said:

What's the general feeling then looking at the models..is this cold spell mid month very much on..or are these ecm runs on to something,surely we need cross model agreement sooner or later.

Looks to be a rogue ECM run regarding a shortwave heading North from the Atlantic low at T+120, every other model doesn't have it, or like the GFS makes much less of it, ensemble pack of ECM also doesn't really so looking good currently.

[feb1991blizzard]

Still not seeing enough of this for me to get really excited.

 

[Severe Siberian icy blast]

26 minutes ago, Daniel* said:

that’s true but there’s a major difference in outcomes between fleeting northerly cold winds and something more sustained. I just feel it is important to stress and express the uncertainty, recent runs from EPS have deteriorated and it’s not just the OP. Confidence is low. 

Jan 16th… Jan 20th

 

The second one just spells battle ground to me, greenland still warm and Europe still cold. Its 14 days away lol

[knocker]

5 minutes ago, Mcconnor8 said:

Looks to be a rogue ECM run regarding a shortwave heading North from the Atlantic low at T+120, every other model doesn't have it, or like the GFS makes much less of it, ensemble pack of ECM also doesn't really so looking good currently.

The gfs temp panels through the 10-15 period show an increasing number of members along the same limes as the ecm. Oh suffering succotash

Edited January 6 by knocker

[Cheshire Freeze]

1 minute ago, knocker said:

The gfs temp panels through the 10-15 period show an increasing number of members along the same limes as the ecm.

Yeah a few lemons in there for sure

[Mcconnor8]

1 minute ago, knocker said:

The gfs temp panels through the 10-15 period show an increasing number of members along the same limes as the ecm.

The ECM det run never got the cold in to begin with, by day 9 only 2/3 of the 30 member GFS ensemble failed to do so so not really any support for the ECM run within them, evolutions post day 10 temp wise in the GEFS I'm not paying much attention to yet until the initial evolution is sorted out.

[WYorksWeather]

9 minutes ago, Midlands Ice Age said:

After the good response to the above post, I decided that (as opposed to the best values above), I would produce an average expectation as to what the CET might be from both the UK 850's and the 'Source' 850's. IF IT WAS POSSIBLE...

 Before I present any results I think people need to understand the restrictions -

1) The cold weather investigated had to be at least a 5 day spell of cold temperatures from the CET record and not to average above 0C.

2) as stated above data for the 850's was taken from the Meteorciel Renalysis charts either north of Iceland (for Greenies or a straight northerly) and Moscow for either Scandies or Russian  cold.

3) To match up roughly with the CET, I used 850's from 20 miles within the UK coastline, hence eliminating extreme Scandy highs in the North Sea, etc.

4) The CET  was the average of the 5 days or more of the event, and  if extended or repeated in the same month/year would show as separate events. Thus 1962 - 63 has 3 events (one for Dec, one for Jan and one for February).

 

It looked a reasonable simple task with 4 events for Nov,  31 for December, 33 for January, 30 events for February and 4 for March, in the database.

 

I discarded the November and March data as they were few in number  and had occurred mainly in the last 50 years . Why has this started to happen?

Also it was clear that the heat of the sun was a factor.

 

So the total number of events in my database was 94 - a reasonable sized data repository.

So fairly obvious (I thought) was to plot the 850 against CET  -          but this produced spaghetti.

So I repeated by month and it gave some results which showed that cold in December was better from the 850;s , than January, however  they were both  in  the main much colder CET wise  than February.   HOWEVER,  there were 2 events in early February which were exceptional (vis 1/2/56 CET 4.08C, and 5/2/1895 CET 4.81C) which were brutal. They both had severe Arctic  blizzard outbreaks before clear skies moved in from the East allowed the night time temps to plummet for several days.. 

 

So turned to compare the 'Source 850's' against the CET.

Immediately it was obvious that a better linear pattern was present for 85 of the events, with some sticking out above  (the Feb ones above), and some below - which all turned out to near UK based highs -  where the cold temps had built up 'in situ'  with high/medium  850's, which hadn't changed much.

However,  for the vast majority of cases the linear fit was -4.5C for -30C for Source 850's -  at the top of the range;   to +0C  CET for an -4.C Source 850 at the lower end of the range.  The above straight line graph also includes values  of about +/- 0.75C CET  on either side for each 850 of Source cold for the distribution.

To help a bit I have produced the below table of the straight line graph.

Source 850        -30C       -25C       -20C         -15C         -10C            -5C              0C

CET                    -4.6C      -3.8C      -2.9C         -2.0C        -1.2C         -0.4C            +0.5C

Please remember that values can be increased or decreased by about 0.75C from these valued depending upon factors like cloud and wind, over the UK, which can change the average position.

The most extreme position I have heard of was in the 17th century when a 50 mph blizzard directly from Siberia kept the temps around -10C for 3 days solid in SE England, whilst the rest of us 'boiled'  in -6C CET's..  Must have been some watch!!! Pity we didn't have N.W.  in those days.

The above are only intended to be a guide,     possibly for the less experienced.

!I hope you will find these useful many times in the future, and lets hope for top quality weather watching  this year.

MIA                              

 

Was the one you mean from the 17th century 1683/4?

That was the year with the most famous of the Thames frost fairs, and one in Leeds on the River Aire. Allegedly cold enough for English oak trees to explode in the cold, and sea ice to form in the Channel. I have no idea what the temperature must have got to, but I wonder if that winter might have achieved -30C at the surface.

 

[Bricriu]

2 minutes ago, knocker said:

The gfs temp panels through the 10-15 period show an increasing number of members along the same limes as the ecm.

It's rare a cold spell wouldn't have a relaxation period.  I wouldn't  be concerned about that  really. Although you being a mild lover might hope for it:)

[tight isobar]

Shabby they are not- but highlighting the ec’s not wanting -clear cut routes!!… 

[That ECM]

6 minutes ago, knocker said:

The gfs temp panels through the 10-15 period show an increasing number of members along the same limes as the ecm.

Would that not be expected if lows are to approach from the west, south west? Some getting all the way through, some not making it and some sliding underneath the cold? I’d appreciate your thoughts as I respect your analysis.

Edited January 6 by That ECM

[Beanz]

12 minutes ago, Lukesluckybunch said:

What's the general feeling then looking at the models..is this cold spell mid month very much on..or are these ecm runs on to something,surely we need cross model agreement sooner or later.

It’s 10 days away, you won’t get cross model agreement at that range.

[bluearmy]

Well now we have at T228 

GEPS                                                             GFS                                                               EPS

     

                                                                                                                                        
 

This looks ok in general. The geps are incredibly consistent whilst the eps are already showing signs of losing the amplification to our west. 

there is no doubt that compared to three days ago, the output for this date has slipped on the eps.  Whilst a few days ago we had some posters commenting that we were possibly on the cusp of a prolonged deeply cold period, that view will have undoubtedly changed now.

my view is that I really hope we can get something white on the ground across large parts of the country before this flattens out.

we shouldn’t be surprised to see the nwp showing inconsistencies given the upper strat reversal ongoing and it’s ebb and flow. 

i really hope we get some cross model agreement by Monday on the push sw of low heights, irrespective of the result. then we can at least know how much attention to give to the output!