The Enforcer

2024 in Bramhall began in the same vein as 2023 ended - pouring with rain.

A new low bar for the 18z ensembles - P22:

There will probably be a 'worse' ensemble member on GEFS 18z.

At least it's better than P17 from GFS 12z:

Sounds like the ECM version of P17.

Remove the word 'more' and I might be able to agree.

How is it possible to identify which P-numbers those ensembles are from the ensembles chart? UK high now appears to be touted as a 'certainty', but not for me until the P17 option has been wiped from the suite. It makes more sense to me to micro-analyse P-numbers that show things going awry than it does to cherry-pick the coldest one(s) in the hope that they become the norm in the many runs that follow.

I made one once, which was countered by a Day 10+ chart.

Do you know what pattern the orange red/runs on the 8th that reach about +4C are showing at that timeframe?

OK, but presumeably the jet has shifted south at the longitude in which the UK resides? In terms of actual weather I can't say I have seen any difference: rain, wind, overcast when not raining and hardly any sun - the same as last week and the three weeks before that.

After the GFS operational spews out a series of mild outliers, it has a tendency to show a colder solution, just at the point where its ensembles and other model sets decide on a milder outcome after all.

I'm glad you agree it's useful. As for performance, it all depends as to what was actually diagnosed - blocking of the jet stream away from the UK or polar air flow over the UK (accepting that it does not cover the type of details that in any event often prevent snowfall from occurring)? Either way, it is not possible to conclude on round 4 until it has actually occurred or indeed hasn't occurred.

Ah yes, another 10 day+ chart. However, I could be persuaded after a further 20-something runs.

I am sure GSDM is a useful forecasting tool, but there hasn't been a cold period yet.

I'd actually go a stage further and argue that not all ensembles are equal, such that those closest to the average line for the time of year ought to be weighted more heavily compared with those further from the average line. For arguments sake, let's say the average was 0C 850 then an outlier at 0C would be the equivalent of 5 runs at -5C and 10 runs at -10C.

Need to get to a position to be able to delete the word "almost".

I tend to look at mild outliers differently. It matters not if they are the operational run or one of the pack. Their existence serves to illustrate that a cold spell, even if promoted by the vast majority of runs, could still 'go wrong'. And in my experience, when the possibility remains that it could 'go wrong', it usually does.

Cold enough for rain.

They are usually pretty effective at blocking until March, when they then take their annual 9-month sabbatical.

"Oh no it won't!"

"And win the greatest victory since the Winchester flower-arranging team beat Harrow by twelve sore bottoms to one!"

There's a Blackadder quote that can be paraphrased for every situation: "E-E-Exactly! And that is what so brilliant about it! We will catch the watchful MOD thread totally off guard! Doing precisely what we have done eighteen times before [show a cold spell at T+some high number only for it to disappear by T+120] is exactly the last thing they’ll expect us to do this time!"

It's not the overall amount of rain that's the issue, it's the incessantness of it. No respite at all. 1. Fronts don't clear properly, leaving 'baggage' behind. 2. Troughs appear in gaps between fronts at T+0 3. Regardless of the direction of the flow, the outcome is the same: rain. Even high pressure brought rain when there were no trapped fronts and nothing on the radar. 4. Virtually no sun, any breaks in the cloudcover always overnight, so everything stays waterlogged and never dries out.

Don't bank on it. There was a ridge of High Pressure about a week before Xmas. Outcome? Rain.

T+216 is not what I would define as "early" and therein lies the problem.