And that's exactly what it is ..………..a guessing game , last year was a farce and certainly taught me a lesson lol

I was going to stay quiet this year, as this annual run-in with you is a bit irritating.

You really need to learn how probability works, but suffice to say this year has been quite different in the run up to Christmas model-wise compared to the past four. Work out what that difference is and you'll find out why the models coped well even with the data loss.

Something more seasonal to look forward to 

So different to the 6z 

Yeah, right.

Anyway, I'll explain it more fully for you as you missed the point.

This year we had a remarkably well organised polar vortex. In the weeks running up to Christmas the models were unwaveringly showing a zonal setup with a roaring jet across the Atlantic.

The models handle this very well, as they're designed to - it is, after all the default scenario.

You would no doubt have noticed that this winter the models have shown virtually nothing when it comes to blocking - even in the ensembles, -10 850s and easterlies have been almost entirely absent.

In such a scenario, with such high confidence and high probability of the outcome (zonal), losing even 15% of the data wouldn't make much difference - it might mean instead of zonality being say 90-95% likely, it would be 85-100% likely (i.e. a widening of the envelope, but not enough to make an appreciable difference).

On the other hand, the previous few years have had much lower confidence in a given outcome, "Shannon entropy" as it was called by some. They still leaned towards zonality (the models after all will generally revert to the climatinc mean), but the chance of that was much lower than it was in the run up to this Christmas. In all those years, -10s and so forth were far more common in the ensemble suites.

So, when you remove data you change a relatively uncertain scenario into a very uncertain one. One of the effects of this is that the jet stream is even more poorly modelled and hence you get highs / blocking popping up where it won't actually occur in reality. As the data comes back and the confidence increases, the more outlandish outcomes disappear again.

You would also have seen this to a limited extent this Christmas too. There was a (feeble) attempt at blocking on the 6z run on Christmas Day - and again today, also on the 6z run of the GFS. You'll brush it off as coincidence, of course, but it's damned funny how it keeps on happening at times of lower data input!

Firstly, Happy New Year to you Rob and to all TWO members.

What a number of the operational model runs now seem to be indicating is a return to the set-up we had for much of the time during the fortnight or so prior to Christmas with a more southerly-tracking jet possibly returning. If that happens, it should mean a return to temperatures nearer to the average for this time of year, although unfortunately it will also mean a return to more widely unsettled weather.

These are diabolical IMHO wrt cold weather. Things could change quickly but at the moment there isn't any sign of prolonged cold and snow.

Its or should be of no surprise , we are the uk after all .

I think we might have to wait for the next solar cycle if i'm honest 

Yes just when you thought things couldn't get any worse...! Buy hey ho. After a seasonal day on Saturday it's business as usual then I guess.

At least there are numbers in the snow row , all 0's and its as bad as it get can get 

almost worth capturing as a screen saver. Truly dreadful and akin to some of the worst zonal dross “winters” of the past 30 years. 

It seems that I have to keep explaining that chart to you. It now shows verification stats for 00z on 31 December; take away 6 days and that's the forecast charts issued at 00z on Christmas Day. That has as a starting point the data collected on Chrismas Eve, so the alleged/debated data loss on Christmas Day is still not going to have manifested itself.

Tomorrow's chart will be for 00z Boxing Day and will include the data gathered on Christmas Day, so that should be the one to look at for possible accuracy issues.

I only remember him saying that February was likely to be the coldest month. I don't recall him saying that December would be cold.

www.gavsweathervids.com/winter-2019-20-weather-forecast.php

We think all three winter months as well as March should have cold outbreaks but December and February could verify as the coldest months of this winter with January perhaps on the milder side.

Thanks. It just shows how rubbish my memory is these days.

To take up another point that some people have been making, regarding the issue of the models having less data to work with on public holidays, I don't really buy into that. Most measuring and reporting of weather data seems to be automated nowadays. A large proportion of the data comes from satellites, and most weather stations are automatic. I suppose there might be a few less reports from ships and aircraft, but I doubt that would be very significant.

I think the best I can hope for is HP to be in close proximity giving foggy days and low temps 

ECM 216 - Dire

I never cease to be baffled by why anyone still cant understand that forecasting a season is impossible. Actually  impossible. As such it’s a completely pointless exercise

Every field of expertise needs its pioneers Matty. Fair play to these guys for studying it. (I couldn't do it)

It may seem like a fruitless task but without the likes of Gav P etc who dedicate a lot of effort into this we would never be able to crack the LRF code.  

Hello Matty Boy 

My post wasn't an  OMG , how can this be?? post 

It really was highlighting how impossible it is mirroring your thoughts , we have so many variables in this country . I am not a believer in pattern matching , I just don't see how it can possibly be accurate .

But just to reiterate I do find Gavs videos very interesting and informative.

Sponsor required for the Bartlett pattern...  

It looks relentless at the moment. Until that High pressure shifts from the Biscay-mainland Europe area we're destined for a period of mild, mostly benign weather for at least the southern half.  

At least everyone's heating bills will be much lower.  I'm not complaining, since we can't change what's coming.