Model Output Discussion 12hrs. 01/10/13

[Frosty.]

It's the change in pattern that pleases me more than anything, I have enjoyed the autumnal warmth but I will be very happy if it does turn a lot colder, I love it when forecasters say, it will be turning much, much colder through the next week.

[SMU]

Looks like a long winter ahead of model watching indeed, already the shredder blades are sharpening and a bigger model bin is required. For those reading that aren't familiar with the phrase binned, binning etc It isn't just abuse, the pros over the pond use this term frequently to discard wayward solutions.

 

UKMO and ECM - clearly the one's to watch. I really fear GFS will be referred to with the same respect as CFS this season. Still , I guess we wouldn't have it any other way!

 

Nice post Lorenzo- although for the GFS saying the 06z is the best may be correct for the NH, but is certainly the worst for locations closer to home within that-

Last year we put up the page that showed the results per cycle of GFS for every area & the 3 closest- Europe, Atlantic & arctic I think it was miles away-  but I cant find that link on the performance page as its changed-

 

 

Anyway the models have settled down now with a burst of cold polar air that will get mixed down to the surface before high pressure remains in situ to the NW, certainly good for Ireland & possibly the west for great autumnal but chilly weather-

 

 

One to note as well as Ive made references to it- the NOAA map that everyone rates has been just as bad with the assessments placing the upper ridge to far east- why, to much GFS bias applied probably to the thinking........

 

S

[draztik]

UK Outlook for Friday 11 Oct 2013 to Sunday 20 Oct 2013:

Many areas will be largely dry at first with sunny spells and light winds. However it will be cloudier across eastern parts with some showers, perhaps wintry over Scottish Mountains initially and more persistent rain possible in southeast England. Feeling rather cold, especially across eastern parts. Into the next week, conditions are likely to remain similar at first, although western parts of the UK may become rather more unsettled and breezy, with outbreaks of rain. In contrast, eastern parts will tend to become drier through the week, with more sunshine and temperatures recovering to near-normal across many parts. Thereafter, northern and western parts are likely to become increasingly unsettled and windy, with showers or longer spells of rain. Sunshine and showers likely across southern and eastern areas, with near-normal temperatures.

Updated: 1210 on Sun 6 Oct 2013

source : MetO

Edited October 6, 2013 by draztik

[J10]

Except for the south east where they could see rain on and off for the majority of the output, in fact anywhere south of a line from Merseyside to the Humber could be susceptible to outbreaks of rain over the coming period with the low over Europe pushing bands of rain in from the continent.

For many southern and eastern areas this is a massive downgrade on conditions from potential warm and sunny southerlies to cyclonic north easterlies and outbreaks of rain. Going to be very bleak from Thursday onwards.

The 06Z GFS actually beats lasts nights ECM in terms of awfulness with the low actually re-invigorating at the end of high resolution.

I guess I should get some sandbags in preparation........

That's why I said to me. personally give me North Easterlies over South Westerlies 100% of the time. Better still South Easterlies but we can't have everything.

 

I would imagine western areas not doing too badly if this forecast actually came off.

Edited October 6, 2013 by J10

[stewfox]

I don’t know if this is the right thread but what models would one use in their arsenal when looking and making comments on the future weather. I’m sure a lot of people who just read this thread would like to know.

 

In the last 24/48 hrs we have seen a fundamental change in the out put the GFS .Some seasoned forecaster are calling the GFS an embarrassment. Its  gone topsy turvy with its output changing dramatically..

 

Is the GFS seen as poor because it run  4 times a day or is  just less capable at picking up ‘signals’ and is that real or perceived ?.

 

GFS ensembles , can we use the ensembles as a guide or are there better references ?  I always thought generally speaking you can look at ensembles as a guide but don’t take them as gospel ?? Also don’t jump on every 6 hr GFS update but look at the trends that can be shown in the ensembles is that still the thing to do. ?

 

Is the ECM the one to watch this winter because its done well recently or does just performs much better??

 

Do people use a mixture and include looking at  fax and 500mb charts out to day 15 before posting their views

 

Before we go into winter proper what would people use for reference before making a forecast ? I’m sure there are personal preferences

Edited October 6, 2013 by stewfox

[Captain Shortwave]

That's why I said to me. personally give me North Easterlies over South Westerlies 100% of the time. Better still South Easterlies but we can't have everything.

 

I would imagine western areas not doing too badly if this forecast actually came off.

For those in the north and west of England and Wales you wouldn't want the low to shift any further west or next weekend could be pretty miserable for most of the country. The metoffice already have the South east under persistent rain. So what's to say that persistent rain could effect a wider area of the UK.

Shame the south easterlies on most models 48 hours back have gone.

Edited October 6, 2013 by Captain shortwave

[Frosty.]

Here are some of the highlights from the Gfs 06z op run, much colder air incoming by the end of tuesday from the northwest with a very amplified pattern for a time and we will all be in the cold air by midweek with a lot of showers for areas bordering northern & eastern coasts with strengthening Northerly winds but with sunny spells inland, most of the showers further east and cold enough for snow on the scottish mountains and higher hills and hail/sleet to lower levels in the far north. Later next week, winds become more NE'ly as atlantic high pressure begins to topple across the far north with low pressure to the southeast, the strong NE'ly flow will push a lot of rain further inland and it will feel very cold as a result. On this run, it then remains cold and unsettled across the south/se but high pressure across northern britain brings a lot of sunshine, light winds and cold nights with a touch of frost, eventually the 6z shows a more typical mid october pattern with atlantic lows spreading in and a chance of increased amplification with polar maritime incursions beyond mid month.  ENJOY.

[Cheshire Freeze]

GFS ensembles , can we use the ensembles as a guide or are there better references ?  I always thought generally speaking you can look at ensembles as a guide but don’t take them as gospel ?? Also don’t jump on every 6 hr GFS update but look at the trends that can be shown in the ensembles is that still the thing to do. ?

 

 

 

You can't spot a trend in the ensembles if there isn't one. It's all or nothing- i.e complete agreement for 5/6 runs (you'd think the outlook was fairly certain) and then a complete flip in the next set.

 

In this instance, it would be reasonable to expect that one of the perturbations would have picked up on the pattern for the coming week- there must have been a good chance of this with the alterations to each separate perturbation to account for error. But no, not one picked up the actual pattern which to me suggests that there is a flaw in the algorithms somewhere which is only corrected once the period in question comes into a certain range where the model can't fail to pick up the correct pattern.

 

At 120-168 hours out NWP in this day and age should at least be able to project the general long wave pattern- and if not the scenario should be in the ensemble suite. Neither of these things happened.

 

It could also be argued the CFS (yes the CFS!) did a better job at hinting at the scenario which will play out this week. The monthly anomalies pointing towards HP being situated further N for October than the Euro high the GFS wanted. It's saying something when a long range model is picking out the pattern with more accuracy than its short-medium range sibling.

Edited October 6, 2013 by CreweCold

[phil nw.]

 

 

Some good questions there Stew.(post 456-above)

 

I know Steve M is not a fan of the GFS suite, and that's probably an understatement, but we all tend to remember the rouge runs and maybe overlook the many times when they are closer to the mark.

Yes i have seen the performance statistics too but all models have their moments good and bad.

An example being the changed outlook for this coming week both the ECM and GFS  switched after the UKMO.first picked up on this. 

I try to view all the models including the means/anomaly charts and graphs.

A good idea is to maybe watch for trends over a few runs.

In essence use as much of the available outputs as we can to get a balanced view.

Edited October 6, 2013 by phil nw.

[Jonan92]

Has anyone seen this chart in FI?

 

 

Artic region - 1030hpa

Edited October 6, 2013 by Jonan92

[BlueHedgehog074]

It's looking more and more likely that next week will indeed deliver a cool/cold setup with winds coming down from the North for a time. Some adjustments to next weeks Northerly flow still expected, such as how amplified the High Pressure system out West can really get (as that could have an impact on the potency of the Northerly). Plus, some slight adjustments Eastwards or Westwards of the High will have obviously have some affect on where the main area of the Northerly becomes concentrated over, and also where the coldest of the airflow heads. I think, however, the UKMO has done well for being one of the first models to pick out a Northerly setup.

[SMU]

I just wanted to clear up things with the GFS-  its ok in its place & once in a blue moon it comes up trumps-

 

The problem the UK has ( & obviously then us) is the GFS weakness is in 2 or 3 key locations across 60N, these 3 locales are always where we see our blocking highs develop-

So when the GFS is wrong it tends to be miles wrong-

 

In this scenario it incorrectly moved the low off the eastern seaboard East when infact it goes North at 72

here you see it on the ECM up Greenland-

http://www.meteociel.fr/modeles/ecmwf/runs/2013100600/ECM1-72.GIF?06-12

 

As a result of this the GFS creates to much energy moving east & doesn't allow any southern movement of low pressure over Scandi-

So the net fail over Greenland creates another fail over Scandi -  the GFS went for a flat jet but over the last 24 hours its 'seen' the low towards Greenland better so it changes the trajectory each run-

As a result as the low goes from East to NE to north the jet bends further & further SW across scandi- hence that low projected location moves SW - once it was near Svalbard in the GFS eyes, then in a matter of 24 hours its pulled it back to the North sea-

that's a LONG correction-

 

Its worth noting in this scenario ( & remember this for ALL winter) the GFS ensembles are worse still at over acceleration the jet, so they are often another 24 hours behind the operational in bending the jet to the correct location-

Hence you get whats perceived on here by everyone an the operational being an outlier- when its infact the WHOLE of the ensemble suite that is wrong.

 

 

Also, the third fail from the GFS is its inability to 'see' a highly meridional jet- IE one that bends back west after digging south, - this is VERY clear in winter again when we start to see the building blocks of a scandi high forming in the 144-192 area of modelling-

 

The GFS will always THROW any trough digging through Europe to far East, subsequently the pattern builds over the top in its eyes & the UK doesn't get any cold air-

The ECM & UKMO are much better & quicker at identifying the jet profiles that will indicate westward bending of the jet- hence they pick up the cold quicker.

 

 

This week is a prime example of how if you know the GFS failings you can dismiss it very easily-

 

For those who just 'pop' in here to see what the models are showing have a choice, you can follow loads of people, or just 1, digest all bits of information its up to them- I see a lot praise for posts that provide a balanced view- & that's great if you don't want to go any further- however a balanced view or a lopsided playing field is sometimes futile- as all your doing is wasting your time.

 

People mostly know me, know my posts- I will always dismiss the GFS off the bat if it displays its usual hallmarks of its mega East fast bias-

 

Hopefully people will not confuse dismissing a model for its bias reasons rather than me dismissing it just because I want it to be cold...

 

 

S

[syed2878]

i'm rite last year ian f did say that the met doesn't use the gfs output beyond 5days. I'm sure some1 will correct me if i misquoted ian.f.

[Timmytour]

 

The GFS will always THROW any trough digging through Europe to far East, subsequently the pattern builds over the top in its eyes & the UK doesn't get any cold air-

 Maybe there's too much use of historical data from pre-2009 driving the algorithms :-)

[SMU]

Yes otherwise known as the 'mod-ern era'...

 

They need to reprogramme it with the 'Post mod-ern era....'

 

lol cant believe that word is in the swear filter....

 

S

Edited October 6, 2013 by Steve Murr

[Gaz1985]

No escaping the nose dive in upper air temps by Wednesday, luckily it won't last, back to average by next weekend 

 

[bluearmy]

steve, you're warming up too early - you'll be worn out by the time december has arrived.  in this instance, you are on the money. a few weeks ago it was less clearcut and the solution turned out a little way from where you expected. however, the less experienced winter model watchers on here would do well to follow your thoughts as they are more right then wrong.  i do hope the winter season delivers - it seems to have more anticipation year on year.

[A.J]

 

 

Maybe there's too much use of historical data from pre-2009 driving the algorithms :-)

 

 

Yes otherwise known as the 'mod-ern era'...

 

They need to reprogramme it with the 'Post mod-ern era....'

 

lol cant believe that word is in the swear filter....

 

S

These 2 posts should be pinned in the model thread for all eternity IMHO as a reminder that all model output is the product of complex computer algorithms, and not, as so many members seem to think, that model suites are sentient beings subject to mood swings and 'going off on one'.....Always makes me smile and laugh out loud when I see posts describing models and their 'personalities'....

[BornFromTheVoid]

A change to more seasonal weather looking very strong now.

 

From t0, we begin to see the first important steps. 

First of all, we get the jet stream becoming very amplified and ridging towards Greenland within 24 hours (a feature we were very familiar with in the summers of 2007-2012). You can see the ridge pushing north, starting off just south of Hudson Bay at t0, before moving strong towards Greenland at t24. A strong trough forms in the mid Atlantic also.

 

...................t0 ...................... .................. t24

 

By t48, that ridge extends further toward Greenland, while the trough has almost become cut off from the jet stream. This allows a ridge to build up toward Greenland, just west of the Bay of Biscay.

 

Cut off jet loop t48 ............ ......................... upper ridging t48

 

By t72, the cut off jet loop is clear to see, while the ridging toward Greenland intensifies to our south and west, rapidly pulling the high pressue along our west toward Greenland, putting us on the eastern, cold side of the high pressure.

 

jet loop and ridging t72 .......................... upper ridging to the west t72 ........................... upper air temps t72

 

After that, we see the ridge continue to push northward and eastward, while a new cut off loop in the jet stream begins to form over Europe, introducing an easterly airflow to the British Isles

 

Ridge pushing north and east

 .......... ..t96 ...................... .............. t120

 

New jet stream cut off loop

................. t96 ................................ ......t120

 

 

After 5 days out, it seems that the upper ridge becomes dominant and spell of strong northern blocking takes hold, with remarkable agreement from the 8-10 day mean geopotential height anomaly comparison between the GFS and ECM.

 

 

 

 

Looks like the first proper spell of northern blocking and -ve NAO of the Autumn. We haven't seen a -ve NAO month since March, so perhaps it's time for a pattern change?

[Timmytour]

2013

1962

Some similarity here. Not implying it means anything, just wouldn't mind the same development this time around over the next few months as developed 51 years ago :-)

Edited October 6, 2013 by Timmytour

[Frosty.]

I love these fax charts, they show progressively colder air digging in behind the fronts as they swing southeast across the uk with troughs also swinging south within the early blast of arctic air, the N'ly becomes strong for a while with a very unstable airflow and heavy showers further east and north with showers turning increasingly to snow on the scottish mountains and higher hills, a sleety mix with hail lower down, it's the opening salvo of hopefully many more progressively more potent cold blasts. Later in the week, atlantic high pressure cuts off the cold airflow across scotland and winds fall light but further southeast, the winds back more NE'ly with cold persistent rain, especially for the southeast, if we get these type of charts from late october onwards, there will be a good chance of snow across the uk, even london, as happened on 29th Octoberrrr 2008 when snow fell widely and especially across the south and southeast and left a covering.

Edited October 6, 2013 by Frosty.

[summer blizzard]

Yes otherwise known as the 'mod-ern era'... They need to reprogramme it with the 'Post mod-ern era....' lol cant believe that word is in the swear filter.... S

Given the timings of the modeling change and the low west of Greenland I'd suggest that the GFS was expecting a stronger and barlonically enhanced TS Karen while after genesis the Euro had a much better grip on the vertical structure of Karen which eventually meant that much less energy was going to go into the low over Canada and ergo we see a modeling shift.Yet another example of why modeling from August-October are essentially useless when the tropics are active.

[snowy weather]

Has anyone seen this chart in FI?

 

 

Artic region - 1030hpa

what does that mean?

[johnholmes]

Maybe there's too much use of historical data from pre-2009 driving the algorithms :-)

another misunderstandin I'm afraid Timmy. ALL models take the data at T+00 and run through to the end of their period using complex maths to solve the almost as complex physics equations, there is NO historical data fed in. Write to UK Met, ECMWF and NOAA to see what they say?

[Cheshire Freeze]

2013

1962

Some similarity here. Not implying it means anything, just wouldn't mind the same development this time around over the next few months as developed 51 years ago :-)

 

Virtually identical but probably means little. One can hope though!