July CET

[damianslaw]

I am going to stick my neck out and say with some confidence that we will record a below average month, I just don't see any sufficient sustained spell of weather that can help the CET to recover to above average temperatures.

However, I think that it will only just be a below average month largely due to fairly mild nightime minima, had we seen high pressure exerting itself more readily towards the end of this week then I think we could have recorded some fairly chilly nights but although the end of the week will see some cooler minima i don't expect we will see anything that particularly cool.

[Stu_London]

Just nowhere near what the BBC/Met O are forecasting Stu. Are you just using the infamous GFS temperature chart by chance?

BBC for instance forecasting routinely 13C tonight in the CET zone. As HP is now slipping away I think you may find we have that curious dichotomy of relatively warm nights but cool daytime maxes (although 24C forecast for tomorrow).

Where on earth do you get your figures from?? Firstly I am refering to an average for the rest of the week, so by using tonight when there is the last of the TM air slipping away is not really representative and then you still proceed to put up figures that are wrong

A selection of BBC mins and maxes

London - not in the CET zone, urban and therefore almost certainly a few degrees warmer than CET land

Max 23 Min 12

Now places in the CET zone for tonight/tomorrow

Oxford Max 20 Min 11

Pershore Max 20 Min 10

Cannock Max 20 Min 8

Redditch Max 20 Min 10

Stroud Max 21 Min 12

As for warm nights, I wouldnt be surprised to see ground frost mins in one or two favoured areas of England or Wales by the weekend

Edited July 17, 2007 by Stu_London

[SP1986]

Where on earth do you get your figures from?? Firstly I am refering to an average for the rest of the week, so by using tonight when there is the last of the TM air slipping away is not really representative and then you still proceed to put up figures that are wrong

A selection of BBC mins and maxes

London - not in the CET zone, urban and therefore almost certainly a few degrees warmer than CET land

Max 23 Min 12

Now places in the CET zone for tonight/tomorrow

Oxford Max 20 Min 11

Pershore Max 20 Min 10

Cannock Max 20 Min 8

Redditch Max 20 Min 10

Stroud Max 21 Min 12

Wouldnt say the last of the TM air is slipping away because itll be warmer tomorrow in many places than it was today. The airmass tomorrow will be the same as today, albeit drier and warmer (and still humid), although nights will be slightly cooler.

[West is Best]

Where on earth do you get your figures from??

The main BBC forecasts. Never use the regional or internet-based ones: they are always inaccurate. Stick to the Met O / BBC main forecasts on the hour (on News 24).

Hadley 15.6C to the 16th ...

http://hadobs.metoffice.com/hadcet/cet_info_mean.html

London - not in the CET zone,

Actually London is used in the CET compilation.

[Stratos Ferric]

...

Stratos - I'm glad to see you admit the 30 yr base is more robust than 10 yrs. I have no problem using the most recent 30 yrs, but for practical purposes it is fairly pointless to consider it necessary to update it any more regularly than every 10 years. ...

I've never said anything else, however, the fact remains that if the climate starts to change then early years in a longer period become an anachronism. The current period seems less warm compared to the recent decade than it does c.f. the recent thirty years, or even the reference period 1971-00. If a cataclysm occured tomorrow to bring Tamara all her dreams of a nuclear winter then there would be no value whatsoever in the thirty year reference period as a means of calibrating the current weather.

Back to matters at hand: the mid month projection is suggesting a very cool month by recent standards, and as Stu stated previously, 15-15.5 now looks like the landing zone if GFS is to be believed. I suspect this is still slightly low and would hunch 15.3-15.7, but whichever, those who punted low this month should be rubbing their hands in anticipation.

...

Actually London is used in the CET compilation.

Not unless you're being very generous with the boundaries it's not. If memory serves me right UKMO use Ringway (M'cr a/port), Squires Gate (Blackpool airport), Malvern and Rothamstead. Rothamstead is near Harpenden; more Luton than London.

CET is taken to represent the triangle between Manchester, Bristol and London, but of those three only Manchester is vaguely represented, and Blackpool lies outside this "rough" zone.

...

As for warm nights, I wouldnt be surprised to see ground frost mins in one or two favoured areas of England or Wales by the weekend

That's almost as remarkable a suggestion as was the famous (alleged) sighting of sleet in Telford last autumn when air temps were around 11C. The 850s are well above freezing, and though the air is sluggish pressure is low-ish and I'd not reckon on air clear and dry enough to allow temps to fall below 5-6 even in favoured rural spots.

[Stu_London]

That's almost as remarkable a suggestion as was the famous (alleged) sighting of sleet in Telford last autumn when air temps were around 11C. The 850s are well above freezing, and though the air is sluggish pressure is low-ish and I'd not reckon on air clear and dry enough to allow temps to fall below 5-6 even in favoured rural spots.

Telford was the said favoured spot I was referring to :lol: :lol:

Anyway 54.84%of the way through the month (up to the 17th) and these are the Manley figures from www.climate-uk.com

CET 15.5% (-0.7C)

EW Rain 63.0m (201%)

EW Sun 85.0hr (77%)

Suprised to see the CET not rise yesterday considering it was a very mild night on Monday (last night was slightly less mild), although the above figure is rounded down. Now an above average month rain wise, whatever transpires in the next fortnight.

Edited July 18, 2007 by Stu_London

[West is Best]

Not unless you're being very generous with the boundaries it's not. If memory serves me right UKMO use Ringway (M'cr a/port), Squires Gate (Blackpool airport), Malvern and Rothamstead. Rothamstead is near Harpenden; more Luton than London.

CET is taken to represent the triangle between Manchester, Bristol and London, but of those three only Manchester is vaguely represented, and Blackpool lies outside this "rough" zone.

SF we'll have to agree to disagree about what I see as your faddish 2-minute up to date average with an ultra-small sample size. Fortunately it has no support in metereology. I do think this is the time to be extremely careful, and all the more rigorous, about one's use of statistics. There has been so much loose guff written about climate change that we don't need to be sucked into it. Anyway, now I'm sermonising back at you, and I can see we won't see eye to eye on it. I happen to think you're completely wrong about this, I dare say you may think the same of me!

Re. the CET triangle although 4 stations are used for the final caculation, they are calibrated from the readings of 7 other automatic stations on the daily feed - including Northolt in London. Hence, why I said 'actually London is included'. Admittedly I was waiting for someone to query it, but the fact remains that the actual CET triangle is tested against a wider 'CET' net for accuracy. Rather a neat idea actually.

Yes, the Manley figure rose slightly to 15.5C rounded down. It remains much too early to assume a below average month is inevitable, although it is looking more likely. Patterns do not seem to me to be clearcut in the models at the moment, and with 2 weeks of the month left there's some uncertainty.

Edited July 18, 2007 by West is Best

[Bottesford]

Anyway 54.84%of the way through the month (up to the 17th) and these are the Manley figures from www.climate-uk.com

CET 15.5% (-0.7C)

EW Rain 63.0m (201%)

EW Sun 85.0hr (77%)

What an appalling set of figures for July - about as opposite as where I'd like them as they could possibly be!

[summer blizzard]

SF we'll have to agree to disagree about what I see as your faddish 2-minute up to date average with an ultra-small sample size. Fortunately it has no support in metereology. I do think this is the time to be extremely careful, and all the more rigorous, about one's use of statistics. There has been so much loose guff written about climate change that we don't need to be sucked into it. Anyway, now I'm sermonising back at you, and I can see we won't see eye to eye on it. I happen to think you're completely wrong about this, I dare say you may think the same of me!

Re. the CET triangle although 4 stations are used for the final caculation, they are calibrated from the readings of 7 other automatic stations on the daily feed - including Northolt in London. Hence, why I said 'actually London is included'. Admittedly I was waiting for someone to query it, but the fact remains that the actual CET triangle is tested against a wider 'CET' net for accuracy. Rather a neat idea actually.

Yes, the Manley figure rose slightly to 15.5C rounded down. It remains much too early to assume a below average month is inevitable, although it is looking more likely. Patterns do not seem to me to be clearcut in the models at the moment, and with 2 weeks of the month left there's some uncertainty.

http://www.wetterzentrale.de/pics/MT8_London_ens.png

If thats not clear cut, i dont know what is.

[Weather-history]

I went for 17.2

To get to that it would need a daily CET mean of 19.3

It hasn't even got to 19.0 once this year so far.

[Stratos Ferric]

SF we'll have to agree to disagree about what I see as your faddish 2-minute up to date average with an ultra-small sample size. Fortunately it has no support in metereology. I do think this is the time to be extremely careful, and all the more rigorous, about one's use of statistics. There has been so much loose guff written about climate change that we don't need to be sucked into it. Anyway, now I'm sermonising back at you, and I can see we won't see eye to eye on it. I happen to think you're completely wrong about this, I dare say you may think the same of me!

Re. the CET triangle although 4 stations are used for the final caculation, they are calibrated from the readings of 7 other automatic stations on the daily feed - including Northolt in London. Hence, why I said 'actually London is included'. Admittedly I was waiting for someone to query it, but the fact remains that the actual CET triangle is tested against a wider 'CET' net for accuracy. Rather a neat idea actually.

Yes, the Manley figure rose slightly to 15.5C rounded down. It remains much too early to assume a below average month is inevitable, although it is looking more likely. Patterns do not seem to me to be clearcut in the models at the moment, and with 2 weeks of the month left there's some uncertainty.

WiB, it isn't a "two minute" up to date average, though it is sensibly up to date. As I've already argued previously, I can see no logic, nor have I seen one presented, for the lag in the use of the stepped thirty year baseline. There is some argument to be made for "intemittent stability" i.e. the stepping, but why the five+ year lag. My suspicion is that it's largely due to manpower issues and priorities in the UKMO, though eualy it might well just be that nobody has ever got around to reviewing policy. There is NO other area of government stats that I'm aware of that has this lag given the short period of measurement upon which summaries are based. The "meteorology" doesn't agree is something of a sweeping statement, because so far as I'm aware there is no global standard for baselining.

Re "London included" - it clearly isn't. The use of seven sites for calibration is not the same as inclusion. If London were exceptionally hot and the other sites on the dayy were not, and there were sound reasoins for that, I would not expect an adjustment to be made in CET, though that would depend on exactly HOW the calibration operates. I'd read it more as a "fail-safe" test than an automatic correction mechanism.

[summer blizzard]

I agree, there is no reason to use the 1961-1990 average, it is as relevant as the 1831-1860 average.

[West is Best]

If thats not clear cut, i dont know what is.

All I do know is that 2 weeks in British weather is never clearcut. About 5 days with a following wind and a bit of good luck is about as far as even the best metereologist can push it.

You don't know what the weather's going to be doing in the UK 10 days from now. I don't. John Holmes doesn't. And I doubt even God does. In terms of probability you are right I think ... I'm just suggesting we can't really be certain at the moment, especially with an apparent pattern change. Wouldn't even surprise me to start seeing some continental plumes re-appearing on the charts ...

[Stu_London]

All I do know is that 2 weeks in British weather is never clearcut. About 5 days with a following wind and a bit of good luck is about as far as even the best metereologist can push it.

You don't know what the weather's going to be doing in the UK 10 days from now. I don't. John Holmes doesn't. And I doubt even God does. In terms of probability you are right I think ... I'm just suggesting we can't really be certain at the moment, especially with an apparent pattern change. Wouldn't even surprise me to start seeing some continental plumes re-appearing on the charts ...

Agree with your sentiments, a lot can change before the last week, however the longer it stays below average and is not rising the harder it becomes make up the shortfall.

[West is Best]

SF - I still think we're discussing two slightly different issues. On one I'm more or less in agreement: use the most up-to-date rolling average. Having said that I think we should be a bit reasonable. There's a huge amount of data switch over in producing a new 30-year base every year and in some ways I think the Met O have better things to do.

On the second issue, I remain in disagreement. Your 10 year rollover, with only 10 input data points, provides in my view for an insufficiently stable and viable statistical base. I do want to repeat my previous point too: the wild fluctuations that can occur with such a small sample size can be used against AGW as well as for. In extremis we have Terry Wogan (admittedly amusingly) taking a pop on a daily basis through this summer about climate change because it's such a p-awful summer. A properly long, seasoned, baseline enables us to take the heat out (ho-ho) of the debate and use mature sensible arguments about whether and how the world's climate is changing. I'm concerned the climate change argument is going to get lost in the midst of a host of wild, unsubstantiated, sloppy use of available data ... the media are at it big time, but they're surely not alone?

So - agree on one, still very much against on the other!

however the longer it stays below average and is not rising the harder it becomes make up the shortfall.

Very true ... I'm starting to think below average is really likely (just not certain lol!), but you're right about this.

[Stratos Ferric]

Agree with your sentiments, a lot can change before the last week, however the longer it stays below average and is not rising the harder it becomes make up the shortfall.

It's very hard to see beyond below average now. I'm not even convinced that there is any immediate pattern change in the offing either.

[VillagePlank]

SF, WiB,

Surely there's some middle ground here?

A ten year rolling average will describe what the most recent climate is like. A 30 year rolling average will put that in context of change. A fixed point average is useful for describing change. It about the use, not the mechanics.

You two, of all people, do not need to be reminded that if I put my head in an oven, and my feet in the freezer my body temperature will still remain average.

Edited July 18, 2007 by VillagePlank

[Stratos Ferric]

SF - I still think we're discussing two slightly different issues. On one I'm more or less in agreement: use the most up-to-date rolling average. Having said that I think we should be a bit reasonable. There's a huge amount of data switch over in producing a new 30-year base every year and in some ways I think the Met O have better things to do.

On the second issue, I remain in disagreement. ...

So - agree on one, still very much against on the other!

I think overall we're in agreement. The argument for a retained reference period is that it avoids the bizarre appearance of a season or year being, say, absolutely warmer than the one before, yet relatively cooler. The man in the street is not sufficiently adept at interpretation to get his / her head around this. My arguments are more based on statistical precision than overall utility. Here I am somewhat contrary though: I totally concede your point about robustness of a ten year sample, however, I keep returning to the point, a long time series is LESS valid if the data trend is consistently in one direction. The point re GW that you're making is actually EXACERBATED by a 30 year comparison. Almost any recent month looks warmer reference to the thirty year baseline than it does to the rolling ten. It's precisely this sort of situation that is capitalised upon by spread betting firms. Everybody knows Man Utd will beat Margate in the FA Cup so there's no real market; ask "by how much" and suddenly you have a market. At present most months are warm c.f. the long comparison period because we're warming, hence recent months are much more likely to be warmer than the thirty year mean. Against the shorter (admittedly more volatile) ten year measure this is less true and each year can be assessed in a more appropriate contemporaneous context. If a client of mine wants an assessment of corporate performance the fact that in the past thirty years they have always been profitable is of no utility whatsoever if a financial abyss suddenly opens up because background conditions have changed. You only have to read the pages on N-W in the run up to winter to see how strong the desire is to remain rooted in a past that no longer exists; I have to say, given that you are often vociferous against the "cold come what may" fraternity, I'm slightly surprised that you're hanging on to the argument favouring the 30 year baseline in ALL circumstances.

I completely agree with your overall hypothesis as a general rule, but as I keep stating, these are not normal times. I agree with the 30 year baseline as a standard, and any comparison I use is certainly not "instead of", but rather "as well as", intended purely to reflect more appropriately what appears to be the current trend in our climate.

...if I put my head in an oven, and my feet in the freezer my body temperature will still remain average.

Not for very long if there's gas in the oven! Don't try this at home folks, especially if the gap from freezer to oven is more than 5' or so.

[Stu_London]

I prefer a one year rolling average and am now absolutely convinced that the ice age is coming as July comes in a whopping 4.5C below the 2006-2006 rolling average

[West is Best]

SF - I think you're right that we're actually perhaps not so far apart. I do agree about your 10 yr point in terms of the exacerbation. I suppose my real worry is the way the CET is being used on the assumption of global warming. How do we know GW is occurring except by very carefully studying statistics over a sufficiently long timespan? You have assumed GW, and are using the CET as a tool to demonstrate the relative ups and downs of the net result. I'd rather see the CET employed as one local geographic example of how climate change may or may not be occurring over decades and centuries. I'd rather see it within a much longer timespan context: and how that global phenomenon may or may not have an affect on this little section of the world with its long-standing temperature series.

Bear in mind that I'm one with you on the fact of climate change. I just think we need to be stringently careful and factual about demonstrating it. The tabloid sloppiness of the way some AGW is being reported is going to bring the entire argument into deserved disrepute. If that hasn't happened already: if I were in the sceptic camp I'd be having a field day with some of the rubbish being reported ...

Stu - very droll!!!!!

Edited July 18, 2007 by West is Best

[Stratos Ferric]

I prefer a one year rolling average and am now absolutely convinced that the ice age is coming as July comes in a whopping 4.5C below the 2006-2006 rolling average

Stu, there's plenty on here who use the three day rolling average to substnatiate more fanciful claims re the end of the world under a thick layer of ice.

WiB: quite agree re (risk of) distortion, though when I want to make an argument for climate change I like to think I use a long time series and appropriate plots to make my case either way. My use of ten year series tends to be restricted to discussions in these (CET) threads regarding likely outcomes for current months, when my arguments regarding anachronisms in the early years of the thirty year series are pertinent. As this month aptly demonstrates however, short term volatility still has an amplitude greater than medium term change.

[Duncan McAlister]

Assuming the CET remains more or less where it is now for the next 5 days (I think it will, with rises today and tomorrow being offset by a cooler weekend under the rain), it will take an average temperature of 19C from the 23rd onwards to bring the month to average. I fail to see where this will come from and so can't understand why some people are predicting an average or slightly above month. In fact I don't think it will quite reach the 1961-90 average.

Edited July 18, 2007 by Duncan McAlister

[kold weather]

I think the chance sof an above average month are pretty much all but gone now I suspect given the models, it'll probably hold pretty close to where it is now looking at the output, Thursday will probably see a slight rise followed by another flal on Friday given fairly surpressed maxes (14-17C I suspect) but I think we are going to see a rare month below average, maybe moderatly below the 71-00 average at that.

[shuggee]

The length of the cooler spell is interesting in that many people in the past used the argument about how cooler spells 'fall between months' so their effect on the CET graph is somehow of a lesser extent than would otherwise have been the case had it started on the 1st and ended on the 20th. Not so this month - shows how stats balance themselves out

[West is Best]

I'm not sure ... I think I'd put it about 75% at the moment. The reason for the 25% is that the pattern is in a curious state of flux, with enormous variations in the models run by run, day by day. I still wouldn't rule out a plume set up and with mild nights still on the cards I'm just not as certain as some of you about the drop off. The old 61-90 average is certainly still possible I think, though that's not really the point. The real reason for wanting to leave some room for change is simply that there are 13 days of the month left.