General Model Output Discussion:

[TEITS]

There is no trend in the ensembles, it's a complete mess.

If you look at each member then you will always find some disagreement. This is why its much better to look at the mean. I always use the GEFS/GEM ensemble means at +240 and beyond.

I will add using the GEFS mean proved very accurate during Dec. I distinctly remember you dismissing the chance of a cold spell in mid Dec!

Edited January 26, 2010 by THE EYE IN THE SKY

[cooling climate]

true ennough KW. even the extended ens underestimated the length of time the current block would hang around to our east before either retrogressing or sinking. it seems as though the ops are too progressive to get to the next step. its then not a straightforward delay to the evolution shown as by the time the block begins to move, other factors may not be conducive to evolve to what we currently see in fi. so whilst a scandi high/greeny high/ mid atlantic high/ uk high looks to be likely, the area where the block will end up is a fairly large spread. as a result, we could just as easily end up with a repeat of the past fortnight in nomans land as a frigid easterly.

CC, you're new to this i guess !!

I only joined Netweather last November but I have been watching the models and viewing this and other forums for

several years. I do feel certain from what I see (charts, teleconnects, MMW etc, etc) of an impending very cold and

prolonged spell of weather and surely there is no harm done in stating that. Even the more hesitant can see there are real possibilities of this occurring.

Edited January 26, 2010 by cooling climate

[cuckoo]

Riga was -21C at noon yesterday!!!

I noticed the area i went on holiday to Poland in autumn was -28°C this morning.

If the easterly in FI was to follow the northerly (and it was a straight easterly) there is certainly some cold surface air available at the moment!

Edited January 26, 2010 by cuckoo

[sandstorm2]

I only joined Netweather last November but I have been watching the models and viewing this and other forums for

several years. I do feel certain from what I see (charts, teleconnects, MMW etc, etc) of an impending very cold and

prolonged spell of weather and surely there is no harm done in stating that. Even the more hesitant can see there a

are real possibilities of this occurring.

Yes,you are quite right posting what you feel and think are right after all that is what this model discussion is all about, the charts that are currently showing must bring with them a bit of caution.

I love cold snowy weather but there are a lot of vulnerable people out there who can't afford to have the heating on, so lets just be careful what we wish for.

SS2

[Country Boy]

This will be interesting if it stays this way

[Frosty.]

Blowing my own trumpet a little here but after banging on about a prolonged cold from the

start of the northerly at the end of the week and then progressing into a east of north a

airflow for about a week now if not more. It is nice to see it coming to fruition and read

correctly the signs that I was seeing from the northern hemisphere charts.

It hasn't happened yet though so hold off with the trumpet for the time being. The N'ly will probably have a stuttering start but friday should bring down the main thrust of Arctic air and continue through the weekend with a bit of luck. As for the E'ly, way too early to be confident about that but no sign of any particularly mild weather in the outlook either.

[bluearmy]

The below quotation is from the ECMWF Ensemble page as posted by GP (I hope they don't mind). I suppose the questions below are directed in particular to anyone who has professional or in-depth experience of how models work.

'In the European Centre for Medium-term Weather Prediction, the operational model is re-run at a lower resolution (the control run) and this is then repeated 50 times, each with slightly different starting conditions. These ensemble runs are shown in the graph above as the grey lines; the average is shown in yellow.'

(my emphasis)

1/ What is different about the starting conditions? Is it the choice of exact data used, or is the data very slightly manipulated by the modelling tool to represent tiny changes in real life which would naturally be happening?

2/ Are the ensemble runs done at a lower resolution purely for speed (i.e. if data is very slightly manipulated anyway, to create the range of possible outcomes, there is no point in having fine tuning)? Is it because of the ways the data is supposed to be interpreted (i.e. saying, "Oo, I wish we could have seen how number 32 got down to -20 deg C for next week" is pointless since it was never designed to be seen - but rather to show the likely spread of possibilities, with the yellow average being the mid-point of greatest likelihood)?

3/ When a model is upgraded (like the recent Parallel GFS becoming the new GFS), what is changed? Actual measured data? Programming within the modelling tool itself?

Thanks for any answers

Nick

how approprauet that you should ask number 3) when the upcoming ecm 12z is the first live run of the latest upgrade. there is plenty of blurb on the ecm website about the upgrade. its mainly to do with a lower resolution in the horizontal axis down to 16km. there are other programming improvemnts though.

wrt the ens, i dont know the answer, but i suspect that speed has something to do with it couple dwith the fact that the starting data has been modifed anyway. i'm sure that studies have been made to see if running the ens at high res makes a substantial difference to the mean outcome across 50 runs. if it did, no doubt there would be investment in the processing power needed to do this.

[kold weather]

Very interested to see whatthe 12z GFS does, but it does need to be remembered that it is a rather fragile evolution upto the point where we get HP over us, once that occurs then things will stay cold hopefully but gettinbg there has to be the main worry, esp if the upper high becomes very stubborn to relocate, which I've seen happen time and time again in the past, but however equally we have seen the evolution the 0z ECM puts forward a good 2-3 times already this winter...

[snowmadchrisuk]

There is no trend in the ensembles, it's a complete mess.

I'd actually like to point out the fact that the GFS predicted -3 to -4C widespread last night, and yet the temperature in many places did not go below 0C, I'm certainly not convinced this GFS update that was brought out in September can handle programmed temperatures, also I'm not convinced it will be able to handle a warm up whenever that occurs.

I thought the GFS update was brought out on the 15th December , unless you mean the test period. I agree the GFS is not good at predicting minimum temps , although it seems to be much better at predicting maximums . I have to say though Stephen , you are quite negative sometimes when it comes to a new patten , you have to look at all the facts and at the moment they are not exactly showing screeming South Westerly's and I can not find any model that is predicting anything but continued Cold really.

[Lewis]

I thought the GFS update was brought out on the 15th December , unless you mean the test period. I agree the GFS is not good at predicting minimum temps , although it seems to be much better at predicting maximums . I have to say though Stephen , you are quite negative sometimes when it comes to a new patten , you have to look at all the facts and at the moment they are not exactly showing screeming South Westerly's and I can not find any model that is predicting anything but continued Cold really.

lol, since i signed up 5 years ago, he's always been a negativist I think it's because when hes negative, he normally does quite well but i doubt he'll admit it haha.

[weather eater]

The fact that it's all in F.I pretty much sums it up. Where as personally i'm the sort of guy that looks at short term and will sometimes ponder deep into F.I trying to look into trends/patterns. The GFS would be bottom of my list when looking for patterns/trends deep in FI.

It did pick up the current evolving pattern before the ECM, again this was the post I made on the 18th.

A big HP ridge gets thrown up across the Atlantic, leading to heights being raised to our north and a situation where the cold air to our east can get activated towards us. Similar evolutions have cropped up again and again in recent GFS runs.

[nick sussex]

A good gfs run so far, importantly the trough remains stuck over Europe, no way of milder air getting in if that remains the case, pressure should rise to the ne.If the shortwave near Iceland heads se then its game on.

Edited January 26, 2010 by nick sussex

[Lewis]

U SEEN THE 12Z WHOPPPPPPPPPPPPPPA

If this verifies, i'll see you in spring!

[cooling climate]

This 12z GFS run is showing a even more progressive way to an easterly, a much colder

run into the weekend as well. Although the GFS surface temperatures do not reflect the

very cold 850 temperatures and are overdoing it I think.

[kold weather]

12z is very good and if anything its an upgrade on the northerly flow and everything a touch further west on this run...

Actually I suspect in the flow the GFS is progging maxes of 1-3C seem very likely and mins well below freezing, flow now holding on through to the 1st of Feb so an upgrade in length as well it seems...

Also this flow will help to keep the SST's over the North Sea pretty cold and also help to drag down the SST's to our north somewhat as well, could wel see a big xhunk of the Baltic frozen up, which means if we get any easterly of substance, its going to have some real bite to it...

[Southender]

Ching, Ching!

I'll bank that one thanks.

Why have I got the feeling that this is going to be an outlier.

Edited January 26, 2010 by LeighShrimper

[Iceberg]

A very big upgrade on the 12Z's GFS out to 168 and UKMET to 144.

UKMET takes more prolonged snow across many areas at 120/144.

GFS brings in a NE to easterly by the 2nd of Feb with NO milder interlude.

[kold weather]

The 12z GFS shows EXACTLY the pattern I was mentioning yesterday as the best case set-up to Nick S, though I have to admit I didn't really think it'd occur but the 12z GFS nails the pattern, as long as we get some sort of NNE/NE/ENE/E airflow then the very cold air dragged down into Europe will come back westwards.

Make no doubt about it, the GFS surface temps would be 3-4C too high IF the 12z came off, you'd easily be looking at widespread ice days in that sort of flow as mins would go very low...possibly just about the coldest set-up without snowcover you can get!

[nick sussex]

12z is very good and if anything its an upgrade on the northerly flow and everything a touch further west on this run...

Actually I suspect in the flow the GFS is progging maxes of 1-3C seem very likely and mins well below freezing, flow now holding on through to the 1st of Feb so an upgrade in length as well it seems...

Also this flow will help to keep the SST's over the North Sea pretty cold and also help to drag down the SST's to our north somewhat as well, could wel see a big xhunk of the Baltic frozen up, which means if we get any easterly of substance, its going to have some real bite to it...

The UKMO looks good aswell, generally good agreement on the pattern but its crucial the trough remains stuck over Europe, regarding the Baltic this is the latest snow and ice map:

http://www.rap.ucar.edu/weather/surface/snowAFnh.gif

Apparently its the first time in many years that we see ice between Denmark and Sweden!

[Southender]

Mammouth upgrade.

I cannot believe that this is not going to be a coldy outlier...

-12 850's into EA by 2nd Feb.

Edited January 26, 2010 by LeighShrimper

[Lewis]

510 DAM close to the East Coast!!

This is one hell of a run. This will certainly bring the CET down a peg

[cooling climate]

The GFS is showing 2m temperatures of freezing but the maximum temps a couple of degrees

higher. I have noticed it do this several times in the past but with 850's of -10c I would expect

maximum surface temperatures no higher than freezing or +1c at the very best especially at

this time of the year.

[snowmadchrisuk]

Nice run , We lose our system on Monday 1st , that would have brought Snow to some but it is a sacrifice well taken IMO .

[nick sussex]

Mammouth upgrade.

I cannot believe that this is not going to be a coldy outlier...

-12 850's into EA by 2nd Feb.

I think you might be surprised, i suspect there will be decent support for this in the ensembles, its not often you get the frigid cold over the continent for this long. The ensembles will still likely have alot of scatter for the key period but there should be a grouping between -10 and -15 and even colder the further south into Europe.

[Southender]

I think you might be surprised, i suspect there will be decent support for this in the ensembles, its not often you get the frigid cold over the continent for this long. The ensembles will still likely have alot of scatter for the key period but there should be a grouping between -10 and -15 and even colder the further south into Europe.

Here's hoping Nick.

I'll bank that run all the way, thank you very much.

Oh and here comes the beast of beasts in deep FI (Again!)

Edited January 26, 2010 by LeighShrimper