weather eater

Must have just seen the latest GloSea5 and EC Monthly updates and taken a peek at the UKMO BBC weather presenter, Rachel Mackley, faints live on air during forecast https://www.youtube.com/watch?v=qnCNQ3s5BQs

Couldn't agree more Karl. In fact the GFS has done really well here, it picked out the general pattern and stuck with it. Only on a couple of occasions has it hinted at a longer cold spell and it's our hopes and expectations that lead us up the garden path at times. We all have to face it, we are a tiny island and 50 miles N,S,E or west makes a big difference. And I really do feel that we need to keep in mind what is going on under the hood of these models. Some serious mathematics with numbers that are already old even before they are run. Far from being poor they do a fantastic job every single day of the year. But the further you get from zero hour the more they are going to be wrong. What is needed by those of us who view them every day is a change of onus. From will this run verify to how much and from what point is this run going wrong. For the UK I would suggest that's day three at best most of the time. Of course the ensembles are supposed to be a guide for this but all too often they are either not up to the task or they are cherry picked for cold.

Yes a broad brush Nick but if the broad brush is not indicative of a proper cold spell then there will be no proper cold spell. The details may well feature a transient colder spell as we are now seeing but the fact is that for most of us are looking for something more than a 2 day northerly. However, my main point is the futility of expecting runs to be consistent post a couple of days, it's never going to happen. We could point to 2010 but even then the details kept changing in the run up, what made it stand out was the fact that the changes stayed in a cold theme.

The issue is not the models it is people's faith that runs, any run, any model will verify post a couple of days. I keep saying this, the models run on data that is constantly refreshing, They are always going to less and less accurate the further they get from zero hour. Consequently what we see is not the future synoptic pattern but an approximation of it gained from old data. I do not understand and have never understood the fixating people have in wanting to see fancy cold synoptics post day 2, they are pointless 99% of the time. I can see full well why JH and Mushy use the charts that they do. They may lack detail but they are indeed a far better guide to what may occur in the mid range.

Small Island, small changes, big differences. The ECM is pretty similar to the GFS except the track is further north. Looking at it from that stand point, it is the ECM that is inching towards the GFS evolution.

Indeed we are nick I've already this year had to repair a great chunk of fence and tonight I have left our gates open because they keep taking a battering, absolutely sick to death of storms. On a more happy note every run at the moment looks better and better and as I said many times before Stella charts in FI no matter how wonderful they look fall over 99% of the time, only 2010 was picked up at long range by the models. It has always seemed odd to me that we almost always see cold spells predicted in the mid to long range fall apart but rarely see one evolve quickly from a base that seems innocuous. Mind you that could be famous last words, it could still be all downhill from here. We saw something similar, what 11-12 years ago, the easterly that never was.

Very wise of them., I am optimistic but we know that if it can go wrong it will go wrong and on a tiny island small changes have big impacts.

It's the expectation that it could develop further that has got people interested. Great charts that ninety nine times out of a hundred just end up disappearing as they come within a realistic time frame are pointless. At the moment we are seeing the opposite, poor prospects looking better by the day. Of course the most likely outcome is that nothing will come of it. But that's got to change someday and maybe just maybe this is it.

What I like about this is that it's not FI mega charts, that you just know will never evolve. I keep saying this but it is evolution, all runs are based on the start data but the data does not stand still. It is always old data, thus runs verifying is just a pipe dream. This is evolving seemingly out of nowhere it's not close enough to be certain because there will be more changes but it's not FI either. Just for once can we watch something evolve from something poor to something good or great, rather than what we usually see, which is evolving from something great into something rubbish.

Almost good and then it sinks the high, not sure I take any of it seriously any more

firstly I don't think we can call any model poor both the GFS and UKMO sit out on their own. Both have been as consistent as I would expect, with both showing variations on their respective evolutions. Poor might be of some use to describe one or the other once we get some resolution. Rather like some sort of mathematical tug of war the other models have chopped between one and the other, first lending support for one solution and then the other before swapping back just before the winning line is reached. Big ECM coming up, well we have been saying that for days. Real proof will be when one or the other (UKMO/GFS) or possible both, finally admit defeat. As has been discussed Alex may be pivotal in that. For those hoping for cold and snow it's a rollercoaster but from a pure fascination with modelling the weather, I simply can't think of a more interesting period.

Problem is verification stats tend to iron out the creases and that may mean that particular models may have particular issues in some regards. 2 its ranking also does not mean it is always better than the next on the list, just usually. Why it comes to HL blocking patterns that are directly influencing our weather, I tend to work to the theory that if the GFS is not buying into it, then it will not happen. I cannot recall when this was not sound advice.

UKMO output this morning should come as no surprise. As over proggessive re the Atlantic the GFS is, the UKMO is with HLB. I have no idea how good its verification stats are but I have never trusted it 120hr plus output. Overall my opinion on the models is still, that we will see the UKMO backtrack further. The ECM seem utterly unsure and the GFS still looks to me like it has been nearer the mark all along. Changes in its output have for me just been standard inter run variation and evolution not small steps to the solution the UKMO has been showing over the last couple of days. I hate to bang on with this word but evolution is constant with modelling synoptic patterns and even tiny changes have big impacts for the UK. Consequently It is perfectly possible that we may get a little bit more mileage out of this spell yet. currently sleeting in my little corner of the fens and getting rather windy.

A welcome piece of news though I'm not sure we need Ian to tell us that, as it seems apparent in the models. Not to mention we should all be doing that anyway.

I wouldn't worry about 120hr plus charts anyway. The models will continue to evolve the pattern and Stella charts at 120hrs plus are about as good as a marzipan thing beginning with D. Much better to see these appear at a closer range.