Model Output Discussion - mid Autumn

[Metwatch]

2 hours ago, johnholmes said:

UK Met on their Fax outputs going for really cold air into the northern half of Scotland, 528 DM line along the Great Glen!

UK Met Office Fax Charts - UKMOMSLP Analysis and Prognosis

WWW.WEATHERCHARTS.ORG

---

UK Met Office . . Europe and North Atlantic MSLP analysis and prognosis charts

 

528 DM line into Norfolk on the GFS 06z operational, 00z had it to Lincolnshire!

Looks like another lovely weekend to come, dry, sunny but perhaps over 10C cooler both by day and night compared to the one just gone. Further north joining as well this time. Then from a north westerly flow a UK high moves in, and potentially moving east / north eastwards which becomes a Scandi high mid next week. Interesting to say the least for mid October!

4 days of sunshine symbols from the Met office forecast here starting Saturday.

 

Fits nicely with the MJO, set to stay around phase 8 -1. Composite anomaly reanalysis for phase 8 looking fairly similar to the output for next week but some key differences. Similar case for the phase 1.

Edited October 11, 2023 by Metwatch

[AltonMatthew]

Every year its the same with model watching. Have we not learnt by now that comparing previous years proves ineffective for predicting whats in store for us in the coming season. I would personally prefer to use a model watching forum for looking at the output of models readily available now and make an assumption based on that. 

[Methuselah]

Today's GFS 06Z looks like a case of 'anything could happen', to me; not that's unusual: 

Day 5:        

Day 10:      

Day 15:      

So nothing worth worrying about and no portents (mild or cold) for winter, either. 

And the GEFS ensembles diverged rapidly after around Day 6:

               

               

Edited October 11, 2023 by Methuselah

[danm]

 

[LRD]

31 minutes ago, danm said:

 

The MetO's presentations are excellent

[Mike Poole]

GFS 12z once again brings in the easterly.  Will be interested to see where this sits amongst all the ensemble suites later.  T240:

It’s interesting also that the core of the developing trop vortex is displaced away from Canada.  

[Metwatch]

2 hours ago, Methuselah said:

Today's GFS 06Z looks like a case of 'anything could happen', to me; not that's unusual: 

That is further emphasised tonight with the latest GFS and GEM.

Both aren't overly different in terms of the pressure pattern across Europe, but the airmass is extremely different over the UK.

Both have troughs west of Portugal next Thursday, but on the GEM it's much stronger. The Scandi high is further north and more elongated with the GFS, which allows quite chilly air, especially for Mid October to arrive from the northern Scandinavia / north west Russia. The Scandi high on the GEM is in the far south of Scandinavia, almost into Germany, struggling to move more north.

The ICON 12z is almost at that time frame too and seems to be closer to what the GEM wants.

Edited October 11, 2023 by Metwatch

[sheikhy]

Gfs is very cold for the time of the year!it just aint letting go of that scandi high!!boy its gona be biting from saturday onward especially by night!!

[CoventryWeather]

12 minutes ago, Mike Poole said:

GFS 12z once again brings in the easterly.  Will be interested to see where this sits amongst all the ensemble suites later.  T240:

It’s interesting also that the core of the developing trop vortex is displaced away from Canada.  

Way out in lala land, but interesting to see the GFS pretty much has no proper organisation to the Polar Vortex, apart from a weak core in Siberia. Seems to be a theme recently for a much weaker PV at the end of the month into November on recent zonal charts - bonds well for blocking next month perhaps. 

[Rain All Night]

56 minutes ago, Mike Poole said:

It’s interesting also that the core of the developing trop vortex is displaced away from Canada.  

Do we know of any reason why it would be over there at this time? I assume nothing to do with the stratosphere, since I've seen you mention that the two layers aren't coupled at the moment?

[Mike Poole]

2 minutes ago, RainAllNight said:

Do we know of any reason why it would be over there at this time? I assume nothing to do with the stratosphere, since I've seen you mention that the two layers aren't coupled at the moment?

Well, the trop PV tends to end up over Canada later into winter when it is coupled to the strat vortex.  So I would say the reasons for it being where it is now / modelled to be is not to do with the stratosphere, just an evolution of quite interesting tropospheric patterns - permitting blocking, we’ve not been able to shake high latitude blocks for most of summer, seems to be continuing when perhaps one might expect a period of zonal weather into the second half of Autumn.  Beyond that, I don’t know.

[Lukesluckybunch]

if we had what the gfs is showing,,in 3/4 weeks time we would probably be looking at very cold and snow showers..with this set up the cold easterly..ECM will be interesting shortly..UKMO wasnt as good as gfs btw

[Derecho]

I hope this is a proxy for the coming winter....

Looks like the high will stick to Scandi later next week rather then sink over central Europe so quite chilly overall.

UKMO has the centre of the high over the North Sea though so perhaps not as chilly. Should be some useable weather at least in this instance though so I'll happily take either.

I'll admit I still prefer warm weather at this time of year though not too fussed either way. When the clocks go back I get into winter mode.

Edited October 11, 2023 by Derecho

[Lukesluckybunch]

9 minutes ago, Derecho said:

I hope this is a proxy for the coming winter....

Looks like the high will stick to Scandi later next week rather then sink over central Europe so quite chilly overall.

UKMO has the centre of the high over the North Sea though so perhaps not as chilly. Should be some useable weather at least in this instance though so I'll happily take either.

I'll admit I still prefer warm weather at this time of year though not too fussed either way. When the clocks go back I get into winter mode.

with the current set up predicted..would the gfs produce snow eventually..or is it too early and also a lack of cold air in north and east europe at this point?

[Derecho]

5 minutes ago, Lukesluckybunch said:

with the current set up predicted..would the gfs produce snow eventually..or is it too early and also a lack of cold air in north and east europe at this point?

Too early to get snow from the east in October I think! The few times I've seen snow in October are from long fetch northerlies (2018, 2012 and 2008 being those instances).

The earliest snow in somewhat recent times from a Scandi high was mid-November 1993.

[Addicks Fan 1981]

5 hours ago, WYorksWeather said:

Tend to agree with most above - a prolonged cold spell is not on the cards until both models agree. I'd be very sceptical of taking GFS over ECM - it has a lower resolution and all else equal is more likely to miss something that the ECM will catch.

The ECM has cooler conditions persisting for most of us through T+144:

It then gradually evolves to this very mild pattern by T+240:

The mean also supports this evolution.

I tend to trust the ECMWF a lot more, and I think this evolution is very accurate.    

[Cambrian]

The switch around over Scandinavia between day 5 and day 9 in terms of heights is profound, from around 30 dam below average to over 30 dam above average in just 4 days on the 12z GEFS mean.

 
 

The ensemble spread at day 9 is remarkably small for heights over Scandinavia for the marked change, a slightly greater spread further south over Germany, suggesting that’s where the degree of ambiguity is greatest regarding the shape of the high.

 

The equivalent chart for the T850s spread shows a generous spread along the whole of the southern flank of the high, suggesting low certainty in the coldness of the air over a wide tranche from Kazakhstan west to the UK and Ireland, at this stage perhaps indicating a fundamental lack of confidence in how cold it will get through Europe so early in advance of proper winter.

Nevertheless, it’s a rocking signal for a big change of flow over Northern Europe, the extent and suddenness of the change to a setup like this within a month of the autumn equinox is justifiably raising eyebrows. Very interesting already for just the middle of autumn. 

[Man With Beard]

ECM D10 for 850s - That's about as much of a contrast in weather as you can get across northern Europe:

[Addicks Fan 1981]

15 minutes ago, Man With Beard said:

ECM D10 for 850s - That's about as much of a contrast in weather as you can get across northern Europe:

About right ️ to be honest mate and the NOAA anomolies back it too.  Amazed that the NOAA anomolies aren't given a mention as often they see things a lot more than the other models.    People get too preoccupied with looking at op charts that spark their individual interests.    

[Metwatch]

Quite a funky pattern towards the end of the ECM 12z, similar to the 00z and its ensemble average, with a lot of deep lows around the Azores and south west of us instead of the typical Azores / Euro heights you would expect during most zonal Octobers.

We're quite close to the colder airmass over central and eastern Europe, but that very anomalous warmth right over us yet again. Probably nothing extreme in terms of surface temperatures from average though. Very sharp contrasts between the warmth and the chill. Interesting to see if that colder air from the east will reach the UK or not later next week. GFS seems to be most bullish and out of the pack regarding the cold.

Edited October 11, 2023 by Metwatch

[WYorksWeather]

I'm going to have a look at the GFS and ECM in detail, to try to see what is causing the divergence. GFS shows a very prolonged cool/cold spell, whereas ECM shows a fairly rapid return to typical or even above average temperatures.

Here's 12z GFS and ECM at T+72. Not much to see here.

 

Still not much to see at T+120, except that the ECM is a bit flatter, with a weaker Greenland high and a weaker low over Eastern Europe.

 

By T+168 though, a small but very significant difference. Both models develop a low to our south that begins to pull warm air northwards. The GFS has the low weaker and over mainland Spain, whereas ECM has it a bit stronger and over the Bay of Biscay.

 

End result is that at T+168, the mildest air is kept well to our west on the GFS, whereas the ECM scores a direct hit.

 

And by T+240, the runs are poles apart. Both show an Omega setup, but GFS has us on the cold eastern side, and ECM in the middle and very mild.

 

The question is whether one model is right or wrong, or whether they blend their solutions in the end. If either one turns out to be right, it'll be very memorable. GFS produces a prolonged colder than average spell which is something we haven't seen at all this year. Not record breaking, but impressive given the prevailing conditions, with probably multiple ground frosts for most of the country and some hill snow and air frosts in places.

ECM might be record-breaking in a very different way! Below is the 850hPa temperature output for the UK at T+240. Incredibly, it gets the +15C isotherm in yet again. I'd be interested to know, is anyone aware of us ever getting the +15C in this late in the year? If it did verify, might well be good for low 20s at the surface, which would challenge date records.

Personally I'd trust ECM over GFS given the choice. Either way, one of the models is going to look silly in a few days. If we do get a prolonged cold spell I'll apologise to the GFS!

[damianslaw]

Interesting model output, synoptics we haven't seen for I can't recall how long. A chilly NW/ N flow through the weekend, heights build in behind, but instead of sinking away to the south as often the case, we see low pressure develop to the south, propping up the high. The evolution theron remains uncertain, GFS builds heights strongly to the north of scandi, this has the affect of pulling heights NE and we draw in a cold easterly. ECM keeps lower heights closer to the south, and makes less of heights building to the north of scandi, this has the affect of sending core of heights ESE, and we end up with a very mild SE flow with low heights to the SW poised to then move in. Both models though differ in eventualities, are similiar in they don't show an active zonal flow setting in. Indeed ECM would likely trend cyclonic with every possibility heights to the north block the atlantic. 

[Alexis]

A real dry run for winter today.

ECMWF / UKMO  v GFS / JMA

with the GEM sort of a blend of solutions.

Edited October 11, 2023 by Alexis

[minus10]

Gds12z seemingly doubling down on the cold pattern from the ene with potential reloads..

...quite an amazing set of 850 ens from gfs as well..

..although at the bottom from around 18th..the op is supported by a small cluster...think it was mentioned in the met 10 day trend that the easterly option was given approx 10 % probability..so still unlikely given the ecm run referred to in the excellent comparison analysis by @WYorksWeather above, however still a possibility and certainly not at present anyway showing the typical zonal flow you may expect at this time of year...

 

 

 

[WYorksWeather]

20 minutes ago, Metwatch said:

Quite a funky pattern towards the end of the ECM 12z, similar to the 00z and its ensemble average, with a lot of deep lows around the Azores and south west of us instead of the typical Azores / Euro heights you would expect during most zonal Octobers.

We're quite close to the colder airmass over central and eastern Europe, but that very anomalous warmth right over us yet again. Probably nothing extreme in terms of surface temperatures from average though. Very sharp contrasts between the warmth and the chill. Interesting to see if that colder air from the east will reach the UK or not later next week.

I'm not sure about the lack of extreme temperatures at the surface. Those ECM 850hPa anomalies are pretty extreme for the time of year, widespread +10C:

It's enough for temperatures like this on the last four frames from T+222 to T+240:

   

Spot maxima in the high teens, meaning that in practice accounting for model bias you are looking at temperatures probably scraping into the 20s.

 

 

 

Edited October 11, 2023 by WYorksWeather