Yes, that’s a fair question. Although the same could probably go for the other models that are issued 4 times a day. Observationally (with no statistics at hand to back it up) I would suggest that even with an extra couple of runs a day the GFS op runs rarely get a “head start” on the ECM runs.

Sorry Neil - I don't - I only had this bookmarked to my favorites and thought of having a look.

Meanwhile - i see the Daily Express saw the 12z Operational run:

Dr Claire-Kennedy Edwards??

https://www.express.co.uk/news/weather/1388636/UK-Snow-forecast-weather-warning-met-office-snow-charts-winter-ice-London-weather-latest?fbclid=IwAR2ikQIltS8GNpvqJhlmPkI83_g3xDWpigGD4CtmSmOTIJ4qkxFlOQ2ufhY

What happened to their 20" of snow?

For me I personally think the ECM twice a day is more consistent. I find 4 updates a day to have wild fluctuations and ultimately eventually returns similar output to those which run x2 a day

That’s not to say there aren’t many examples where this is wrong but on the balance of probabilities I don’t think 4 runs a day are required?

Para's doing it's best to save the situation.....

I knew the 18z would swing mild - or even worse the exact opposite. Getting a bit of a deja vu now me thinks with flip flopping this even 06z showed a northerly post 300z, the 12z showed an easterly with scandi High post 300z with low pressure over southern Europe and most recently the 18z showed a south westerly post 300z with high pressure over southern Europe? What next for the 00z run? We have seen every possible scenario from azores high, northern blocking, to barlett high - just couldn't make it up could you!

I guess the Express would have to re title their front page to "Heatwave to approach the UK"

At least the 18z Op was a milder option of the GFS Ensembles. The para and control has potential - Control shows an easterly post 300z. 

GFS Control pretty eye catching too, and the GEM is still showing a cold outcome - so anything past 144 hours is speculation right now: Clearly eye candy runs are still plausible because they keep popping up, so, as they say, we will see: Feb has promise.

The forecasters are struggling to figure out the Northerly progression of the milder air over the next few days, so predicting what's going to happen in 7-10 days time in the current set up is extremely difficult. 

Small tweaks in the set up in the short term could lead to significant differences down the line. I would be sceptical of the MO output beyond 3-4 days at present. 

Just a glance at the model verification stats would tell you that accuracy starts to fall away quite quickly beyond 5 days. At 10 days and more the operational runs are generally pretty useless.

They were still changing their forecasts for Sunday’s snow just 12 hours before - and still got the detail wrong!

I try not to take notice but from a weather enthusiast and from a researching perspective it is frustrating because they (the models) are so inconsistent this winter - maybe it could be down to the SSW not sure. We are all use to seeing occasional flip flops but ... From a northerly to easterly and now to a south westerly - we have seen all options and wouldn't surprise me if tomorrow morning's 00z Control run flipped mild but the 00z operational run flipped cold again yet the para had southerly tracking lows and the ECM was somewhere in the middle. But yes totally agree with you!^_^ 

Tell me about it! a day or so before there was uncertainty about the extent of the snow for Sunday and on Friday the forecast was for the snow to affect Mostly wales, south west England and central and southern England when it took a more northerly track and the main focus for the snow was actually in the midlands and the south esp coastal areas was in the warm sector and got cold rain, sleet and wet snow at best with any accumulating snow was falling away from coastal areas of the south and over hills/moors.

Back to here and now - it's not all gone there is still hope - but we have seen this last week when the parallel model was going for an easterly - we all know what happened with that and a few days ago the control run which flipped mild and then flipped cold again (seen in tonight's 18z) and of course the Operational run which flipped from cold to mild south westerly (as seen in tonight's 18z).

I won't give up hope - we have all of February if nothing happens this weekend and there are still some sub -10c uppers in the run also there could be another SSW too but a rise of 25c @ 10hpa instead of 45c. - I think this is the main reason why the computer models are struggling.

18z ENS for London:

A lot of scatter despite the milder Op run the ENS mean is still remaining colder than average from 2/3rd February.

Indeed. All the forecasts, and indeed most of the main models showed the main belt of snow wouldn’t reach as far north as me. And even now, most of the snow depth charts showed no laying snow where I am straight after the event.

These are truly very complicated situations to resolve currently, for a global location that always seems very close to the borderline between mild south westerlies / westerlies or freezing north / North easterlies/ Easterlies , on an island in the middle of a warmed ocean and at the edge of a continental land mass.

4 or 5 days has been relatively reliable timeframe for forecasts and models to deal with in recent years. But in this situation everything could change on the ground for Britain’s location by a sheep farting in New Zealand.

All we can rely on (cling our hopes on) no matter what our best wishes are for, be it freezing cold or mildness is trends and signals for anything beyond 3 or 4 days

Model following is like betting on a 10 runner race where the tipsters all predict a different winner.The quiter activity on here suggest punters have all done their money in on this winter.

I think we should focus Gem as it seem doing well and consistently for a long while now.  

Jetstream strong and over or just to the S of the UK with only slight meandering to Wed 3rd . More chaotic with big loops appearing Thu 4th, Mon 8th and Thu 11th in each case placing the loop to put UK on the cold side

http://www.wxmaps.org/pix/prec4 main feature is the amount of rain of the western coast of Europe for the next two weeks; earlier promise of cold from the east on the companion chart has come to nothing though it remains very cold in extreme N Europe

GFS - a rather messy picture with LPs bombarding the UK from Scotland to Cornwall, passing through quite quickly in each case. A  bit more action when one works its way into the N Sea Mon 8th 985 mb with following N-lies; and the traditional end-of-run tease on Thu 11th with a large and deep area of cold LP covering the N half of the UK.

GEFS - around the seasonal norm to Thu 4th, above and below in the S but mostly below in the N, after which so much variability as to prevent any trend being picked out (maybe cooler for a while???). Rain to be expected on most days in the S, more clustered in the N around Thu 28th and Tue 2nd with drier interludes later.

ECM - like GFS though pressure gradients associated with the LPs tend to be smaller

Im putting all my faith in the GEM and I don't see how the model discrepancy regarding this weekend can continue too much longer - although FI has at times come down to t72, so I would say a resolution of the general pattern affecting the UK by 30 January (Sat) will be achieved by tomorrow morning 27th, at the latest.

One gets the feeling Sods law will operate here, but you never know.

Will be same old for me whatever model is right😉 

My guess from looking @ the models cold will never be far away, and their will continue to be snow chances for central and northern parts in particular, as has been the case for the majority of this Winter.

It been solid for very cold easterly to come in early Feb and now we are only 5 days left to exit Jan and it have to happen at some point.  I am with WI post too.

The op runs may be jumping around but the GEFS for the first third of Feb have been very consistent in recent updates.

https://www.theweatheroutlook.com/premium/gefs-tables.aspx?run=00&fv=tmp850&loc=London&lat=51.5&lon=0&p=&cty=

Absolutely just for fun, I'll take the end of the 00z GFS v16 run...