General Model Output Discussion:

[Yeti]

No the GFS didn't call it right until the final hr. The GFS had the LP so far N that my max temps today would of been 7C. However my max temps will be around 3/4.

As for the data issue may I ask did you follow every run of the GFS last xmas and this xmas. If not then to be honest you can't comment. I did and the modelling was very odd indeed but its funny how on the 27th normality resumed.

Sorry but some of the comments on this thread this morning have been very bizzare indeed. The quest of looking for 20ft drifts is causing members to miss the bigger picture. I shall explain why when i've returned from my shower.

Last Christmas I don't remember it going back to normality on the 27th, in fact I believe people were still talking about missing data right into the New Year!

Anyway I can't understand either why there seem to be so many disappointed people! The 6z is still an excellent run, and anyway, it will be wrong!

[Methuselah]

Please try and stick to the topic???

[kold weather]

Just ran through the ensembles out to 180hrs and out of 21 ensemble members (the control as well) only ONE run supports the op run, all the others are further south then the GFS op run.

Therefore its safe to assume for now that unless more ensemble members/models come onboard on the 12z suite, its something of a outlier, with only very minimal support for that evolution...

[casparjack]

I don't get all the GFS bashing

They called yesterday's event spot on they said that the front would get as far as the north and it did they also said the snow would be marginal and it was so their data can't be that screwed ATM, GFS was the only model that called yesterday right

I think that the "flight data" is just a caveat used by people when a run doesn't suit their agenda

I have to say I agree, despite variations as the event neared, I feel that the GFS did indeed call the system further North and more marginal than the others - just as it happened in the end...

[nick sussex]

In answer to a PM question i received in terms of why this run won't verify as shown.

Firstly we have to look at the general model guidance as a whole and not take a single operational run in isolation especially being the gfs with a known bias for mishandling shortwaves.

Secondly we should weight the solutions of the ecm and gem above the gfs and ukmo as these have been consistent upto 144hrs with the current pattern, the bias of both the ukmo and gfs initially was to retrograde the block too far west, the point of main uncertainty regarding the block is the strength of this not as was a few days ago where it would end up.

Thirdly NOAA have gone with a compromise solution of the ecm/gem ensemble mean this morning to iron out some smaller scale upstream differences, no mention of the gfs!

Forthly and this is most important and I quote from NOAA

TELECONNECTIONS WITH A NEGATIVE ANOMALY /THE POLAR VORTEX/ NORTH

OF THE NORTHWEST TERRITORIES OF CANADA COUPLED WITH A POSITIVE

MID-LEVEL ANOMALY /A CLOSED HIGH/ OVER GREENLAND FAVORS A

SLOW-MOVING BUT PROGRESSIVE VORTEX NEAR THE NORTHEAST/ATLANTIC

CANADA...

The gfs 06hrs run holds this back and doesn't edge it ne or eastwards, because of this it's run is very suspect.

[Optimus Prime]

I'm pretty sure the LP that was forecast to be further north by the GFS ended up south in realtime. I would have recorded 8c yesterday, instead the maximum was 4.3c.

GFS always overdoes Low Pressures to the south it's a common problem the model has.

[SnowBallz]

I have to say I agree, despite variations as the event neared, I feel that the GFS did indeed call the system further North and more marginal than the others - just as it happened in the end...

I seem to recall reading that "the GFS has backed down at the last minute" (bringing it into line with models forecasting the more Southerly track) So, the GFS seemed to have the mid-term signal right, but the immediate short-term as woeful as the other models. Can't really deny that, because that IS what happened.

[kold weather]

Plus Nick there is the simple matter the GFS 06z op run has very minimal support from its own ensembles!

Anyway some utterly amazing ensemble runs, some would probably lock this country into a very long period where temps struggle to reach 0C...just about the most prolonged cold spell I've ever seen on some of those ensembles for Scotland!

[Iceberg]

Just for info and to put some of this mornings more extreme comments into real life perspective.

The lowest daytime temp for London on the 00Z ECM was 0C. (notice no -2 to -8's ! )

It was an incredibly dry run for London (as were nearly all of the 50 ensembles).

The good news is that it mirrored the ensemble mean very well, also that ensemble mean keeps temps at +3 daytime or below right out to the 14th of Jan.

[kent]

The ECM will verify, the GFS wont etc. The probability is that none of the models will verify as people are seeing what they hope for their own areas. History has taught us that. To avoid further disappointment I really believe that 72 hours hence is difficult enough and as for 216 plus its all about hopecasting.

[Duane S.]

It really is painstaking reading this thread at times.

The GFS, whilst on its own is still a superb run for January in the UK, is a VERY different run and which is unlikely to verify for the reasons given earlier by what must be an exasperated Nick and others. I'm not saying its modlling of the LP in question is impossible, and if the block was to edge West towards Canada then this could be a possible outcome. All eyes on the Eastern Seaboard over the next few days, as long as that trough sits there and moves slowly NE it will stop our GL high migrating away., leaving us a bit less exposed. That said, if we can get the cold air embedded then that very scenario could bring impressive snowfalls to anywhere in the South.

Oh, and, NO... the GFS didn't call it right yesterday, it progged the low pressure system itself much more North than it actually got.

The bottom line is that the charts are sooooo good at the moment that downgrades are virtually impossible to avoid now as we near T+0, whenever that is!

[Guest North Sea Snow Convection]

Just for info and to put some of this mornings more extreme comments into real life perspective.

The lowest daytime temp for London on the 00Z ECM was 0C. (notice no -2 to -8's ! )

It was an incredibly dry run for London (as were nearly all of the 50 ensembles).

The good news is that it mirrored the ensemble mean very well, also that ensemble mean keeps temps at +3 daytime or below right out to the 14th of Jan.

None of those computer calculations will take into account the extra effects of radiation cooling over snow cover for example. That was clearly demonstrated in the pre xmas cold spell - with lower temps recorded in reality than by the models.

So I think you can subtract, realistically a couple of degrees or more from those estimates - IF we are assuming the synoptics verify.

Edited December 30, 2009 by North Sea Snow Convection

[casparjack]

I seem to recall reading that "the GFS has backed down at the last minute" (bringing it into line with models forecasting the more Southerly track) So, the GFS seemed to have the mid-term signal right, but the immediate short-term as woeful as the other models. Can't really deny that, because that IS what happened.

I'm not saying it got it absolutely 100% correct, none of the models did. However, what it did do is carry the system further North than the others (I know it ovedid the intensity of the Low).

Anyway, whatever it's done now - we should be looking ahead, not picking the bones out of yesterday...

[Cecil Hogmanay Melchett]

Anyone else looking forward to the 15day outlook today?

If the update reflects ukmo/ecm then its going to make very interesting reading,hopefully it wont reflect the 06z op run.

[leroys35]

Just for info and to put some of this mornings more extreme comments into real life perspective.

The lowest daytime temp for London on the 00Z ECM was 0C. (notice no -2 to -8's ! )

It was an incredibly dry run for London (as were nearly all of the 50 ensembles).

The good news is that it mirrored the ensemble mean very well, also that ensemble mean keeps temps at +3 daytime or below right out to the 14th of Jan.

i find it strange how can you get a very dry north or north easterly things will change nearer the time i guess

[Iceberg]

None of those computer calculations will take into effect the extra effects of radiation cooling over snow cover for example. That was clearly demonstrated in the pre xmas cold spell - with lower temps recorded in reality than by the models.

So I think you can subtract, realistically a couple of degrees or more from those estimates - IF we are assuming the synoptics verify.

Morning Tamara, they do take in this cooling effect (honest), how well is open to discussion. ECMWF has according to some pro forecasts to much of a negative bias on temps when snow is forecast to be on the ground.

You might be right if London does get snow (but ECM is not progging snow, but keeps it very very dry from the 1st onwards.

i find it strange how can you get a very dry north or north easterly things will change nearer the time i guess

I do agree post 120Z ECM is very bad at not including SW's, troughs etc, but I wouldn't want to assume that snow will be widespread either !.

[casparjack]

This will surely give some 'back edge' snow to Southern areas, not much, but enough to be going on with...

[TEITS]

A couple of points before I shoot off.

First of all the LP system than many are talking about. I said a few days ago that when you have model disagreement the answers often lies inbetween. This proved to be correct because the LP wasn't as far N as some GFS runs indicated but was further N than the ECM suggested. Now don't make the mistake that the GFS was right because of the lack of snow. The lack of snow was due to the marginal temps. In some respetcs I blame the Met O because alongside their warnings they should of said 50% risk of disruption on lower ground. They should add percentages along side their warnings on how possible disruption could be.

Back to the models and I find looking for a severe cold spell in the UK is the same as baking a cake. You need all the ingredients for the perfect outcome but sometimes you can get away with one.

Here are the ingredients we currently have.

1. Negative AO which is off the scale!

2. S,ly tracking jetstream.

3. Block of HP over Greenland.

4. Pool of cold air over the UK (will have from Friday onwards).

Now if like the 06Z a LP manages to get far enough N then we will experience heavy snow due to No4 ingredient. However if the LP is further S i.e ECM then as it tracks E this will pull in even colder NE,ly bringing more snow. However if the block moves further W this exposes us to a risk of the LP moving much further N bringing snow and then milder weather. At the moment this scenario isn't suggested by the models.

I don't need to ramp the potential because around lunchtime the Met O 6-15 day forecast will do that for me!

Just add that the ECM from +168 would not be dry. I can tell plenty of snow showers would develop just by looking at some of those charts. This is why the Met O forecast suggested more widespread snow middle of next week.

Edited December 30, 2009 by THE EYE IN THE SKY

[bluearmy]

Morning Tamara, they do take in this cooling effect (honest), how well is open to discussion. ECMWF has according to some pro forecasts to much of a negative bias on temps when snow is forecast to be on the ground.

You might be right if London does get snow (but ECM is not progging snow, but keeps it very very dry from the 1st onwards.

I do agree post 120Z ECM is very bad at not including SW's, troughs etc, but I wouldn't want to assume that snow will be widespread either !.

do the london temps take into account the 'heat island effect' of the city ??

[Guest North Sea Snow Convection]

Morning Tamara, they do take in this cooling effect (honest), how well is open to discussion. ECMWF has according to some pro forecasts to much of a negative bias on temps when snow is forecast to be on the ground.

You might be right if London does get snow (but ECM is not progging snow, but keeps it very very dry from the 1st onwards.

I do agree post 120Z ECM is very bad at not including SW's, troughs etc, but I wouldn't want to assume that snow will be widespread either !.

Morning - it would be very surprising if there was no ppn on the face value of the latest ECM, with slack low pressure over the continent and a NE drift with very cold air over a relatively warm north sea.

And edit: as you say, there would be at least one or tow minot features at the very least depicted in nearer time on UKMET faxes etc I would think.

[Upgrade]

do the london temps take into account the 'heat island effect' of the city ??

No . The heat island only works in some weather situations. Strong winds tend to wipe out the effect. Strong enough North Eest winds would render the UHI effect insignificant.

[Snowly does it]

06z - major snow event (away from the South Coast) followed by a prolonged freeze retaining snow cover.

Other ensembles = colder as the LP does not get as far north, less snowy as the LP does not get as far north.

Be careful what you wish for.

[Lewis]

Well put aside the arguments on this thread, it's safe to say it's getting colder !

Edited December 30, 2009 by Storm Force Lewis

[Guest North Sea Snow Convection]

To close current posting, I would say that the ECM and other model output is much more preferable to the GFS 06z route. Not just for my region (I live only 10 miles now from the south coast) but anywhere south of the M4 they will be preferring the low further south, and most importantly on the basis that a colder evolution arrives on a reload NE'erly - that in itself will generate more snow opportunities for the good of everyone more nationally too. So a possible snow to rain event (which could grow over wider areas northwards in time), plus one which wouldn't then neccesarily go on to show the later synoptics of the GFS 06z is not desirable imo. No just for the south, but butterfly effect elsewhere too.

Fortunately at this time, the latest GFS 06z has very little support

I know what I am wishing for!

Edited December 30, 2009 by North Sea Snow Convection

[Cloud 10]

Very little support for the 06z operational run at the +144 timeframe.

Looks to have the low to far north again,which is a very similar situation to what we had a few days ago.