seb

Ok I’m talking about the weather warning which was issued about 4 hours later

Certainly the case as every single automated app (* I refuse to look at BBC Weather but MetO, Accu, WeatherPro, Weather Channel and Apple/Dark Sky) has picked up on it in the last few hours and is showing snow for my location on Sunday for several hours (and into Monday) so clearly sufficient model consensus to warrant the warning.

Not sure if it's 30% (the text states confidence is very low) but considering the struggle of the models at the moment, I wouldn't expect anything else.

MetO warning for Sunday's SE feature just out. 2-5cm widely, 10cm in places.

This is what my model blend is producing

In fact not just winter. We don't have to go back further than summer which the GFS constantly wanted to kill off with fronts sweeping in from the Atlantic, just to then fall in line inside the 3-5 day range. If the GFS would have been right every time it brought the Atlantic in over the last year then most of the country wouldn't have been in a drought!

Great charts this morning. ECM and GFS almost swapped places with the GFS now on its own, but overall more cross-model agreement to keep us cold. Still, nothing is nailed on so best to stick to not taking anything post Sunday for granted.

There is of course also the possibility of lots of snow with only a minor incursion of milder air (short change to sleet/rain) and straight back to cold and snow before clearing. Much more rare in the UK than many other places but not unheard of

Wow those are some amazing charts this morning. Yes, as always a lot of caveats if peeking into FI (still sticking with my 4-5 days for now) but it has been a long time since no matter what solution is being churned out by different models and different model runs that all of them end up in a variation of cold/very cold for the foreseeable future. Lots of exciting options re snow, lots of potential for re-load after re-load and cold, cold and more cold.

Beautiful and indeed interesting charts across all runs tonight. Plenty of plausible cold and potentially very snowy options ahead once the cold is properly established. Further afield the deep lows being churned out are also of interest as very unusual. And yes, let’s keep them far away and then slide

Agreed. However, we all know that once the cold is properly established in the UK surprise snow features pop up at short notice. One of the few occasions where our geographical location is an advantage.

Much more sensible runs across the board this morning, stepping away from things that are just not plausible or possible. Still some uncertainty but if you’re after cold and potentially snowy surprises you couldn’t ask for more!

I think that is the point people are trying to make Struggling = not being able to provide any kind of agreement even at 5/6 days whereas during more common setups we can usually rely on a good few days beyond that. The models are doing what they’re meant to do, it’s just that the output isn’t helpful to make much of a forecast. That’s not withstanding random hurricane-like low pressure systems being modelled lol.

Snap

Time to stick to 96-120 hours for now, regardless of model. It’s clear that none of them agree and each one of them throws out different solutions with each run. Tbh it’s no different to what happened in 2010. I went back through some blog posts etc. from back then earlier.

GFS

The perfect-perfect storm lol. You’re right though… it’s borderline impossible.

Mainly GFS.

I go out for a few hours and come back to the most bizarre runs I’ve ever seen. Very, very interesting times ahead.

Fully aware of the risk that I may sound like that lady who was in charge about 5 prime ministers ago, I’m just going to say “Nothing has changed” and copy and paste what I said earlier. Substitute GFS for ECM and so on Arguably last night’s ECM wobble in the run up to what looks more and more likely to be a decent and long-ish period of cold weather was one of the most extreme I have ever seen. But, it underlines further just how much models struggle when having to deal with something out of the ordinary. It happens every time and I don’t think it’s the last wobble we’ve seen over the next few days (or weeks depending on how long the pattern remains). I hope it was a good learning experience for those who aren’t used to it (not through their own fault but simply because these setups/events don’t occur too often). The number one rule is not to take individual model runs as gospel, especially if they make no sense and with cold weather in particular to stop at day 5, sometimes even day 4. For what happens after that period, apply common sense and logic.

Arguably last night’s ECM wobble in the run up to what looks more and more likely to be a decent and long-ish period of cold weather was one of the most extreme I have ever seen. But, it underlines further just how much models struggle when having to deal with something out of the ordinary. It happens every time and I don’t think it’s the last wobble we’ve seen over the next few days (or weeks depending on how long the pattern remains). I hope it was a good learning experience for those who aren’t used to it (not through their own fault but simply because these setups/events don’t occur too often). The number one rule is not to take individual model runs as gospel, especially if they make no sense and with cold weather in particular to stop at day 5, sometimes even day 4. For what happens after that period, apply common sense and logic.

It really isn’t showing on the GFS.

Which, unless I’ve gone blind, isn’t shown anywhere! Instead GFS now randomly produces a deep-ish channel low out of nowhere.

There certainly was hype but at no point were we anywhere close to the current setup. There was also model agreement across the board around 5 days before that it would most likely be disappointing. Very different to models jumping about and wobbling a bit. They always, always do this with such a huge amount of cold air and high pressure sitting there. In fact 2010 had individual runs up to 3 days before showing all sorts of mad things that would have caused cold rain instead of a foot of snow and -12 temps. Likewise, the “reload” a few weeks later was not shown at all up to 7 days before it happened. I only ever come here when exciting stuff may be happening but I’ve been doing this for a long time and unless the next few runs stick to what the ECM showed this morning and what the GFS has produced an hour ago, there simply is no trend. Just as there was no trend yesterday - just look back at the doom and gloom after the ECM run and then the hype after the GFS run.

That is all correct but not at the time frame some people are concentrating on. I remember Feb 2013 very well and I think it could be a good idea if you also look at the charts leading up to the non-event.