Model discussion highlights

[Kirkcaldy Weather]

Right I thought its a great idea making a post on exactly what ingredients are best plus discuss general requirements for snow  

My personal snow checklist is the following 

• Dam of AT LEAST 528. Values below this are also perfect
• THE MOST CRUCIAL ASPECT -- YOU NEED A DEWPOINT AT OR BELOW 0C. A dewpoint higher than 0C will be sleet you often notice rain starting to mix with wintry precip from about 1.5 > 2C dewpoint but when it drops to and under zero degrees its pure snowfall given you have other fitting parameters but DP really is king.
• 850hpa temps you start to notice the turning of rain to wintry precip from -5 generally but perfect ingredient being 850hpa at -8 with the snow changing its structure ie wet to dry with the lower you go at 850hpa 

Looking closely at this pattern I've identified the key timing which will be crucial if it is indeed snowfall your expecting.

Very akin to how Jan into February 09 played out

Another fine instance of the separating energy with the Rossby Wave Breaking

Here's what you'd look for in this setup its the energy along the southern isobars of the Scandi High. 

Ideal scenario would have a similar upper level cold pool generate but this is dependant on how much energy goes west into said cold pool vs how much moves east into the cyclone upstream

 

We've got a blend of the 2009 setup with 2018 and 2010 really, 2009 never really retrogressed to a strong Greenie high and that elongation from Scandi through Iceland over Greenie is exactly as played out in 2010 and 2018

Obviously taking what month it is occurring in gives better logic its just common sense your easier getting lower dewpoints and 850hpa values in December - say March and April than October.

Another big point will occur with the pivot shortly before the retrogressive state, this is a trait that models have struggled to comprehend even going back 13>14years and I recall with 2018 it was modelled farther south across Europe for plenty of days before the pivot was understood and the majority of times when this occurs if you track along the isobars its natural path more often comes to UK.

Elongating high in 2010

Setup we will be bringing in soon 

Looks like the pivot begins later Thurs with the WELL below average 850 temps slide in anywhere from 10 > 20C below average

Start of the week into Thurs as previously discussed. Initial cyclone northern Spain pairing to UK ridge and initial upstream cyclogenisis from Scandinavia south. Next part of the Rossby Wave begins to break sending energy and new cyclone southeast to Portugal + Spain

Thursday teatime we have the upstream cyclone deepen and increased pressure through Scandinavia particularly North. Pivot gets underway then its the question how does the energy separation develop, GEM sends a good chunk of energy westward a la 2009 which ends up being sufficient for a new cyclone crossing the UK again like 2009

Goodness some ridiculously sexy scenarios evolving at pace 

 

 

At this speed it gives the vibes this is setting the scene for Winter 23 / 24 to be 

 

 

 

Original post: https://community.netweather.tv/topic/99428-model-output-discussion-mid-autumn/?do=findComment&comment=4937130

[WYorksWeather]

Some have already had a look above but I'm going to do what is now becoming my usual post on the GFS vs ECM comparisons.

To T+72, broad agreement now on an area of low pressure sitting just to the west of Spain/Portugal.

 

Both GFS and ECM blow this up into a major low and send it blasting our way by T+120. Have to say that looks quite nasty on the wind speed charts - as an indication, UKV has widespread gusts into the 60+ mph range, some isolated spots reaching 90mph. Too early for details yet though I think. It's also quite mild though under the SE flow - ECM at 19C on raw output which will likely get to 21C in reality.

 

At T+168, both GFS and ECM still look quite miserable. The easterly never gets going properly, more of a south-easterly, so we're stuck with lows floating over us, and temperatures probably about average with not much of a diurnal range, maybe slightly above average with the mild nights.

 

It remains messy right out to T+240 - more low pressure on both GFS and ECM. ECM perhaps slightly better for the SE corner but GFS is pretty wet and windy for everyone. An ex-tropical storm is sitting in the bottom left of the ECM chart, which we should probably look out for in a few more days when we may have a better idea of what it might do.

 

Heading out of the ECMs range and looking at the GFS extended, it probably goes into silly season with this ridiculous low at T+384 (one to add to the list of GFS being drunk). A 950mb low generated from an ex-tropical storm impacting Greenland, with close to hurricane-force winds (note that this is a mean wind chart, not gusts!). I don't think we're going to see this happen somehow...

 

Overall, I'd say it's just generally quite a messy outlook. Lots of low pressure, probably not as cold after the next few days as we were perhaps expecting, but equally no signal for anything combining mild and settled weather. Probably disappointing for everyone. If like me you'd still value a few more settled and mild days at this time of year, you're probably out of luck for the moment. But for early cold fans, probably not much joy after the next two or three days either.

Both charts are fairly representative of the ensembles as well. Never really straying too far from the mean.

 

Original post: https://community.netweather.tv/topic/99428-model-output-discussion-mid-autumn/?do=findComment&comment=4937217

[Met4Cast]

Now we’re talking! 

Huge WWB (westerly wind burst) now taking place over the Pacific. This should help drive the MJO and in turn lead to more positive AAM (El Nino) conditions.

Mid-long term should be a little more predictable as a result.

The recent (albeit brief) spike in AAM tendency associated with a previous smaller WWB is the reason I suspect for the Scandi high/cyclonic (UK) pattern we’re moving into now. 

It’ll be interesting to see how things develop in the tropics & sub tropics in the next week or two.

With the MJO likely moving into phase 8, assuming we see the above rise in AAM as a result we could be looking at an incredibly wet & unsettled November. 

All the while fighting against blocking to the NE.. vortex under pressure, anyone?

Original post: https://community.netweather.tv/topic/99428-model-output-discussion-mid-autumn/?do=findComment&comment=4937707

[MattH]

28 minutes ago, sheikhy said:

Ukmo cold scandi high again!!gfs moves to ukmo as well!ecm just seems to struggle to slide the low into the continent and leaves it parked over england for days on end!now that would be worrying!!

Very difficult synoptics through the week ahead, especially regarding the position for frontal features and resultant large rainfall totals and which areas at greatest risk. The block to the NE certainly a key feature in determining the position and orientation of frontal systems as they move up into the UK from the SW. Primary areas seem to be well highlighted at the moment, with E and SE Scotland for real concern for a significant weather event later in the week from rainfall totals.

With regard to the colder ingress through the forthcoming weekend it does seem more likely than not, clearly this is October of course and not January, but it should still lower temperatures and create quite a cold/raw feel with the E'ly wind into the weekend. The 00Z EC seems to be messing about too much with the low over England and Wales, not clearing it quickly enough and the deterministic frontal position highlights this well, over E England on Sat. It isn't without support, mind, but as things stand especially for more central and northern areas, at least, temperatures should fall away again this coming weekend.

Cheers, Matt.

Original post: https://community.netweather.tv/topic/99428-model-output-discussion-mid-autumn/?do=findComment&comment=4937781

[johnholmes]

Not looked in detail at the usual charts since Thursday last week so today

Monday 16 October

First NOAA 6-15 chart

The large +ve values in the east Pacific of the last few days continues with a –ve value beneath this; so we have slight contour ridging where normally there is a trough and a small trough beneath this ridge, into a general weterly flow over N America into the distorted pattern south of Greenland into western Europe, with a marked ridge off western Norway towards Greenland, with a +be anomaly of around 180+ DM. Again this has been showing for several days. The overall pattern across the Atlantic and into much of western Europe is westerly, with a change to a more S’ly over the UK and far NW France/Scandinavia. The 8-14 loses most of the +ve heights along with much of the ridging.

For the ECMWF output for 22-26 October (much the same period as NOAA 6-10) the 500 flow starts with an very similar one to NOAA. It becomes less distorted over time but remains first N of West and then more westerly by 26 th. The 850 temps are generally showing around + 5 C with some warm up towards the end as the +5 C moves into the SE of England.

Overall the two suggest somewhat less cold air slowly developing as the trough slowly warms out slightly. The settled dry spell initially probably becoming less settled with time.

Climate Prediction Center - 6-10 Day 500mb Outlook

WWW.CPC.NCEP.NOAA.GOV

Original post: https://community.netweather.tv/topic/99428-model-output-discussion-mid-autumn/?do=findComment&comment=4937813

[WYorksWeather]

Here we go again, GFS vs ECM. Compared to the ensemble, GFS is a little cooler for the most part. ECM ensemble not out yet.

 

At T+72, deep low pressure associated with Storm Babet is on its way in to the UK from the south west, leading to a mild but very wet pattern.

 

 

By T+120, the wind gets pulled into an easterly on both models. The Scandinavian high is a bit closer on the GFS, so it isn't quite as wet. Still a very wet outlook for both models though, and temperatures fall away as the 0C isotherm comes down across the whole of the UK.

 

At T+168, GFS has produced an impressive blocking area of high pressure of 1040mb over Greenland, and generally keeps the Scandinavian high going. However, ECM is beginning to break it down, and not much sign of the heights over Greenland. Result is that GFS stays colder for longer than ECM.

 

 

At T+240, the GFS has broken down the Scandinavian high, but the deep low pressure over the UK is associated with north-westerly winds which will probably feel quite cold - 0C isotherm back over most of the UK. The ECM is much more typical autumn fayre - temperatures much closer to average, pulling the wind in from a westerly or south westerly, transient ridging for the south but much wetter in the north. Once more, it's primarily a cold rain vs. mild rain scenario.

 

Rainfall totals out to T+240 are extremely high. GFS has 50mm quite widely, and over 100mm in parts of Scotland and NW England. ECM rainfall chart isn't out yet, but based on what has loaded so far, it looks wetter still.

 

 

Original post: https://community.netweather.tv/topic/99428-model-output-discussion-mid-autumn/?do=findComment&comment=4938098

[Cambrian]

The changes in the models over the last few days regarding the relatively easy establishment of lower heights over the UK and Ireland to the southwest of the Scandinavian high later this week reminds us that it is, after all, an “early season” block that we’ve been looking at, with no deep cold behind it, much easier for the Atlantic to unpick it at its southwestern flank, even with a southerly tracking jet stream.

Nevertheless, this tussle between pulses of energy running through the Atlantic jet stream and the resolute northern block is still very much the main theme for the next few weeks.

The fine balance is nicely shown by the 0z GEFS mean for day 12. A westerly surface flow for the UK and Ireland is indicated on the pressure chart, while the heights anomaly chart not only shows the heights still there to the north - they are there all the way through - but by then also recovering over the open Atlantic, leaving the door open for the possible development of a northeasterly flow as low pressure moves through to the continent. Indeed, the low heights have extended east to Germany by then. 

 

The northern  block continues to exert its influence on the outlook, by means of directing low pressure systems from the Atlantic into Europe, potentially pulling us into colder air around the northwestern flank, the interesting “something to watch out for” step for the models over the coming days being the possible firming up with this idea of the cutting off of the low pressure from the Atlantic trough, as heights recover to the northwest over the North Atlantic which lets the low drift into Europe.

This is also hinted at in yesterday’s ECM weekly pressure anomalies for the next three weeks, 16-23 Oct, 23-30 Oct, 30 Oct - 6 Nov. These vary a fair bit from day to day, but in recent days consistently show the main focus of the Atlantic low pressure slowly slipping east into Europe, while higher pressure anomalies are maintained throughout over Scandinavia. 

  

Taken together, these charts really are an illustration of a continued southerly tracking jet stream - and the sequence - easterly, then cyclonic, then potentially back to an easterly - fits with the yesterday’s ECM regimes forecast consistently showing a combination of blocking (red), Atlantic ridge (purple) and NAO- (green) signals putting the squeeze on the NAO+ signal (blue), the negative NAO signal often being as strong or stronger than the positive NAO (I’ve cropped it at bonfire night because it gets ambivalent after that).

Taken together, this all suggests a continuation of the northern blocking through to the start of November - and also of the tussle with the ebbing and flowing of low pressure running along to its south, leaving us with slow transitions between occasions when the trough sometimes gets stuck just to the southwest or over the UK and Ireland, and other times when it gets pushed south over the continent, allowing us a few days to dry out in a chillier flow from the east. 

Original post: https://community.netweather.tv/topic/99428-model-output-discussion-mid-autumn/?do=findComment&comment=4938313

[Kirkcaldy Weather]

"The PNA will be having major say in the evolution of 500hpa setups progressing through October"

Indeed. Fast forwarding to Oct 24th we really see this switching from positive to negative phases as the Alaska trough flips to an elongating high as discussed previously this will connect with our Scandinavia / Northeast blocking and the opposites also occur as the United States west coast high pressure shifting to a trough and finally a bit of respite for the North East US toward the maritimes with the trough switches into a building high which carries the exact orientation as the previous trough from the Gulf of Mexico

Our trough will also connect energetically to the west coast trough at least briefly 

Goes in line really superb as a blending of the PNA NAO and MJO

Did have me questioning when the phase 1 setup would start cropping into play 

 

This all begins the process linking various factors from now 

First step is to get a cut off low moving over the west USA high

As bolaven enters stage left our cut off low begins to strengthen and become elongated similar to the Bolaven influence within the Pacific troughing which also boosts and increases the Alaskan high pressure 

Lovely to note also as the MJO bounces between the Western hemisphere and Pacific basins which allows the New TS - Hurricanes which play a huge role within the evolution 

The hurricane yet again on that Baja route creates the bonding of energy with the switching to western trough 

High pushing East. The MDR system > next Potential hurricane might start as a bit of problems around Turks and Caicos island then feeds its energy into the lows in the North Atlantic which are flowing into the UK and Atlantic Wave Breaking

Not surprised in the slightest seeing rainfall concerns across the UK, these are all characteristics through the PNA switching, Having the slow pace across phases 8 and 1 which is basically the algorithm for battleground setup and pair that with essentially a constant feed into Rossby Wave Breaks its not surprising because these cyclonic developments cannot maneuver further east therefore the break process is sat within Britain for longer periods

Also not forgetting into Europe as downstream cyclonic conditions continue to be prone for cut off cyclones and excess rain events ie Portugal with Babet

We experienced very similar dynamical evolution in 2021

Europe experienced unprecedented hot and wet extremes in the early summer of 2021

Extremes resulted from recurrent blocking and wave breaking caused by a persistent meridionally amplified flow over the Atlantic

Three extratropical transitions of hurricanes in close succession were critical in repeatedly amplifying the flow

Plain Language Summary

In the early summer of 2021, Europe experienced widespread weather extremes, including record-breaking floods in Germany and prolonged heatwaves in the Baltic. These extremes resulted from a complex and particularly persistent atmospheric circulation over Europe and the North Atlantic. The interaction of three closely spaced tropical cyclones with the mid-latitude circulation led to repeated amplification of the circulation, resulting in frequent wave breaking and blocking episodes that caused the weather extremes. Understanding what led to this unusual season has important implications for current and future climate risks, and is a first step toward improving sub-seasonal forecast models to better capture such events.

'JMA with phase 8 

Retrogression underway via Scandi --- Greenland

Italian Cyclone

Troughing through southern US ie GOM

Deep Russian Troughing'

There is the Italian Cyclone signals coming to reality pairing with the drier conditions along the retrogressing Scandinavia Greenland block

 

From Friday through Tuesday this is the first phase of impacts 

SCOTLAND AND NORTHEASTERN ENGLAND 

️EXTREME Impacts 

IRELAND Severe Impacts ---️EXTREME Impacts

WALES AND REST OF ENGLAND 

Severe Impacts

PORTUGAL, FRANCE ESPECIALLY SOUTH AND SOUTHEASTERN THROUGH INTO SWITZERLAND ITALY CROATIA

️EXTREME Impacts 

SPAIN 

Severe Impacts

There is a shot for a snowstorm to occur with the energy separation of the northern section of the low in Europe as I covered, GFS being most keen ATM

Original post: https://community.netweather.tv/topic/99428-model-output-discussion-mid-autumn/?do=findComment&comment=4938428

[Cambrian]

The Atlantic jet stream remains heavily invested in its southern arm over the next 10 days on the 0z ECM op, it’s tracking well to the south throughout, over Iberia and increasingly following through into the Mediterranean. There are occasional more intense pulses of energy running through it, around days 1-2, 5 and 8. 

There is also some activity in a northern arm, especially during the next couple of days, pulsing through Greenland, Svalbard and eventually north of Scandinavia. Innocuous as it looks, this little arm sets up a pincer movement with its southern counterpart, day 1, and cuts off the Norwegian Sea heights from those over the Arctic at day 2. 

 

It’s this pincer movement that helps shift the heights northeast, rather than letting them roll back west, and leads to a more substantial relinquishing of the UK and Ireland, setting us up for a decidedly autumnal weekend. 

Back to the northern arm, to rub salt into the wound - when the resulting low pressure clears through Svalbard into Siberia by day 4, the focus is on building heights north again into the Arctic, and so the desertion of the UK and Ireland is consequently a quite lengthy one, several days waiting for the trough to get to the east of us.

However, by day 10, there are signs of heights building south again, down through the Atlantic to begin with, expediting the isolation of the low pressure from any troughing further west over the Atlantic and a growing signal for the low to move east over the continent, over northern Italy by day 10.

  

The jet stream trough is comfortably over the Mediterranean by then and any northern arm activity is negligible - 0z GFS op day 10 similar to the end of the ECM animation above (the 6z GFS op looks to track the low a little further north into Europe and keeps a bit more life in the Atlantic trough, but holds the same broad idea regarding the jet stream), 

 

however, importantly, the 6z GFS mean for day 10 shows the eastward transfer well underway, cyclonic but hinting at a subsequent northeasterly flow…

  

…which in conjunction with the reassertion of heights over Iceland and northern Scandinavia looks like a very reasonable precursor to a more general easterly setup as we head into November. The ETA is tricky, but this is looking increasingly like the road we’re on.

Original post: https://community.netweather.tv/topic/99428-model-output-discussion-mid-autumn/?do=findComment&comment=4938812

[Captain Shortwave]

Initially it seemed like windward coasts would get the worst of the conditions, now the easterly seems to have been lessened (especially further south where that chillier airmass never arrives). 
Now we are face with an area of low pressure becoming slow moving across the UK and much larger amounts of uncertainty regarding rainfall.

GFS up to Saturday evening;

 

Southern England looks pretty bad along with an area from the Firth of Forth to Aberdeen.

Arpege

 

Aside from Eastern Scotland, this shows a very different story. Eastern areas are badly hit whilst towards the south west things are not too bad.

The red warning looks right but aside from Western Scotland (Which looks like escaping without much rain), any other area could be given a proper soaking depending on the behaviour of that low.

Original post: https://community.netweather.tv/topic/99428-model-output-discussion-mid-autumn/?do=findComment&comment=4938972

[WYorksWeather]

A bit later than usual, but here's the GFS 18z vs ECM 12z comparison. I'll be comparing the same valid date/time, e.g. GFS T+66 vs. ECM T+72.

The GFS 18z ensemble chart isn't out yet, but the ECM is close to its mean on the 850hPa temperature.

At T+66/72, GFS and ECM both have the remnants of Storm Babet to our south east, though the low is deeper on the ECM. Both charts show a Scandinavian high, and a deep Greenland low.

 

At T+114/120,  GFS swings in another low from off the Atlantic, whilst ECM doesn't bother with that and instead brings the remnants of Storm Babet back up the east of the UK for a second go. Quite an unusual pattern. In terms of the upper air temperatures, GFS looks quite a bit colder with the 0C isotherm covering the UK between T+66 and T+144, whereas ECM never really gets the coldest air across the country - it's mostly restricted to the north.

 

At T+162/168, both GFS and ECM are beginning to break down the Scandinavian high, though less so on the ECM. ECM turns briefly colder as the wind goes into a north-westerly, whereas GFS is fairly close to average at this stage.

Both charts though show tons of low pressure to our west and nothing much to stop it - significant rain totals are likely. GFS has an ex-tropical feature or remnant in the bottom left which may become important.

 

  

Finally, at T+234/240, GFS barrels in yet another deep low from off the Atlantic. There is quite a strong Greenland high, but the Scandinavian high has all but gone, so it's just a continuation of wet weather. ECM also shows the deep low but around a day earlier, and so by T+240 the pattern is a bit flatter. Still looks pretty poor if you're looking for settled weather.

 

Bonus chart is the cumulative precipitation charts from GFS and ECM to T+240. No sugar coating this - heavy rain and possible flooding risks, hence the weather warnings. Risk of 150mm or more in the worst hit areas, and 50mm almost everywhere.

 

 

 

 

Original post: https://community.netweather.tv/topic/99428-model-output-discussion-mid-autumn/?do=findComment&comment=4938667

[WYorksWeather]

Once more I'm going to do a 12z GFS vs ECM comparison, and look at rainfall accumulations over the near term given the interest from Storm Babet.

The GEFS is showing that tonight's 12z GFS sits somewhere near the middle of the pack except at the very end where it goes somewhat colder. A fairly eye catching +15C isotherm ensemble (P3) is worth keeping an eye on to see if it gathers any support over the coming days, but most likely it will fade away. ECM ensembles aren't out yet. 

At T+72, both models have a major low over the UK after the initial impacts from Storm Babet in the next couple of days. Some variation around where the worst impacts will be as a result. GFS focused more over the south, ECM closer to more north-eastern areas.

 

At T+120, both GFS and ECM have a low to our south west, though ECM makes more of it. The result is that temperatures are close to average on GFS under a south-easterly but possibly a little above average on the ECM with more of a direct southerly influence. Likely still quite showery. 

 

By this point, accumulated rain totals from GFS and ECM are very high. Some differences in location as mentioned earlier but certainly few areas seeing less than 40-50mm and many areas seeing 100mm or more. Of course the highest totals to be found in eastern Scotland as per the weather warnings issued today.

 

Moving on, at T+168 GFS already has another low encroaching from the west. ECM is a bit slower with the low but it is looming in the Atlantic.

 

By T+240, GFS goes on to pass a deep low right over the top of the country, whilst ECM at least keeps the core of it a bit to our west. Still a pretty miserable weather pattern though in either case.

  

In the extended, the GFS finally starts to lift the heights a little, possibly allowing the south to dry out a little, but still low pressure dominates further north.  Turning colder right at the end at least in terms of 850hPa with more of a north-westerly flow. Most likely, this is due to a strong Greenland block which reaches 1050mb. Nothing overly cold though for the time of year, perhaps a little below average, as low pressure will typically bring cloud and hold up night-time minima. Of course this is a long way into FI and hence not to be taken too seriously.

 

Overall though, it's fair to say that the next 7-10 days is looking pretty miserable. Which is brilliant timing, because I've booked a week off work 

 

Original post: https://community.netweather.tv/topic/99428-model-output-discussion-mid-autumn/?do=findComment&comment=4939089

[Cambrian]

Going by the 0z ECM ensemble mean, stuck in this seemingly relentless battle between the Atlantic trough and the heights to its northeast, it looks like the brief opportunity at day 7 for cutting through the Atlantic trough with a link-up ridge from Iceland, around the back of the UK / Ireland low, that might eventually lead to a cut-off and bring us into something colder and drier…

 

…is scuppered by the low heights established over northern Canada, around the Hudson Bay / Baffin part of the world, reinvigorating the trough by day 10, reconnecting it with the low at its southeastern end and preventing its escape into Europe. 

 

The 0z ECM op is similarly accommodating to the escape but also leaves a bit more leeway for evading recapture…

 

bringing us into an easterly supported by maintenance of those heights around Iceland, extending further south towards the UK and Ireland.

 

More importantly perhaps, is the decidedly southeastward extension of the trough from Labrador into the mid Atlantic, as it gets squeezed along the western flank of the heights and gets disrupted - a very similar southeastward trajectory as the trough on the chart for today…day 1 / day 10…

 

except by day 10, the trough is lying a good deal further southwest - the action is south of Greenland and looking to be heading down well to the southwest of the UK and Ireland (the day 7-10 animation shows this nicely)

giving more of a chance for the colder, drier conditions to consolidate for a few days.

The day 1 and day 10 setups are synoptically very similar, bar just a few hundred miles, the kind of difference that can be masked by the means and the weekly means, but potentially provide a very different weather type for the UK and Ireland - and if realised at day 10, a very different feel to the end of October. 

Original post: https://community.netweather.tv/topic/99428-model-output-discussion-mid-autumn/?do=findComment&comment=4940005

[Rain All Night]

The ECMWF seems to have again expanded its repertoire of machine learning models to include one of its very own, which it is calling the "AIFS" (the formal name of the regular ECM model being the "IFS"). We can now view outputs from this model alongside those of the other machine learning models we have already seen from Nvidia, Google and Huawei. My understanding is that the reason the ECMWF is able to run so many of these models at once is that they require vastly less computing resource to run than traditional models.

https://charts.ecmwf.int/chart-set/Latest forecast?base_time=202310210000&productNames=medium-mslp-wind200&productNames=aifs_medium-mslp-wind200&productNames=fourcast_medium-mslp-wind200&productNames=graphcast_medium-mslp-wind200&productNames=pangu_medium-mslp-wind200&projection=opencharts_north_atlantic&valid_time=202310310000

You can see that at 0z this morning, all were showing broadly similar outcomes for the UK for Hallowe'en in ten days' time.

For comparison, a more familiar Z500+MSLP view of the regular ECM model output shown top left above:

As a reminder, the way that these machine learning models differ from traditional atmospheric models is that instead of trying to "fast forward" the state of the atmosphere by using (approximations of) the laws of physics, they use re-forecasts of historical weather going back decades (which are produced using the regular ECM model, meaning that that model is still needed) to "learn" how the atmosphere works, and then use the patterns they have "learned" to try to "fast forward" the current state of the atmosphere.

Original post: https://community.netweather.tv/topic/99428-model-output-discussion-mid-autumn/?do=findComment&comment=4941081

[Kirkcaldy Weather]

So looking at what's happening in the next 10 days approx first the transition to the negative PNA which will factor greatly into the overall synoptics

Taking the setup from July 1st - July 15th

Why am I using these dates to compare with the pattern through late October 

The strength of this negative PNA matches nice with the 1st part of July, obviously some teleconnections will be different as was in July however taking the standard negative PNA setup 

️ Blocking High to our East Northeast

️ Increased high pushes Northeast through the east---->Northeast US

️ Alaskan High strengthens with the East Asian trough associated Jet to its south 

️ West coast cut off low and troughing 

On 17/10/2023 at 17:23, Kirkcaldy Weather said:

Indeed. Fast forwarding to Oct 24th we really see this switching from positive to negative phases as the Alaska trough flips to an elongating high as discussed previously this will connect with our Scandinavia / Northeast blocking and the opposites also occur as the United States west coast high pressure shifting to a trough and finally a bit of respite for the North East US toward the maritimes with the trough switches into a building high which carries the exact orientation as the previous trough from the Gulf of Mexico

Our trough will also connect energetically to the west coast trough at least briefly 

 

This switches the recent west coastal heat into a cold plunge with the heat transferred East 

I've enjoyed researching the PNA further extending more information than the bits I knew already and having found multiple papers there's a clear connection with the PNA and upstream blocks either Scandi Greenland or a mix of both which stems from Rossby Wave Breaking in most scenarios as we are and have witnessed lately 

With the stalling MJO still sat between phase 8 and 1 we are already seeing how this sets the next evolution into reality

First the PNA switching 

As the outlook chart shows by Halloween the PNA rises as Current signals have a huge cyclonic and upper level trough signal moving East southeast across Canada 

This is likely linking with our configuration into November with further energies from it feeding into further Atlantic RWBs with renewed unsettled and most probable STORMY patterns prevailing 

  

There is however an opportunity which could come if a RWB causes a cut off low to go south whilst the blocking to the Northeast connects across Greenland and the Alaskan high 

We will see the Russia and extension across Siberian high begin to develop too and an eye catcher from the EC which ramps up the Strat warming as that gets going 

As this cyclonic trough over Canada develops through Halloween we are seeing the initial warming which I had discussed as a genuine possibility  

Full house with the stretching first 

A number of output attempt an October Split which is madness

Amounts increase at 50hpa

Original post: https://community.netweather.tv/topic/99428-model-output-discussion-mid-autumn/?do=findComment&comment=4941251

[Tamara]

Uncertainties, in general abound. Not least, the superimposed forcing exerted from excess stratospheric water vapour following Tonga continuing to add such uncertainty to controlling influences through late autumn into winter patterns

What is known is that Pacific feedbacks have changed considerably over the mid autumn period with the strongly -ve PDO reversing towards a neutral position. In tandem with continued tropical convection standing wave signals close to the dateline. This means less Nina-esque resistance to the downstream pattern, so a likelihood of amplified Atlantic troughs & secondary features running around it, continues to consolidate further heading into late autumn. Where blocking structures set up based on this highly meridional jet stream configuration will no doubt be of much interest as winter approaches.

Overall the background as described maintains an unsettled, at times very unsettled, theme with little to shift the anchor of the upper trough away from its present position. Modelling in extended periods has been trying to point to a flatter more traditional autumn pattern but there is a tendency for this to come unstuck in closer time periods. This accounts for some of the longer term erratic signals in ensemble products.

Based on the PDO switch lifting its suppression on global windflow propagation of AAM anomalies between tropics & extra tropics, and the citing of tropical convection feedbacks in the Pacific, the sense is that angular momentum is on a more linear rising trend heading towards the back end of autumn and this contributes to the interesting implications for the longer term. Irrespective of personal preferences & purely from a synoptic diagnostic point of view.  

Babet, Aline & today newly named Bernard, clearly evident from this satellite imagery heading towards the Algarve here in Portugal).........

.....have been features of this looped very southern tracked branch of the jet stream accelerating SE out across the Atlantic from Newfoundland bringing high impact conditions to the Iberian continent - and then as very well known, being sucked northwards in this loop towards NW Europe, and impacting UK especially, closest to the anchor of the upper trough.. 

Bernard is set to repeat his predecessors, moving quickly northwards across central Spain tonight, through France tomorrow, and being incorporated within the gyre of the upper trough close to UK tomorrow

Long term: How longer term developments evolve in the polar stratosphere are going to determine how this persistent split energy pattern plays out, especially as seasonal wavelength changes under this particular background of macro scale tropospheric pattern, as described above, are not conducive to an organised vortex. 

 

 

 

 

 

 

Original post: https://community.netweather.tv/topic/99428-model-output-discussion-mid-autumn/?do=findComment&comment=4941435

[MattH]

The developments are becoming very interesting in terms of the evolution of the N Hem pattern as we push deeper into autumn. Obviously, we are weeks away yet from the onset of winter and, as is always the case, the primary watch through November is on the development of the stratospheric polar vortex and how this interacts (or doesn't) with the tropospheric polar vortex. It should come as no surprise at all that if there is an organised sPV which then links and connects to the tPV into December there is usually only one outcome and that is an enhanced westerly, zonal flow, this is the "standard" evolution of the N Hem pattern in early winter. The likes of the MJO etc can help to alter this, a good example being the first half of December last year, but that is always the first port of call to keep analysing as November progresses.

The +IOD event remains a troublesome feature mind, amongst favourable patterns and evolutions but we will just have to wait and see on that. Despite AAM having been persistently -ve for numerous weeks, the recent and consistent WWB near and just west of the dateline now seems to be having some impacts. We've seen a pronounced uptick in global MT and the recent tendency plot is 'off the scale', but does usually get watered down as each update progresses, but the broader development here is noteworthy. This links to what Tamara was mentioning with regards to finally a more gradual, linear increase in AAM and, importantly, this may progress and arrive at a time that has implications on the early winter patterns and the sPV too.

The two dominant features are the -ve 850mb wind values near 60E, linked to the +IOD and now the more semi-permanent +ve 850mb winds (WWB) near and just west of the dateline, of which according to the Carl Shreck data is a noteworthy equatorial Rossby wave event. As we all know the MJO has been 'dead' for some considerable time now, the last noteworthy event way back in spring, clearly this links back into the evolution of the summer as well. It will be interesting to see how the MJO develops through the rest of the year as, eventually, a more pronounced development and event should make an appearance and that will also be key to some influence on the early winter pattern too. Generally speaking, the MJO is often more active during the winter months, than it is during the summer.

 

As highlighted numerous times by Tamara over the years, a more Nina-like (-ve AAM) regime in winter is one that often supports more amplified mid-Atlantic ridges, which can build towards mid-latitudes. Amplification of this type, in winter, is often associated with inflated Azores high and with the usual tPV in play, which usually means a general +ve NAO pattern. There are also some links here to this stopping or inhibiting wave activity and flux that is required to help disrupt the sPV as well. The key thing towards winter is that a more atmospheric progression towards a Nino pattern (decline of easterly trade winds) means that the AAM pattern will lead to a greater risk of easterly anomalies at higher latitudes, bringing about a more amplified jet stream pattern.

It is interesting to note that such a pattern has been in play of late, with a more dominant W'ly regime at lower latitudes, blocking, E'ly signal at higher latitudes. This is also why, over the next 7 to 14 days we see a dominant cyclonic pattern but one which is often set to see a more southerly tracking jet with high pressure trying to dominate at higher latitudes. Clearly, November blocking especially Russian and Ural blocking can have significant impacts down the line as well.

The importance of the E'ly AAM anomalies at higher latitudes, compared with W'lies is exampled well if we look back a year ago on the below plot through Nov and into Dec - Note how through late Nov and into Dec, last year, the 50-60N region saw -ve E'ly anomalies dominant, being propped up by W'lies through the sub-tropics, overall it is this kind of AAM pattern that signals amplification and higher latitude blocking. If you then follow through last winter you can see how that pattern reversed with increased W'ly momentum being added to the mid-latitudes into the New Year as AAM fell to more negative territory - the key thing here, is that this year we could/should see the opposite occur with AAM rising.

All in all, with the Nino/eQBO combination in play, some interesting signals already for a weaker vortex and, importantly, a likely progression towards a more +ve AAM pattern then the foundations remain firmly 'set' for a potentially interesting few months ahead, etc.

Cheers, Matt

 

Original post: https://community.netweather.tv/topic/99428-model-output-discussion-mid-autumn/?do=findComment&comment=4941765

[Mike Poole]

One thing I’ve noticed on recent runs at about 10 days is that many of the GFS runs look to be heading towards default with a substantial lobe of vortex taking shape in its more usual location upstream.  Less tendency for this to show on the other models, so I wonder if a bit of GFS bias is showing here..  

This morning’s 0z runs illustrate well, first GFS at T240, lobe of vortex gathering to our west.

Whereas GEM, ECM and JMA show more blocking and a more disturbed vortex:

Original post: https://community.netweather.tv/topic/99428-model-output-discussion-mid-autumn/?do=findComment&comment=4941980

[Cambrian]

Some crucial changes early on in the run determine the ultimate path of the 0z ECM op and lead to some striking differences in the setup by day 10. 

Focus has been on the resilience of the block of Icelandic heights and the equally determined Atlantic trough working around the south of the block into the UK and Ireland. 

It’s Scandinavia where the sleight of hand is happening though, in the push of low heights south between days 2 and 3. 

 

With bridgehead secured, the Atlantic trough transfers the low to the base of the Scandinavian trough at day 5, which pulls it northeast along the trough through the UK and Ireland during the weekend, day 6. 

 

The eastward movement of the low, combined with the resolute Icelandic-Svalbard heights, secures us a northerly flow for days 7 and 8, 

 

before a shallow ridge gives way to repeat of the process of Atlantic cyclogenesis and transfer of the resulting low pressure, another deep one, to the base of the Scandinavian trough during days 9 and 10. 

 

The animations are great value, showing the change for the UK and Ireland setup from Atlantic trough cyclonic to an ongoing transfer of low pressure systems between the Atlantic and Scandinavian troughs. 

 

This should serve to bring us better breaks and colder interludes between the systems as we move into November. Have a great day.

Original post: https://community.netweather.tv/topic/99428-model-output-discussion-mid-autumn/?do=findComment&comment=4941997

[Rain All Night]

Since the model outputs themselves are depressing at the moment, I'll post something else of interest. The ECMWF site is now publishing verification stats (only in 3-month aggregated form for now, covering June/July/August) for their regular model alongside the four machine-learning models whose outputs they are currently publishing.

https://charts.ecmwf.int/products/plwww_3m_fc_aimodels_wp_mean?area=Northern Extra-tropics&parameter=Geopotential 500hPa&player_dimension=area&score=Anomaly correlation

Legend
IFS = regular ECMWF model
AIFS = ECMWF's own machine-learning model
FourCastNet = Nvidia's machine-learning model
GraphCast = Google Deepmind's machine-learning model
Pangu-Weather = Huawei's machine-learning model

This article explains some of the basics around these models, including the key point that they are vastly quicker and cheaper to run than traditional atmospheric models - you can apparently run any of them on any normal computer that has a graphics card (GPU), provided you have access to the ECMWF's licensed starting data (which these models all depend on to be able to run), with the outputs available in minutes rather than hours - no need for a supercomputer. This opens the door to the possibility of ensemble forecasts generated with thousands of members rather than just dozens.

https://www.ecmwf.int/en/about/media-centre/science-blog/2023/rise-machine-learning-weather-forecasting

This second article explains that the ECMWF's machine-learning model uses the same approach as Google Deepmind's machine learning model, which is interesting as these appear to be the two top-scoring models.

https://www.ecmwf.int/en/about/media-centre/aifs-blog/2023/ECMWF-unveils-alpha-version-of-new-ML-model

You can view a selection of actual forecast charts produced by the machine-learning models every 12 hours here:

https://charts.ecmwf.int/?facets={"Product type"%3A["Experimental%3A Machine learning models"]}

Original post: https://community.netweather.tv/topic/99428-model-output-discussion-mid-autumn/?do=findComment&comment=4942239

[Don]

2 hours ago, Metwatch said:

=

Us humans are going to evolve into marine mammals at this rate, albeit less intelligent!

Original post: https://community.netweather.tv/topic/99428-model-output-discussion-mid-autumn/?do=findComment&comment=4942338

[Rain All Night]

On 21/10/2023 at 20:30, damianslaw said:

Hazard a guess but I suspect if you averaged out pressure anomalies for last week of Oct, it would show a mirror image of what is in prospect, i.e. low deep heights over the UK. Settled weather in late October always seems hard to come by. 

I've just plotted the average MSLP for the last week of October, for each of the last ten Octobers, using the tools on https://psl.noaa.gov/, and while there are a few surprises here, in each of the past three years since 2020 there has been an Atlantic trough in play. The only year where it looks like there was an unqualified settled week for the UK is 2016.

Tool link: https://psl.noaa.gov/cgi-bin/data/composites/comp.day.pl?var=Sea+Level+Pressure&level=Surface&iy[1]=&im[1]=&id[1]=&iy[2]=&im[2]=&id[2]=&iy[3]=&im[3]=&id[3]=&iy[4]=&im[4]=&id[4]=&iy[5]=&im[5]=&id[5]=&iy[6]=&im[6]=&id[6]=&iy[7]=&im[7]=&id[7]=&iy[8]=&im[8]=&id[8]=&iy[9]=&im[9]=&id[9]=&iy[10]=&im[10]=&id[10]=&iy[11]=&im[11]=&id[11]=&iy[12]=&im[12]=&id[12]=&iy[13]=&im[13]=&id[13]=&iy[14]=&im[14]=&id[14]=&iy[15]=&im[15]=&id[15]=&iy[16]=&im[16]=&id[16]=&iy[17]=&im[17]=&id[17]=&iy[18]=&im[18]=&id[18]=&iy[19]=&im[19]=&id[19]=&iy[20]=&im[20]=&id[20]=&monr1=10&dayr1=25&monr2=10&dayr2=31&iyr[1]=2013&filenamein=&plotlabel=&lag=0&labelc=Color&labels=Shaded+w%2Foverlying+Contours&type=1&scale=&label=1&skip_vector=&cint=1.6&lowr=985&highr=1025&istate=0&proj=Custom&xlat1=30&xlat2=80&xlon1=-60&xlon2=40&custproj=Cylindrical+Equidistant&level1=1000mb&level2=10mb&Submit=Create+Plot

Original post: https://community.netweather.tv/topic/99428-model-output-discussion-mid-autumn/?do=findComment&comment=4942445

[Kirkcaldy Weather]

I posted the following September 19th 

What could unfold as we go into October?

These are the biggest development dynamical situations I'd be expectant to be occuring. Hazarding a date keep eyes around later of October Week 2 sometime around 10th - 13th

Elongation of European Heights start a gradual movement evolving to Scandinavia---Greenland 

New trough emerging close by most probably a smidge to the East then drifting to the UK 

I'm noticing that trend of low pressures through the United States trends show that eastern coastal Noreaster style as a possible setup

And the following 500hpa animation on 8th of October 

 

New 500hpa sequences following that period which go excellent to my thoughts back in September 

 

'Active West Pacific signals [super typhoon Bolaven]

Californian >> Mexican cyclonic pattern. Major Hurricane Lidia plus Tropical Storm Max. Also the potential next Rossby Wave via Bolaven into new California Cyclonic conditions 

Build of pressure through the UK and Scandinavia which transpires as the Rossby Wave within Atlantic continues its progression further East'

Nothing short of spectacular the manner in which this has really energized the setup pretty much globally 

Using the above JMA phase 7 expectant setups it even gave good indication on the -PNA east American high plus the dual zones over on the western end more on this in a minute 

Really stunning showcase of how a positive PNA swaps to that of a negative phase

"an upper trough trying to align on a negative tilt which might see a severe weather event into the Midwest and west central America as that has the energy from Norma which did indeed move into Baja'

Focusing more at the California Baja >> Mexican troughing this really kicked into life from phase 6 but in particular 7 and 8 we've seen Lidia Max Norma and Otis within like 2 weeks.

Crazy.

Next focus is with the separate trough swinging through the Northwest and Canada the recorded temp differences from AVG are truly incredible already with upto and over 30 DEGREES BELOW AVERAGE

 

 

As the PNA rises this will coincide with these values moving east southeastwards with nearly all of America below average and decently at that, I expect a good number of further minimum temperatures will be at and setting new records particularly in traditional warmer states such as Texas and along the Gulf. Might be some high temp records doing the same within that -PNA high initially however the minimum values really are the main news.

A rare pattern by the very beginning of November in regards to a legit cut off Greenland High, a VERY rare beast bringing back the 09/10 vibes. Note the separation of energy has occurred as our block from Scandi has split with the formation further east connecting through Alaska and the Pacific Northwest with the rest over Greenland as discussed and something I mentioned on October 8th

'I'm eagerly awaiting the approach into Halloween  be looking particularly toward the patterns getting retrogressive in nature with scope for further blocking which looks to have the opportunities of origin towards Scandinavia'

Nice viewing as the composites are starting to show in alignment to the continuing phase 8-1 snailpace MJO movement 

We'll have to keep very alert for week 1 November of the forming CAG with very early signals this has possible connections for the PNA moving once again to positive territory, Florida maybe Gulf states need to be mindful with formation currently signalled by November 4th

This lead up to Guy Fawkes gives another date for cyclone development for us also around 2nd - 6th 

Original post: https://community.netweather.tv/topic/99428-model-output-discussion-mid-autumn/?do=findComment&comment=4942927

[Cambrian]

The fundamentals of the outlook do not appear to have changed for many days now, the UK and Ireland remains at the convergence of Atlantic and Scandinavian troughs at their southern ends, the troughs being kept apart further north by a resilient pocket of heights around Iceland. 

The waxing and waning of each of the troughs is determining their relative influence on our day-to-day weather, with the range of outcomes being dependent on these timings.

Keep an eye on the heights of that zone of the Atlantic to the south of Greenland, and also those over Scandinavia and the Baltic Sea. 

We go from a weaker Atlantic trough, as here on the 0z ECM op for today, supporting a low pressure in the Atlantic, loitering to our west, but making only slow eastward progress. Indeed, the low is to some extent rather left in limbo by a lack of trough engagement at either end.

 

On to day 2, the Atlantic trough is reinforced but this is more than matched on the Scandinavian side,

 

with the result that the low pressure is stretched west-east at day 3, the eastern end now firmly over the UK and Ireland, the low Atlantic heights to the south of Greenland giving way…

 

…with the Scandinavian trough winning this first tussle and the low pressure being handed over, by day 4.

 

Already though, the Atlantic trough is reinvigorating, so that by day 6 (that blue rabbit is in full flight!), the trough is extended all the way from the Canadian Arctic to the southwest of Ireland and the UK. 

 

The surge of cold air southeastward out of Canada meeting with warm air from the Gulf, sparks up the jet stream. There are still low heights and temperatures being maintained over Scandinavia and through the North Sea, so that there is a 24 degree difference between the T850s over the mid Atlantic and those over Scandinavia.

 

The heights to the far north and also to the south of the Azores are squeezing the stark temperature contrast into this mid latitude corridor, and the outcome is some powerful cyclogenesis to the southwest of the UK and Ireland around day 7 and the stormy outlook for day 8 as this second low is handed over to the Scandinavian trough and in the process passes straight through our islands.

  

It doesn’t end there. The fundamentals of heights to the far north and the two troughs converging over the UK and Ireland is still the overriding signature in the 0z ECM ensemble mean charts for day 10.

 

To my good friends on here, I’m afraid it’s all well and truly in a rut, so have a good few days and stay warm and dry. 

Original post: https://community.netweather.tv/topic/99428-model-output-discussion-mid-autumn/?do=findComment&comment=4942645

[WYorksWeather]

Usual 12z GFS vs 12z ECM comparison.

On the 850hPa ensembles, we have a broadly average GFS mean with the operational generally below average, ECM just generally close to average.

 

As yesterday, having a look at the MSLP charts, there is increasing support for a rise in pressure on the GFS 12z, but there is still a lot of scatter, and the rise in pressure doesn't take place at least until day 10, so we still need to be patient on this.

Looking at the OP runs, at T+72, we have low pressure very much in control, and more low pressure waiting in the wings to our west. Similar to yesterday, GFS is much more bullish on the Greenland high compared to ECM.

 

At T+120, both GFS and ECM develop a very complex low pressure setup with three separate deep lows as part of a broader area of low pressure in the Atlantic. GFS gives an early hit to the north with one of the lows at this stage, whereas ECM generally has the system centred further west. GFS and ECM are both continuing to build a Greenland high, GFS much more so than ECM.

 

Stepping through more slowly given the storm risk - at T+144, GFS has the low sitting to our south-west on the previous frame deepening and giving a direct hit to southern England. ECM instead merges all the lows into one and gives a hit more towards NI, NW England, and Scotland. Hence, too early to be sure about impacts in particular areas at this stage. Regardless of location, any landfall over the UK would almost certainly mean a Met Office named storm.

 

At T+168, GFS has the main low clearing away towards NE Scotland, while ECM keeps it mostly anchored over the UK. Naturally, in both cases it is weakening through landfall.

 

At T+192, GFS sends another low through the Channel, possibly enough for a second named storm, though of course a small change of track could see it miss the UK. ECM follows a similar pattern, but is more of a direct hit.

 

Skipping ahead to T+240, GFS fills the low, but of course a rise in pressure comes from such a low base that it is still very unsettled. ECM is slower to clear the second named storm candidate. Possibly the only bright spot is that there does seem to be some sort of attempt to get the Azores high going, but it's still a long way from the UK and not in a position to change the pattern at this point.

 

The GFS extended finally starts to show something of a pattern change. The Azores high begins to build in, but then a deep low smashes into Greenland and completely destroys the Greenland high, and we end up with a Scandinavian high, and an end to the relentless series of lows. What we end up with is a complex battleground. The wind direction charts show wild swings, with milder south-westerlies interspersed with colder northerlies and north-easterlies. The position at T+384 is shown below.

I'll simply allow the rainfall accumulation charts to T+240 to speak for themselves. The forecasts for E Scotland remain quite concerning.

 

 

Original post: https://community.netweather.tv/topic/99428-model-output-discussion-mid-autumn/?do=findComment&comment=4943271