Model Output Discussion - Late October Onwards

[shedhead]

Agreed Mushy, really no point in trying to see things that aren't there. GFS does flirt with rising pressure over Scandi, but inter run it's very inconsistent, whereas ECM continues to maintain the same overall pattern day in, day out. I'm not suggesting we won't see a switch to cold/very cold weather later this month, just like I never ruled out a switch back to cold/very cold weather last Jan/Feb, but until I see some evidence of it I will remain as sceptical now as I did then.

[summer blizzard]

What a stonker from the GFS0z in terms of dry and potentially frosty weather late in the run (good inversion likelyhood there).

Seems to be a classic model war now with the past 3 runs supporting opposite trends.

GFS0z was not an esemble outlier.

[chionomaniac]

What a stonker from the GFS0z in terms of dry and potentially frosty weather late in the run (good inversion likelyhood there).

Seems to be a classic model war now with the past 3 runs supporting opposite trends.

GFS0z was not an esemble outlier.

That to me probably suggests the whole lot of GFS ensembles are wrong then!

[Radiating Dendrite]

The 0z run was right at the top of the enselbles for air pressure.

I think it was an outlier in this respect.

[ANYWEATHER]

Agreed Mushy, really no point in trying to see things that aren't there. GFS does flirt with rising pressure over Scandi, but inter run it's very inconsistent, whereas ECM continues to maintain the same overall pattern day in, day out. I'm not suggesting we won't see a switch to cold/very cold weather later this month, just like I never ruled out a switch back to cold/very cold weather last Jan/Feb, but until I see some evidence of it I will remain as sceptical now as I did then.

Yes, gfs over the last few weeks has been portraying high pressure over towards the East/NorthEast on a good many runs and it has not verified. So im taking all that with a large truck full of Salt. For the short term "normal" weather for the time of year that likely to extend into the medium ,longterm, although it looks like a brief respite for the NorthWest of the Uk over the coming weekend to early next week.....

[Nick F]

Quite clear the differences between GFS and ECM as we head into the medium range looking at the 8-10 day H500 comparisons for the 00z run this morning:

As others mention, given the operational was at the higher end of the pressure scale against its ensemble members - especially for the north but also down south too, we may see GFS operational head towards ECM's less settled towards the NW solutions on subsequent runs:

Aberdeenshire:

London:

[MP-R]

A very interesting model run but everything is a month early! An unsettled week this week but if the GFS is to be believed then a drying up again next week and gradual cool down as we pick up winds off the continent with quite a northerly component.

And there was me hoping for a proper Atlantic spell.

[shedhead]

A very interesting model run but everything is a month early! An unsettled week this week but if the GFS is to be believed then a drying up again next week and gradual cool down as we pick up winds off the continent with quite a northerly component.

And there was me hoping for a proper Atlantic spell.

The above highlighted word being the all important one. Nothing I've seen so far today suggests GFS will be right, in fact I think ECM's more consistent approach to a generally mild, unsettled outlook is far more likely to verify at this stage.

[wolvesfan]

Hello all, ECMWF 00z to be the model most likely to verify from what the top three are showing at the moment,nothing realy to support a sustained rise in pressure to our north.on the other hand pressure to our east well and truly anchored.any ideas as to whats going to break its back anyone?.

[nick sussex]

Although there are still differences between the big 3 models the UKMO and GFS have already edged towards the ECM by reducing the strength of that ridge in their overnight runs.

Looking at the average of the ECM ensemble solutions you'd have to think its more right than the others.

The actual spread though does show that its ensembles are widely split.

http://www.meteociel...=1&map=0&type=2

The pressure spread shows this clearly.

I think we'll know by this evening but either way theres no sign of anything wintry for the foreseeable future, the best given the poor teleconnections is for some faux cold if the ridge can hold on far enough north but the overall northern hemisphere pattern is not condusive to anything other than that for the timebeing.

If you put aside the differences in the eastern Atlantic between the ECM and GFS and look upstream over the USA then they're basically the same with a flat jet and until there are major changes there then this Euro high/ Atlantic trough stalemate will remain.

http://www.meteo.psu.edu/~gadomski/ECMWF_0z/hgtcomp.html

Edited November 1, 2011 by nick sussex

[BLAST FROM THE PAST]

The current model outputs bare no resemblance to the synoptics that we saw during November 2010 (check link below to archive), I'm not sure how you have come to make a pretty bold statement like the one above. You may be right, but we won't know for 3/4 weeks.

ECMWF says no to the GFS blocking setup and brings in the Atlantic in later in the run.

http://www.meteociel...12&map=1&mode=0

Exactly, models will toy with the idea for a while yet and a retreat/stalemate shown by ECM is precisely what I'm on about, no change for a while. It won't IMO be until late Nov for the switch. Now bearing no resemblance who said anything about the synoptics being like last year? What we may well see is a switch from the mild conditions to very cold fairly quickly when it happens and my timing is late Nov [last ten days say]. I think models are tentatively showing [too early] what may be the way forward down the line. Its not a bold statement, just what looks the way forward on my forecast method at this early stage.

BFTP

Edited November 1, 2011 by BLAST FROM THE PAST

[OldMetMan]

Interesting to see the GFS sticking to its blocking pattern in today's 2 runs so far. I'm inclined to agree with others here that the ECM might be nearer the mark with a more indeterminate block, or with westerlies resuming. But I do feel that a more meridional flow seems likely in the coming weeks so I think it unlikely there will be prolonged unsettled conditions.

Quite apart from anything else, it's been a long time since we've had any major, persistent HP anywhere close by so it's about due. I think the longer range UKMO later today will be the one to watch in terms of how and where the blocking pattern will develop.

Incidentally, the slow-moving cold front over us in the next couple of days could cause some significant rainfall problems - depends where it gets stuck.

[ITSY]

GFS certainly wants HP to dominate come FI - it will be the next few days IMO that sees either the other models fall into line with GFS (WITH MORE POTENtial for cold) or GFS drop back into line...one to watch

[Glacier Point]

If we think the atmospheric circulation has 8 broad phases, phase 1-4 representing a La Nina type and phases 5-8 an El Nino type, the anomaly pattern over the Northern Hemisphere since September has been largely representative a phase 1-2 type transition. This is consistent with the onset of La Nina conditions in the tropical Pacific.

The ensemble mean H5 GFS and ECM products are broadly in agreement days 6-10, the ECM perhaps a little less amplified in the Pacific but essentially both are continuing to advertise a phase 1-2 type pattern for the medium range period. This is characterised by a ridge in the Pacific, downstream trough digging into the western half of North America consistent with a -PNA pattern, and a trough over the Atlantic with downstream ridge centred over Scandinavia.

What both models also show in the ensemble menas is an anomalously low height over the eastern side of the Arctic and generally average or slightly above average westerly flow around 60-80N. This is preventing the type of evolution that the extended GFS outputs are hinting at.

I have a problem with the GFS in particular for two reasons:

1) its tropical modelling is biased (there is a biased correction model output even) towards an over-progressive eastward movement of tropical convective waves, which will invraiably try to build heights over the North Atlantic and Greenland over time in this current set up;

2) the evolution for most Novembers under developing Nina conditions (aka phase 1-2 type pattern) is for some height rises to the north, and I suspect that this is setting a default for the GFS beyond day 10.

The issue is that both are not nailed, the MJO forecast by the ECM to be weak and the height rises to the north being hampered by the polar westerlies descending down through the upper atmosphere. That's not to say that these wont happen, and the deeper into the winter the greater the pressure for these factors to give, but for now, stick with the ECM.

[snowlover2009]

The way this is heading, the La nina will end up being weaker, than last year, with how the forecast members have flipped. The grey and blue ones are the new runs!

http://www.cpc.ncep.noaa.gov/products/analysis_monitoring/lanina/images3/nino34SSTMon.gif

I think we could be odds on for a major shift from very mild weather to very cold weather in the coming weeks. Watch this space. Something similar to November/December 1981. Looking in the archives, you would have never have guessed the December coming would be a record cold month. The pattern just suddenly shift. I don't know what you guys think.

I am not saying a repeat of last year of when the cold came early, but some forecasters are expecting the cold to come in the second part of winter. If this is the case, judging by the new cfs link above, I think the cold could come early as next month.

[Backtrack]

The way this is heading, the La nina will end up being weaker, than last year, with how the forecast members have flipped. The grey and blue ones are the new runs!

http://www.cpc.ncep....ino34SSTMon.gif

I think we could be odds on for a major shift from very mild weather to very cold weather in the coming weeks. Watch this space. Something similar to November/December 1981. Looking in the archives, you would have never have guessed the December coming would be a record cold month. The pattern just suddenly shift. I don't know what you guys think.

I am not saying a repeat of last year of when the cold came early, but some forecasters are expecting the cold to come in the second part of winter. If this is the case, judging by the new cfs link above, I think the cold could come early as next month.

Do you have any other information other than perhaps a slightly weaker La Nina?

There's no ensemble support to suggest anything remotely wintry - even for Scotland stretching right to the end of the run:

Even the long range CFS model has very little cold, with just a 4-5 day North Westerly feed around the start of the month, which quickly returns to above average conditions:

To add to that, I just can't see any real blocking being established until at least mid-January.

Not what many people want to hear, but I'm strongly suspecting that December COULD be a complete write off.

[Frosty.]

The Gfs 06z shows something of interest for those who live in the nw of the uk, a building azores ridge pushing into nw britain during the weekend and bringing a settled and sunny spell with light winds for a change with some overnight frost and mist/fog patches whereas the south and east of britain would be more unsettled under a low pressure influence, the fine weather would persist in the north well into next week with a slowly improving picture further south across the uk. The 6z is a very anticyclonic run on the whole and hints that the main high could begin to retrogress towards mid atlantic at the end of the run which could leave the door ajar for a cold blast after mid november (being very optimistic here) but the gfs is so unreliable that the next run could be atlantic dominated in FI. The Ecm 00z is generally a more unsettled and milder run than the gfs with spells of rain and strong winds, especially for western and northern areas but if the 6z was correct, it would buck the trend of the last 6 months with the best weather in the northwest and the dross in the southeast for a change!...but the southeast does need the rain and the 6z would bring welcome rain for those areas...anyway, a very mild/warm start to november with temps around 6c above the seasonal average across the south but tending to cool down by the weekend from the northwest as that ridge topples into northern britain.

[johnholmes]

To add to that, I just can't see any real blocking being established until at least mid-January.

Not what many people want to hear, but I'm strongly suspecting that December COULD be a complete write off.

and what were you posting this time last year?

For those worrying-needlessly I might add as no one not even Mr Corbyn, have much idea what will happen longer term, take a look at the charts for early November and switch to 20 November>

Have a look at early November 1962 or 1946.

Stop worrying and chasing snowbeams on every run. The weather will do whatever it wants no matter how hard we try and forecast what it should do.

Mild is the key word for a while yet.Nothing in the 500mb charts out to mid November to suggest otherwise apart from the prob of a surface high possibly developing over or near the country giving, depending on cloud cover, some frost and fog overnight.

AO and NAO in the fairly reliable time frame out to 15-20 days are not suggesting much in the way of cold.

MJO and 30mb temp=wait and see at the moment it seems to me?

Regarding the 30mb temp profile. It is my own view with no science to really back it up, other than it has happened on about 3 occasions in the last 3-4 years, IF the temp rises rapidly from where it is within the next 2-3 days then it MIGHT signify colder weather 20-25 days down the line-I said MIGHT and IF?

Edited November 1, 2011 by johnholmes

[Backtrack]

and what were you posting this time last year?

For those worrying-needlessly I might add as no one not even Mr Corbyn, have much idea what will happen longer term, take a look at the charts for early November and switch to 20 November>

Have a look at early November 1962 or 1946.

Stop worrying and chasing snowbeams on every run. The weather will do whatever it wants no matter how hard we try and forecast what it should do.

Mild is the key word for a while yet.Nothing in the 500mb charts out to mid November to suggest otherwise apart from the prob of a surface high possibly developing over or near the country giving, depending on cloud cover, some frost and fog overnight.

AO and NAO in the fairly reliable time frame out to 15-20 days are not suggesting much in the way of cold.

MJO and 30mb temp=wait and see at the moment it seems to me?

Honestly last year I didn't understand the models enough to understand what was needed for snow. It was only until about December I was starting to properly understand I think.

Now I need to work out how to understand how to understand those 500mb anomaly charts and I'll be golden.

[Frosty.]

and what were you posting this time last year? For those worrying-needlessly I might add as no one not even Mr Corbyn, have much idea what will happen longer term, take a look at the charts for early November and switch to 20 November> Have a look at early November 1962 or 1946. Stop worrying and chasing snowbeams on every run. The weather will do whatever it wants no matter how hard we try and forecast what it should do. Mild is the key word for a while yet.Nothing in the 500mb charts out to mid November to suggest otherwise apart from the prob of a surface high possibly developing over or near the country giving, depending on cloud cover, some frost and fog overnight. AO and NAO in the fairly reliable time frame out to 15-20 days are not suggesting much in the way of cold. MJO and 30mb temp=wait and see at the moment it seems to me? Regarding the 30mb temp profile. It is my own view with no science to really back it up, other than it has happened on about 3 occasions in the last 3-4 years, IF the temp rises rapidly from where it is within the next 2-3 days then it MIGHT signify colder weather 20-25 days down the line-I said MIGHT and IF?

Agreed, does it really matter about a cold snap in november..I don't think so although I would like a similar event to late last nov but last year we had all the ice and snow in late nov and on and off through december and then the second half of winter was a huge let down for coldies, i would prefer the cold to be spread out more liberally during the main winter months this time around thanks very much!

Edited November 1, 2011 by Polar_ Low

[kold weather]

GFS looks at odds with the ECM, though both do suggest the PV weakens a little by the 8-10th of November and re-positions. That would argue that a change in the pattern becomes possible between day 10-12 and if high pressure can become stronger over Scandianavia it could well begin to have an effect.

ECM still is very dominant with the pattern we've had for the last month.

[feb1991blizzard]

I never knew that you could view the CFS on a frame by frame basis, ive looked all the way through until 1090 and there is nothing showing, you would think that far out there would at least be some potent cold showing at some stage, still we will just have to wait until the next run and see what happens!

[Paul]

The raw, daily CFS is not bias corrected - you need to understand the model and it's inherent biases a bit more if you want to read into it's output, and even then you need to view it for trends over many days/weeks to really start to glean anything..

[Thundery wintry showers]

Central and eastern parts of the British Isles will have a sunny and quite warm day tomorrow, ahead of the fronts, and then the rain looks set to extend slowly eastwards across the whole country on Thursday, followed by sunshine and showers on Friday as the returning polar maritime airmass finally establishes (but only for one day).

The weekend will have a north-south split, most parts look set to be dull, and there will be rain across the southern third of the country, but the GFS outputs suggest that the weather may turn much brighter across Scotland by Sunday with a fairly widespread ground frost and then sunshine by day. The majority of the outputs are then suggesting an anticyclonic/southerly type into next week, meaning most central and eastern parts will be generally dry, but temperatures and sunshine amounts are sensitive to the precise positioning of the high to the east and the lows to the west (whereas in summer this would be a near-guaranteed "hot and sunny with thundery outbreaks" setup).

[feb1991blizzard]

The raw, daily CFS is not bias corrected - you need to understand the model and it's inherent biases a bit more if you want to read into it's output, and even then you need to view it for trends over many days/weeks to really start to glean anything..

Is there anywhere you can read about its biases with regard to forecasting UK weather or is it just a case of viewing its output over a long period and getting a feel for it like with the GFS for example over deepening lows and sometimes being to progressive.