chrisbell-nottheweatherman

I don't know if this is of any use, but I'm just viewing a Youtube video by Simon Keeling from Weatherweb, and he reckons the GFS is worse than before the upgrade as it's dramatically-overplaying low heights in the Atlantic. They are strongly favouring the ECM at present. https://www.youtube.com/watch?v=PkDmsPq5gs8

I'm trying to post a reply in the Mod thrad, but, unless I select an alternative post to quote, it sems to quote a post by "knocker" from a few days back (this is when I select Post Reply. Can anyone offer any advice, please? EDIT: Please ignore - have solved it myself!

At least 16-17C in December gives a feel of alternating autumn and summer, rather than the 5-6 month autumns the south has endured since 13/14.

Given the recent references in the model discussion thread regarding the effect on the weather in the UK of the vortex sitting over Greenland, which I imagine is related to El Nino via. its effects on the vortex, I find myself wondering whether El Nino can explain the seemingly unmovable Euro high. Is there any research that supports my speculation?

Sadly, the latest outputs are showing a blowtorch south-southwesterly for the 25th, with 2m temperature anomalies 7C or so above average. Our daffodil-loving friend may be able to get his outdoor cooking apparatus set-up in his garden as he's been threatening to do!

What's the sun? It's been so long since most of us in the east have seen it that I think we've forgotten what it is!

You can't argue with the raw data. We're on course for an exceptionally mild December, and, IMO, an equally exceptionally-mild January and Februry as well. Of course Atlantic-driven weather is the default British winter pattern, but the point is tht these usually see more Pm air and less Tm as the winter progresses. If anything, current outputs are suggesting the opposite. I joked the other day about charts showng an airmass sourced from near Brazil; some of the outputs being posted in the Mod thread this morningt are showing a switch to just off the coast of NW Africa!

I utterly agree. It's the seemingly endless half-light, even around midday, especially if it's combined with drizzle. Like you, if it's going to be exceptionally mild at this time of year, at least it could be bright. I'm not asking for wall-to-wall sunshine, but it would be nice to have a few sunny intervals.

I didn't say they were, to be fair; my accusation was that they appear to be, by way of cognitive dissonance, separating their love of zonal conditions from the problems in Cumbria.

There are some that love this weather and want to see it continue non-stop through until lte March/early April. What annoys me is that a small percentage of such mild muck-lovers are usually the ones who are the first to moralise on the evils of enjoying wintry weather, due to the accidents and disruption it admiitedly causes, yet, having seen footage of people's homes ruined by flooding seem to take delight in wanting a continuation of the weather set-up that caused it. Fortunately, according to Weatherweb, the 500mb GFS output suggests something closer to seasonal for the end of the month.

Which will no doubt be a relief to those in the north west; thanks for posting these outputs, Matt.. Changing the subject slightly, yet still staying with matters meteorological, am I the only regional Netweather member who normally finds themselves hating incessant gloom, drizzle and unseasonable mildness who's trying to convince themselves that they love it really, and that clear, crisp, frosty days and newly-fallen snow are awful?

At this rate your firewood will have rotted (unless it's in very dry storage, in which case it'll probably fossilize) before you get round to using it.

Just popping into the woodshed to make sure the barbe's in fine fettle Who'd want Christmas dinner in those sort of temperatures anyway? Seriously, for someone in my position whose interpretation skills aren't up to much, I presume that's showing +4-5C to the average temperature, therefore around 12-13C?

Perhaps I didn't mean that entirely seriously? I'm also getting annoyed with one or two of the rain-lovers on here who are welcoming outputs which would see those who have already been affected by flooding affected further just because they (and I notice they live a fair distance from Cumbria) want to see deluges on their own doorstep, and the possibility of endless low pressure systems excites them as makes it more likely that they'd see rain in their area, even though it's the last thing those who've seen enough rain recently to last them a few years never mind a few weeks need. As a moderate, sensible coldie (i.e. my preference is to have some seasonal weather at this time of year rather than 5-6 months of autumn), if I knew that an area was badly-affected by snow, I'd be happy to see a warm-up even if there had been no snow in my area, for the sake of those who were struggling.

What's the latest regarding the drought in California? I understood that the current El Nino was supposed to end it.

I'm contemplating putting him on my ignore list, which disappoints me as he's usually right, but I'm getting a bit tired of "no signs of cold in the woodshed" "daffs doing nicely" and cherry-picking outputs which, as you say, show air sourced from somewhere near the equator!

As usual, someone else sums-up my point in 2 lines when I needed 2 paragraphs!

Fair points. You're much more knowledgeable than I am on the meteorological side of things, so I'm happy to defer to you on that. My point is that there's a difference between calculating a mean on the one hand, and using it as a trend guide on the other. If we had a group of people at a conference or public lecture, for example, we could calculate their mean height. If the distribution of their heights was a standard continuous variation, i.e. a normal or Gaussian distribution, the curve, where heights on the x-axis are plotted against number of persons on the y-axis, would be something like this: In this case, we might expect the mean value to give us an idea of the trend regarding the height of attendees. However, if the curve is closer to this: then we have a situation where the mean will lie between the two clusters (in my example, it might suggest that the trend is for attendees to be around 5 feet 9 inches in height, where, in actual fact, most are between either 5 ft 2 and 5 ft 6 or 5 ft 11 and 6 ft 3). Applying this to the current model situation, my understanding (and please correct me if I haven't grasped this properly) is that the high levels of Shannon entropy as a result of changing underlying patterns and drivers means that the ensembles are tending to fall into one or two (or more?) clusters in terms of output. As a result, while an ensemble mean is still an ensemble mean, the trend it suggests my lie between clusters of members and may therefore infer a trend which doesn't exist (perhaps showing a solution which is neither one thing nor the other when the individual ensemble members tend to fall into one definite trend or the other). Apologies Mods for the off-topic diversion.

If you're referring to the conversation I was involved in, I think all we were saying is that a mean average is only as useful as the data from which it emerges allows it to be. Combining my generic understanding from three years of undergrad biology, plus some practical experience in a masters' programme with the specific meteorological knowledge from others, I get the understanding that periods of widely-diverging ensemble members when the current pattern seems uncertain in terms of whether it will continue as it current is presents typically-non-Gaussian (i.e. unequally-distributed) data. Such distribution makes mean averages without standard deviation information less powerful as it will tend to show a result between clusters of data as opposed to representing the concentrations of individual data points created by the clustering of the data.

After yet another gloomy morning when it feels as though it's still partly dark, the clouds have cleared as the front has moved away leaving a pleasant, sunny afternoon.

I honestly think this needs to be quoted whenever Mods open a new Model Discussion thread. I for one as an inexperienced model followr who relies on knowlegeable people's analyses respect John's posts and try to view them for what they are. It seems at present, based on his analysis that the anomalies are showing no movement towards prolonged cold, but there is still every chance of a short-lived colder spell between lows, with the trend being for a more southerly mean jet location.

Long enough to remember a couple.

Thanks for providing the meteorological explanation on that, Nick - it helps me see the link between the basic side of it I already knew and the details of the current situation. Presumably, your comments regarding ensemble means being more reliable in straightforward zonal set-ups is due to low Shannon entropy, whereas the expected pattern change and resulting high entropy/low confidence is likely to result in a greater spread within the ensembles and therefore the ensemble mean has to be treated with more caution?

Though I cannot speak from any exprience meteorologically, if I might be arrogant for a moment, as a trained scientist, this is something that jumps-out at me. Mean averages on their own, without standard deviations and statistical calculations of confidence intervals are not always helpful, in particular, as you say, when data is clustered rather than a more standard Gaussian distribution.

Model thread is all over the place - everything from "no sign of cold into the New Year" to "northerly and lying snow in 10 days' time", and that's just today's posts. Models are evidently confused and not showing consistent output, and, as usual, the mild rampers are cherry-picking the charts which show our air mass being sourced from somewhere off the coast of Brazil, while the cold rampers are doing the same with outputs showing it coming from the North Pole!