Mark Smithy

Thanks for the clarification Paul. Was somewhat confused by this: "The UK's weather has been stuck in a rut recently, with persistent low pressure to the west and north-west of Britain bringing very mild and changeable weather, with south-westerly and southerly winds. This pattern looks set to persist for most of the coming week, but there are growing signs that the pattern will shift as we head towards next weekend." So whether it was intended to be about this coming week or next week, the signal for high pressure to "build" to the east is really not consistent in the current models. Some show it but others show that we are not through with the mobility yet: the so-called "rut". There is no real "pressure build" on the ECM. Look at the transition here from this at T168: to T192: to this at T240: That's not a significant high pressure build on the ECM but an ongoing west to east mobile flow. Or, perhaps to re-phrase, it's a transient high pressure cell. That's two important runs in the past 12 hours of output which have showed mobility continuing and when we see that it can sometimes be a sign. As you know, I rate the accuracy of using anomaly forecasting for the UK as fairly weak. They regularly seem to be inaccurate for good reasons previously stated about the prevailing conditions in the UK for wind direction, sea currents and temperatures and relative warmth for our latitude. In other words, too many downstream variables come into play to make us rely on them. A useful tool to consider though, for sure.

Morning. The ECM clearly hasn't read Ian Simpson's declaration of a pattern change ahead "The aforementioned pattern change will set in, with high pressure building to the east of Britain." Pattern change on the way next week - the end of the rut is in sight WWW.NETWEATHER.TV Finally, the weather is on the move, with a change in the pattern looking likely later next week as high pressure builds to the east of the British Isles. This might still happen but it's never wise to be too definite about UK weather 10 days out. The 18z GFS and the 0z ECM are hinting at continued mobility.

Edited and re-phrased my original a little more felicitously: I've been watching them for three decades (back in the earliest days of the upstream signal charts) and they are so often wrong. I reckon the reason is the long draw across the Atlantic, the prevailing downstream signals here which are so strong that they often blast away the upstream signals. We have relatively unusual weather here: far milder, for example, than our latitude should produce, because of other factors which come into play: sea and land temperatures, prevailing winds, and of course the Gulf Stream. For these reasons I strongly suspect all model outputs beyond T240 for the UK are worthless. Arguably, from T168 to T240 should also carry a health warning. Most of us who are cold lovers and who have been following the model outputs for as long as I have know deep down that this is true. Come inside T168 and we can take notice. Outside of it? Of little worth.

I've been watching them for three decades and they are so often wrong. I reckon the reason is the long draw across the Atlantic, the prevailing downstream signals here which are so strong that they often blast away the upstream signals. We have relatively unusual weather here: far milder, for example, than our latitude should produce, because of other factors which come into play: sea and land temperatures, prevailing winds, and of course the Gulf Stream. For these reasons I strongly suspect all model outputs beyond T240 for the UK are worthless. Arguably, from T168 to T240 should also carry a health warning.

Yes indeed. Nothing overly dramatic in terms, for example, of storms in the south but a pretty mobile west to east flow. And as we saw on Friday and will see again today, slow moving fronts can often produce a lot more rain. What I still don't understand is what happened to the upstream anomaly signals? I was told fairly categorically that this kind of mobile zonal pattern wasn't going to happen because it went against the NOAA anomalies. Settled high pressure was apparently going to be in charge over the UK for at least the first half of the month. But once again the UK weather decided to do its own thing. The models have kicked into gear with zonal Atlantic flow, fleeting ridges of high pressure, but a generally mild and wet westerly flow. Exactly as you say. Personally I think the anomaly charts are of low value to the UK. They rarely seem to get it right, and where they do, might it just be a case of blind man's bluff? What we cold lovers need to hope is that this zonal pattern now has plenty of time to play out before winter proper begins. The fear would be that, having been held at bay, we could be into a pattern change of zonality lasting many months. I could certainly see this Atlantic phase lasting until Christmas.

As anticipated, many of the 0z GFS members do not buy into the low pressure operational (it's only an actual outlier for a few hours though!). Remember too that the European models also go with the more mobile pressure systems so it will be interesting to see how the GFS operates through the next few days. The ECM is autumnal and zonal e.g.: Are they actually ever right? Or do they not just play catch up with the more immediate model outputs? Too many variables come into the downstream mix to make the long term upstream anomaly signals reliable, for the UK at least.

A MUCH more turbulent 0z GFS. As I'm sure everyone has seen, the GFS has been holding out with high pressure rebuilds, and we've often seen the operational going against the mean in this regard. Well the overnight run could be said to have finally cracked. This is the first time the GFS has gone full tonto autumnal: With repeated bands of rain, sometimes heavy: It'll be very interesting to see if this is an outlier on the pressure charts. If it's not, and it has good support, then hats off in a gale to the ECM and UKMO for picking up this trend and running with it sooner.

0z ECM also distinctly on the nippy side:

Nippy northerly on the 0z GFS: A month of two halves and certainly little that could be described as 'summery' about this September

It's really quite baffling. Here is the very latest 6z GFS which you might think would have sorted out the precipitation? Erm, no. Actual radar for the same time: So when people post, 'not much rainfall' on a GFS run, it's worth bearing in mind just how sketchy the output can be on precipitation.

Here's an example of why precipitation charts from the model outputs should be treated with circumspection. I'm in Dorset this week and it's absolutely chucking down this morning. Here's the radar: The ECMWF got this pretty much spot on: But here's what the 0z GFS predicted for now: Others will be able to explain why the GFS seems to have such a struggle with precipitation forecasting on these shores but the fact remains that for short term modelling it's best to look at the FAX / Met Office outputs and the ECMWF. I believe the Meteo group / BBC / Met Office now use ECM for short-range modelling? GFS is still excellent for medium range and longer term trend spotting. I notice that again on the 0z it's hinting that the high may not stick around. The 0z ECM has now picked up on this GFS idea:

Signs on the GFS 0z that the high pressure may not stick around. It was first hinted at on the 18z and it's more apparent with the 0z. Take it out past T200 and we're back to the Atlantic: Fair to say that the ensembles are unsure about that: It's a signal worth keeping an eye on. Has the GFS picked up on upstream mobility?

Perfect summary of the model outputs there! The question I think will be whether that rebuilding high pressure holds at the end of next week or whether, as per the 0z GFS, there are signs of the Atlantic re-awakening to blow it away e.g.: Not the September many of us envisaged during those long hot summer days and nights. It's like someone flicked a switch.

Still a lack of clarity about what happens next i.e. around 5 days from now: I think it's been said on here a lot that the model outputs struggle when there are ex. tropical storms thrown into the mix. This seems to be no exception with different scenarios occupying the ensemble suites. We wait and see!

Morning CS. Subtle but important differences though between GFS and ECM? In fact, not so subtle. One has high pressure building over the UK, the other does not. One has high pressure building over France, the other has the opposite: GFS: ECM: What might happen with the ECM chart is that high pressure from the mid-Atlantic comes in behind that low to the north? So at around T240 you might end up with something similar but basically the ECM is a more uncertain and mobile pattern, whilst GFS reverts to its default high pressure build. The fact that there's such a lot of pressure scatter amongst the GFS ensembles from as soon as day 5 suggests they don't have much confidence yet about the likely evolution. In the meantime, a lot more squally rain to come - even for those who have so far somehow managed to escape the downpours.

In terms of rising high pressure there wasn't a lot of support for the 18z GFS operational but there was a huge scatter from as early as the 12th September. So it will be interesting to see where the latest 0z GFS operational sits in the ensembles. Basically the models are really struggling with what happens next. Does the brief ridge of high pressure build into something more substantial and long-lasting? Or will the Atlantic continue to push on through? No one knows.

According to the latest model runs, above average temperature, perhaps well above average. More cyclonic than we have been used to of late. But the real significant anomaly looks to be rainfall which could be well above average. So I suspect that, at the moment, going by strictly meteorological measurements this is not at the moment looking like an average September as measured against standard references. Of course, it's only the 4th September. A lot can change in nearly 4 weeks.

I don't think these model outputs are presenting a pattern that we have seen this early in September for some time? What is a particular shock to the system, albeit not yet in the south-east, is just how quickly it changed from summer to autumn. Often September throws up plenty of settled, warm and even hot days. It may do so yet. Re. ex tropical storm Danielle, you are right about the wind intensity. However, these ex hurricanes and storms usually produce anything BUT 'very standard fare for early autumn'. It may not have intensity of wind but they invariably carry high levels of humidity and precipitation, with often very 'heavy' atmospherics. You can see this well on the 0z GFS and especially the 0z ECM: That track across southern Britain is not supported by the aforementioned ECM, which has it tracking across southern France, but look what happens when it reinvigorates over the Alps:

Slight tendency I'm noticing with model output watching is: 1. To assume minute details of e.g. when and where precipitation will fall must be predicted by the model outputs with precision even days not just hours in advance. 2. To assume that just because something shown on a model output isn't happening in one's own back yard right now then it isn't happening at all or isn't going to happen. It's probably best imho to look at the synoptics of model outputs and see the general pattern, rather than treating their hour-by-hour estimates of temperature/precipitation as gospel. Interesting situation at the moment for example with the ppn band in the south west slightly further east than most models were suggesting at this stage, although it hasn't really pepped up yet by the looks of it. Exactly

Are you sure about this though? Next weekend on the ECM does offer a brief respite but only before the next system pushes fronts across. It looks pretty cyclonic, no? A long way out so much can change, including on timings.

I love rain, and so do trees and plants, so I'm desperately hoping it comes off

A wet run throughout from the 0z GFS. Will this actually happen?!

The models are getting a lot more progressive now. As the 12z ECM picked up yesterday, so the 18z GFS, 0z UKMO and 0z GFS are agreeing. Which doesn't mean height rises won't occur instead but there's some genuine zonality entering the fray here and it would be a brave person to bet against this trend. There's even cyclogenesis off the North American east coast. It feels like aeons since we saw that? I suppose it had to happen sooner or later. Meanwhile, rain incoming which looks pretty chunky - for the southwest at any rate. Let's see how this develops (or doesn't) for the southeast. Precipitation off slow moving low pressure cells is notoriously difficult to pinpoint accurately so probably best to give the various model outputs a bit of a break. They've got this mostly right so far.

I was down in the west country today and it absolutely chucked down. Torrential tropical rain. And that wasn't really even supposed to happen today. So I'll take the 'rain, wot rain?' posts with a pinch of salt. For now ...

They still are but it's for the North west of England and West of Scotland. The 12z ECM is really interesting. A significant amount of rain for all parts but that's not the most interesting thing about it. First here's the rain throughout almost the whole of the next 10 days: But the most interesting part is that it is, I think, the first properly zonal model output run since the spring? Edit @JonSnow beat me to it