ukpaul

Mere semantics, you know what he means. He isn't suggesting the models have an independent intelligence! The models are 'struggling' then, in that the data is coming up with wide variations from run to run. That much is clear. ECM and GFS being very different in the medium range suggests nothing else.

Looks like a chilly 12z, those heights rising towards Greenland paying dividends later in the high res., expect FI will go back to default but you never know!

Perturbations are useful when they are looked at as a whole and in comparison to the operational run. Sometimes the op goes off on its own merry little way and the ensembles are a useful corrective.

The strong PV spinning speedily round an Arctic low isn't going to give you northerlies, you need pressure to rise there, maybe from a warming of the stratosphere for which there's a chance in the new year. A strong PV otherwise means you might have to rely on something coming from the east. Please correct me if my self taught understanding of this is astray!

Inconsistency/struggle/whatever, it's just ways of explaining that it's not easy to trust the output beyond a short time frame at the moment. Look at John Holmes pdf post from earlier which explains this simply and very well.

You can follow the individual perturbations coming out before the graphs are issued, via meteociel, I just sampled a few and they look very varied.

A model that shows very different things from run to run cannot be deemed 'right', GFS is fun in giving us four runs a day but it enable people to pick out 'their' run too easily. It's akin to forecasters who say one thing, then say another as a 'revision' and claim they are right afterwards because they managed to cover each possibility! That run a few days ago which showed storm after storm hitting us until the middle of December and eviscerating the Netherlands, where is it now, for example?

GFS starting to struggle in the last couple of days veering from run to run, the problem is its determination to find lows which are then downgraded vastly or disappear altogether. ECM appears to be more consistent and credible in the medium term in my opinion,

Well if that GFS run comes off as is there's probably a lot of damage going to happen with low after low whipping fierce winds around the south, I'm thinking I should leave repairing that fence for a few weeks, it'll only get blown down anyway....

It's got cross model support in a reasonable time frame both are showing a cold snap for next Sunday-ish to Tuesday-ish (from others such as BOM and JMA too), after that and further into FI and things diverge but those are only useful for general impressions anyway. Not going to be much of an effect in this neck of the woods but should be good for many. Here in the South East I wasn't expecting much until mid-Jan so happy for anyone else to get what they can while they can, personally I'm looking forwards to our 'autumn in a week'.

850 Ensembles 18z - they look a bit Jackson Pollock-esque to me, up down, here, there, everywhere. London - http://www.wetterzen..._London_ens.png Aberdeen - http://www.wetterzen...berdeen_ens.png Manchester - http://www.wetterzen...chester_ens.png

In the high res looking more like the ECM but GFS tries to throw a low across the south later which I'd be surprised to see remaining in future runs. ECM looks more convincing to me but maybe this is GFS's halfway house on its way to the actual result.

Well well, take away the GFS's determination to push through those lows and we get ECM's first chillfest of the season, wouldn't last much more than another 12 hours beyond 240 but certainly of interest for the whole of the UK now. Both sticking to their guns but will GFS or ECM be the ones forced to back down?

GFS with a good degree of consistency with its previous run, maybe a little tardier at bringing in the last low pressure next weekend. FI totally inconsistent as you might expect and destined for the bin (ensemble runs are a mix of allsorts, although I tend to look at GFS FI and downgrade the persistence and severity of its lows). ECM at 192 and 216 was very different, is it going to stick to its guns? EDIT: GFS op skimming along the top end of temperature in FI as expected (well, for this neck of the woods) Ensembles 12z - London http://www.wetterzen..._London_ens.png Ensembles 12z - Aberdeen http://www.wetterzen...berdeen_ens.png

The models need to be looked at rather than comparing to some memory of last year, compare the output now to previous years and you realise that there is no typical winter. Looking at the current models without emotion and there is a clear pattern change, more for the north than the south but it is there. If you look ahead to next Saturday then you can see another disagreement between GFS/ECM and that's one to look for. (I'd presumed that talking about models in general rather than specific output was acceptable, is that not the case?)

ECM stalls the low that it and GFS have to our North around the 5th Dec, agreement between them before that for similar lows crossing on Mon/Tue and Wed/Thu next week. Will be interested to see if one comes into line with the other over the next few days.

Mega FI but today's 6z CFS Daily from mid December onwards looks fun.

Had a lovely walk out in the fog earlier. It was positively Victorian out there.

Not just an outlier for London, it's sailing along with 850s higher than any other ensemble member for the last five days or so. (PTB 8 is fun to look at by the way)

Support from about 4 or 5 runs for similarly cold 850s, that's a percentage improvement on the last few days but it all depends on the euro high being finally obliterated.

Barely any support for it from control and ensemble members. Edit: Op on the mild end after the 27th, for this neck of the woods at least.

GFS ensembles are wintry on the whole, much more so than recently I would say. We're still too far away for December but the end of November looks up for grabs.

Hmm, not sure what to make of the ECM 12z, slows down the lows but they are still inexorably approaching us by 240h. GFS says quick, ECM says slow.

Quite a few nice autumnal overnight temps. there, maybe December's going to be the new Autumn......

You can't just ignore the bits of the runs that don't fit. We know that FI might well be different, but then again, it could become more snowy for that region too.