Strider

The 18z is Lettuceing awesome for cold prospects compared to the 12zs... with such a dramatic difference though it can't be taken seriously until we see the esembles.

This is the first time I can remember GFS and ECM showing a pattern change simultaneously. Until it gets within T144 and starts showing on UKMO I won't be taking it too seriously though.

I would. I think it's pretty clear he was talking about relative temperatures and not literal ... ...

Despite the difference in temps I'm sure 08/09 is more likely to be remembered for the February major snow event.

We're going to need the Azores to move northwards in late Dec to have any hope of a cold shot chaps... GFS is tentatively showing this out in FI at the moment. Can't see anything else of interest.

I think Jan/Feb/March 2006 too! Over a winter you would expect to see snow falling at some point. But for major snow events with accumulations of above, say, 5cm is an exceptional event IMO (for most people)

January 2011 was average in terms of temperatures and hardly anyone saw snow

You don't even have to go that far back, if you look at the GFS run on this day 2 years ago it looks pretty similar http://www.meteociel...=2009&archive=1 The weather remained unremarkable until mid-December and it was a cold winter.

I have a bad feeling that GFS's evolution will lead to a Bartlett (obviously I'm not basing that on extreme FI) ECM's is infinitely better which isn't saying much tbh. I think UKMO may be trending slightly closer to ECM but that might just be wanting it to be the case...

Most sensible post I've seen. The evolution back to zonality could herald a change back to windy and wet weather, but nothing especially cold. That's not a great chart for cold lovers.

Can I ask why ECM is more likely to verify than GFS? Just because it shows a cooler evolution is not an acceptable answer and is misleading to inexperianced forecasters reading this thread GFS has very strong esemble support, and besides ECM isn't *that* good anyway - it just leads back to the atlantic.

Regarding ECM's LRF, is that ran at a higher resolution than their normal model? I mean we wouldn't take their T240 as gospel so shouldn't we be a little cautious with this output? ... ... Just sayin'

The irony is the current block over eastern europe which is keeping our weather mild is helping the snow cover to build up faster It's better to have the right synoptics at the right time so I'm entirely comfortable with this. Looking at the models it looks as though we may have to wait until mid-month (a week or so) before Scandi gets a proper pasting

Interesting, I would've thought the same amount of data went into each run. On the link you edited into John's post you talked about different amounts of flight data per run. Do you have evidence to back this up?

Unfortunately Meteociel's archive facility doesn't seem to be working for me atm, can anyone else use it? Anyway I think a degree of patience is required here. Remember, winter doesn't START for several weeks. There's nothing to suggest below average temperatures at present. I feel the first half of November will be very warm (with strong model agreement) with the second half of the month average and wet with slow moving troughs from the Atlantic.

Yes, the best possible solution would be for a mid-Atlantic ridge leading to significantly cooler conditions than we have been experiencing of late.

I was under the impression (certainly from the Met Office) that the official start to winter is indeed Dec 1st. Dec 21st is used by older folk

The right run at the wrong time for coldies, although that said it does help to move the cold air mass into Scandi which could be crucial later on. Also good to see Metociel have the runs coming out at the proper times now

The forecasts aren't based on luck but probabilities. If I say it's going to snow on January 24th then that *is* based on luck as I have no theory to back it up. But proper forecasters make assumptions based on current climatic condtions and connections. Obviously the further away you are predicting the more tenuous those links become, reducing the probability of accuracy. Unfortunately a lot of people see forecasting as black & white so forecasters get a lot of unfair criticism due to ignorance of statistical likelihoods.

Funny there's no posts about FI tonight... hopefully a lesson learned for some. If you base your predications on T300+ you'll have a heart attack before December!! Looks like the usual autumn fare to me, wet and windy for most with Atlantic dominated weather.

This is yesterday's T384 from the 00z - were you saying the same thing yesterday, or just today because it's showing cold? http://91.121.84.31/modeles/gfs/archives/gfs-2011102300-0-384.png?0 I'm not saying it can't happen, I'm just saying that it's not really of note until it gets within at least T200

Knew as soon as I saw T384 it would be referenced in this thread! Remember just because FI shows cold doesn't give it a shred of credence In the reliable timeframe there is little to get excited about as we appear to be entering a phase of westerlies.

Yes, people say January was warm because it was virtually snowless for most. I have to say I'm fearful for this winter. I can't see it being as interesting as the last few atm.

Superb charts across all 3 models. It's really hard to believe considering the output a few days ago was so dire..! Can't help but wish it was 11th Nov though... are we coming too early?

It's not *that* cold... and UKMO chart is showing similar.