I agree they are not great but why does it point to the end of a cold spell and why should ECM fall into line?

The key timeframe in all of this is mid next week. The 00z has an insane amount of scatter after the 23rd with a 15C+ difference in 850s!

I am sure the ECM won't fall into line. If the GEM was showing this you wouldn't bat an eyelid yet I'm pretty sure the GFS is performing worse at the moment!

Really poor overnight runs from the GFS, it's ensembles and the UKMO. Awaiting the ECM but if it follows, then this one has gone and I would imagine Exeter would be revising their outlook later today. 

Don't think they'll be that hasty, somehow  Nothing stopping an improvement in the evening runs. Swings and roundabouts and it is to be expected in such a complicated pattern. You'd think FI would be at a pretty early timeframe at the mo.

The GFS has been consistent in not following the ECM line, if the developing Low runs through the corridor of death as the GFS and UKMO show, then it completely changes the dynamic for the rest of January. 

Our old friend Pete Tong has come to stay in place of His double-barrelled relative - Auntie Azores-high - judging from the GFS, GEM, JMA, and Uncle UkMO and all this morning. Today's modified Arctic plunge apart, Polar Maritime breakfast, lunch, tea and dinner. Cool uppers sufficient for a Burt Ford Forecast - snow on hills. Just the FI tease of Jam in February...from the reliably teaseyy GFS, 

Hopeful ECMF will keep the faith when the current run is fully out.  It's been the reliable model of late IMO.

The UKMO only goes out to 144hr and is similar to the ECM at that point,the GFS is already deviating. I would wait for the rest of the run before jumping to conclusions as it may also indicate what would follow on the UKMO output.

GFS consistent? 

Exactly 🤣🤣🤣🤣🤣

Rotten output this morning, following that ECM run from yesterday which had us all excited. Was ever thus in the model output world.
As The Who put it, we always get fooled again.
ECM t+144 this morning is a massive downgrade on where it was in yesterday's 12z run.
The golden rule is this - never believe stellar charts for UK cold until they reach t+96 and even then, cross fingers in both hands and hold your breath for 4 days!

P.S. ECM may be a downgrade on its previous runs, but it is still a million miles away from GFS and much better than GFS. GFS has completely lost the plot and as others have said, it's reliability must have been compromised by the US state shutdown. The Artic signature from ECM is nothing like the GFS at t+144, nothing at all - and I know which horse I would back in that race and it is not a "G"

Worth noting nobody is on duty to fix the GFS!

GFS

I think Ian means consistent in showing something different to the ECM, rather than model consistency (a humourous thought indeed) 

Yes - a simple rule that some forget year on year. 

For the record here is yesterday 144hr chart (156 is not available in this format)

Not what I expected to see this morning. I figured we were close to a tipping point yesterday and was quietly confident a long cold spell was setting up. This morning the output (I need to look at in more depth to be sure) appears to have firmly come down on the side of a more mixed and less cold picture. In the shorter term it looks as though GFS has made a much better job of the disturbance running in from the west on Thursday/Friday. ECM has been all at sea. I've got to say that ECM gets all the plaudits in the US because it tracks hurricanes more consistently. However, closer to home it is utterly hopeless in potentially blocked scenarios.

This rotten stuff you reference is still a couple of days or so into your FI (agree it's around t96/120 right now) so we will continue to see changes. Now the ECM has reached that point we can see there isn't agreement yet. One only needs to look at the GEFS 00z to see how scattered it is beyond day 5.

Yet still much better than the GFS recently. 

Agree completely. 

Pulling my hair out on the other side at all the posts slagging off the GFS and lauding the ECM. The ECM often over simplifies blocking scenarios in our patch of the world. For all the GFS faults, once it picks out a spoiler, it's usually correct in doing so. 

Statistically for the globe then yes. However, in our small part of the world ECM has been rubbish with the GFS a lot better. I genuinely think that GFS handles the North Atlantic profile more consistently than ECM. It would be interesting for someone to research it because obviously my comments are anecdotal. 

How can it handle things better when it often shows a different scenario every six hours?!

I guess one will be right. Up to about Day 4 or 5 I have not noticed much of an intermodal difference.

My observations are that in the medium term neither have been consistent but the GFS has been more erratic jumping suddenly as  opposed to moving from one scenario to another.

Edit: In the meantime there is excellent agreement on very cold (sub 520dam, circa -10° t850hPa) unstable air moving into the north this evening. FI is 48hr 

Would you not expect more scenarios to play out as it doubles its exposure twofold over the ECM that only runs twice a day? 

Why was it not next stop February last night? Has model accuracy improved so much in 12 hours?