February CET

[reef]

A months cold is a months cold whichever way you dress it up. What difference does the start and end day make?! :lol: Apart from satisfying the meteorological annuls of course. But it doesn't somehow make the climate warmer than it is because the dates don't conveniently match a calendar month!

Lets wait till the end of the month and the whole matter should be put to bed anyway regarding the point you are trying to make. But tbh I am rather bemused by the fixation and significance of this 3c winter CET thing. IMO it is pointless trying to place limits/benchmarks all the time and then moving them around once they are subsequently breached to try and keep a certain POV alive. We have had this with the annual rolling CET, but the script wasn't followed, and now it is happening with monthly winter CET's.

I don't think that this sub 3c CET is that significant - only in the minds of those who feel it is so hard for it to be breeched in the UK for some reason. Still, only 3 weeks to go and you will all know the outcome for this month I guess. But it will happen at some stage anyway.

In summary I wasn't actually trying to illustrate how cold it has been at all - merely that some people are speculating about the unlikeliness of something that has already happened under a differing range of datelines!

It makes a huge difference at the end of the day. Ive never seen anyone on here suggest a 31 day period averaging 3C is probably beyond us. I have however seen members suggest that a 3C month might be (Im not one of them by the way). Theres a very good reason for this however, it is and always has been much easier to record a 31 day period averaging 3C than a calandar month (and indeed, a rolling sub-10C 12 month period). It is more difficult now because unless the cold period exactly coincides with a month, then you require more potent cold, a longer spell or less of a mild correction in the month. This is exactly what people suggesting a 3C month may be beyond us are saying. This benchmark proves how much we have warmed in winter because so far since 1997 we just havent had depth and length of cold required. We have in that time probably recorded a rolling 31 day period (perhaps as recently as mid/late-Feb - mid/late-March 2006?) but again they are still as rare as ever compared to in the past and thats the point people are trying to make.

I disagree, partly because a "complete month's data" is essentially a random dataset anyway.

Much was made of July 2006 being the hottest calendar month ever, which was true, and the press loved it of course, but few media outlets picked up on the fact that mid-Jul to mid-Aug 1995 was actually a hotter 31 day period. Using calendar months actually allows the media to distort weather stories, not put the "event" (to use a new, favourite buzzword) in a proper, historical context. Which is quite the opposite to what you are saying.

Using calendar months is nothing to do with the media distorting anything. Its more to do with consistency, something very important in meterology, its why we define winter as 1st Dec - 28th Feb and not 21st Dec - 21st March. You say the 31 day period was hotter in 1995, but was the 40 day period, or 20 day period? Just where do you draw the line when you are picking out the days of the 'event'? Its actually this that is little more than moving the goalposts to suit an argument. Its much easier to say that July and August 1995 averaged 18.9C, compared to 17.9C in 2006. Anyone can see in summer 1995 the main body of heat spanned a much longer period, simply from looking at the monthly figures. July 2006 was a perfect example of the hottest weather residing almost exactly in a calendar month, which is what sparked the headlines, but then the media are renowned for quoting partial facts for a good story. 'Heaviest snowfall in 18 years' is a recent one, when in fact for most areas outside London and the south-east it wasnt anything so special.

On a personal note, I thought it was a great shame that we just seem to have missed out on sub-3C last month. I do hope we get it this month and it is indeed the best chance in years.

[Guest North Sea Snow Convection]

It makes a huge difference at the end of the day. Ive never seen anyone on here suggest a 31 day period averaging 3C is probably beyond us. I have however seen members suggest that a 3C month might be (Im not one of them by the way). Theres a very good reason for this however, it is and always has been much easier to record a 31 day period averaging 3C than a calandar month (and indeed, a rolling sub-10C 12 month period). It is more difficult now because unless the cold period exactly coincides with a month, then you require more potent cold, a longer spell or less of a mild correction in the month. This is exactly what people suggesting a 3C month may be beyond us are saying. This benchmark proves how much we have warmed in winter because so far since 1997 we just havent had depth and length of cold required. We have in that time probably recorded a rolling 31 day period (perhaps as recently as mid/late-Feb - mid/late-March 2006?) but again they are still as rare as ever compared to in the past and thats the point people are trying to make.

I'm sorry if I'm missing something but I still just don't see this at all. Best just left at that though :lol:

Taking the 1997 stat I also am not moved by what it is to me an insignificant time period. I just think that people read too much into short term trends to make a bigger picture than there really is. And I can anticipate the reply to that. Again a case of just best agreeing to disagree rather than go round in circles

[gmoran]

A months cold is a months cold whichever way you dress it up. What difference does the start and end day make?! :lol: Apart from satisfying the meteorological annuls of course. But it doesn't somehow make the climate warmer than it is because the dates don't conveniently match a calendar month!

Lets wait till the end of the month and the whole matter should be put to bed anyway regarding the point you are trying to make. But tbh I am rather bemused by the fixation and significance of this 3c winter CET thing. IMO it is pointless trying to place limits/benchmarks all the time and then moving them around once they are subsequently breached to try and keep a certain POV alive. We have had this with the annual rolling CET, but the script wasn't followed, and now it is happening with monthly winter CET's.

I don't think that this sub 3c CET is that significant - only in the minds of those who feel it is so hard for it to be breeched in the UK for some reason. Still, only 3 weeks to go and you will all know the outcome for this month I guess. But it will happen at some stage anyway.

In summary I wasn't actually trying to illustrate how cold it has been at all - merely that some people are speculating about the unlikeliness of something that has already happened under a differing range of datelines!

One of the problems with looking for statistical trends is cherry picking, finding data that suits your model and then attaching significance to it in retrospect. Using calendar months prevents that, such that a calendar month of sub 3'C has more statistical significance than finding 31 consecutive days that are sub 3'C.

Since 1860 we have been able to observe a long term global mean surface warming trend of .08'C per decade; since approx 1975 the warming trend is .16'C per decade. The interesting question at this time is whether the shorter term warming trend since the 1970's has now come to an end, and we can expect some cooling or at least some stasis. In the UK it may mean that synoptics since the early 1990's that have tended to preclude cold and snowy winters are now changing. Answers to these questions may take some years to become apparent.

[Nick H]

a calendar month of sub 3'C has more statistical significance than finding 31 consecutive days that are sub 3'C.

How can that possibly be true?

[Stu_London]

What our weather may be about to prove is that it is very silly or arrogant or both to make statements such as 'a sub 3C month is no longer achievable' or 'a sub 10C year is no longer possible'.

Apart from being statistically unsound arguments in the first place, one is bound to run out of luck at some stage when making such assertive predictions about the future.

From I personal point of view, I prefer not to waste time trying to guess where the statistical floors and ceilings lay in favour of sitting back and watching the weather make monkeys out of those who do

[The PIT]

How can that possibly be true?

Because the Monthly figure will published while the 31 days will lost in annuals of History. Unless you're Mr data that is.

Edited February 9, 2009 by The PIT

[Nick H]

Because the Monthly figure will published while the 31 days will lost in annuals of History. Unless you're Mr data that is.

That is true. But "statistical significance" is a very precise term that is quite unrelated to whether the Hadley Centre records in 28/29/30/31 day periods.

[Stu_London]

How can that possibly be true?

whether it is statiscally more significant or not is a moot point

it is certainly more difficult to achieve the mark between set benchmarks rather than a rolling period

[gmoran]

How can that possibly be true?

It's a case of cherry picking: looking for a pattern in data and then attaching undue significance ( http://en.wikipedia.org/wiki/Cherry_picking ). A calendar month is a sample period uncorrelated to the data you are attempting to attach significance to, it is not biased.

[Nick H]

It's a case of cherry picking: looking for a pattern in data and then attaching undue significance.

A calendar month is a sample period uncorrelated to the data you are attempting to attach significance to, it is not biased.

Calendar months are no more uncorrelated to the data than any other 31 day period. There is as much likelihood that a given heatwave will coincide within 1 Jul-31 Jul as there is that it will coincide within 9 Jul-Aug 8, 16 Jul-15 Aug or 25 Jul-24 Aug. Of course, there is far less likelihood that a heatwave will coincide within a given 31 day period, but that does not make it any more or less "statistically significant" for doing so. Note that I am using that term in the statistical sense.

[Weather-history]

It's a case of cherry picking: looking for a pattern in data and then attaching undue significance ( http://en.wikipedia.org/wiki/Cherry_picking ). A calendar month is a sample period uncorrelated to the data you are attempting to attach significance to, it is not biased.

Wasn't the Met Office, cherry picking when they issued a press release about the start of this winter being the coldest for 30 odd years? It was issued after only about a week into the meteorological winter.

Edited February 9, 2009 by Mr_Data

[Eugene]

What our weather may be about to prove is that it is very silly or arrogant or both to make statements such as 'a sub 3C month is no longer achievable' or 'a sub 10C year is no longer possible'.

Apart from being statistically unsound arguments in the first place, one is bound to run out of luck at some stage when making such assertive predictions about the future.

From I personal point of view, I prefer not to waste time trying to guess where the statistical floors and ceilings lay in favour of sitting back and watching the weather make monkeys out of those who do

Great post, other statements that have come back to haunt the poster have been "Bone dry" last September(still laugh at that one :lol: and we then got some of the worst flooding in September in the first week(haven't seen any flooding in my area like it since) and "Winter is over in early Jan" and we got one of the snowiest weeks in a long time in the first week of Feb, not sure why they continue to make such statements.

[kold weather]

As SM has stated from this point onwards the Jan CET started to increase and whilst today and tomorrow will push up the CET a little we have the rest of the week which will have at the very least some pretty cold nights, even if the daytime maxes are a little higher at times.

[Stu_London]

I'm very confident that we will once again be well below average.

Fairly confident on sub 3, tempered only slightly be the ability of February to warm up quite a bit as the days get longer

Sub 2 couldn't be ruled out with confidence based on the ensembles which go right out to month end

[Duane S.]

What our weather may be about to prove is that it is very silly or arrogant or both to make statements such as 'a sub 3C month is no longer achievable' or 'a sub 10C year is no longer possible'.

Apart from being statistically unsound arguments in the first place, one is bound to run out of luck at some stage when making such assertive predictions about the future.

From I personal point of view, I prefer not to waste time trying to guess where the statistical floors and ceilings lay in favour of sitting back and watching the weather make monkeys out of those who do

Spot on. Far too much arrogance based on very little. I hope these people ARE constantly reminded about their wild and unfounded assumptions, when they are proved to be wholly inaccurate. It is exactly this sort of thing that helps feed the crazed media frenzy that has been going on reference AGW for the past xx amount of years.

That said, the usual suspects will have this winter down as a 'blip' no doubt. There's 'blips' and then there's 08/09 'blips', very different indeed.

[AtlanticFlamethrower]

Calendar months are no more uncorrelated to the data than any other 31 day period.

No, but what he was saying was you can't cherry-pick 31 periods in the winter months, out of almost 60! available 31 day periods that fall within winter months, this season and compare with month periods in previous years which were only measured in three specific 31 day periods.

It's far more likely that a cold spell will fall within one of 60 than one of 3. For a fair comparison you need to compare the same set of 31 day periods.

Back to Feb CET... Right now I'm beginning to worry my 2.3C is going to be over 1C out.

[reef]

Spot on. Far too much arrogance based on very little. I hope these people ARE constantly reminded about their wild and unfounded assumptions, when they are proved to be wholly inaccurate. It is exactly this sort of thing that helps feed the crazed media frenzy that has been going on reference AGW for the past xx amount of years.

That said, the usual suspects will have this winter down as a 'blip' no doubt. There's 'blips' and then there's 08/09 'blips', very different indeed.

You talk about wild and unfounded assumptions, but then go on to say theres 'blips and theres winter 08/09 blips', a contradiction if I ever saw one.

I dont remember anyone saying 3C is impossible, nor a sub-10C year, but rather that looking at the trends it was probably very close to being so. Note, the assumptions by these people were based on evidence to hand (i.e data from the past 11 years), not some wild prediction without a shread of evidence (i.e predicting a change to a colder winter trend over the next few years), something which cold biased members are often guilty of.

At the end of the day, with a background warming of 0.5-1.0C at most compared to 25 years ago, it would be absolutely ludicrous to suggest we couldnt have a sub-3C month ever again. I still feel I must remind people, that although looking at the models theres an increasing chance this month might be the one, that we still havent breached that sub-3C mark. Nor have we had a colder winter than 1995/96 yet.

Speaking of which, as it stands, this winter is a blip. That is unless we see some more colder winters in the next 10 years. If we do, then there really would be evidence that winter patterns in the UK had changed. For now, its a rare colder winter (just as 1990/91 and 1995/96 were) in a period of ultimately milder winters. Im sat on the fence with regards to whether its due solely to the warming of our climate, a natural climatic phase or both. Im also not niave enough to let my 'hope for a return of winters of old' cloud my judgement, as many do on here.

Edited February 9, 2009 by reef

[Snowly does it]

Given that sub 3 at least looks a possibility, 2 figures to watch for

2.5 - would be the coldest expressed as a deviation against 71-00 mean in the 21st Century (beating Dec 2008 which was 1.6 below average)

2.2 - first month 2 degrees below average this century, would make winter 08-09 colder than 95-96 and would be the coldest winter month (and of course any month) since Feb 91

[West is Best]

I dont remember anyone saying 3C is impossible [...] but rather that looking at the trends it was probably very close to being so.

Errrrm, unless you think that language never signifies that which it denotes this comes as close as you'll get ...

As I've been saying for a few years now, and despite the fact that cold fanatics don't like it, we STILL HAVE NOT had a sub 3.0C month since 1997. For sure we had a fortyish day spell below that mark, but it was punctuated by some warmth that scuppered the calendar average. All that does is ratehr align with my contention: sustained cold, as sustained and as cold as it occasionally used to be, is now a thing of the past I think ... Every year the same people rail, every year we do not get a sub 3C month: in fact it is now 35 winter months since our last sub 3. Can anyone tell me when we last had a run that long?

I still feel I must remind people, that although looking at the models theres an increasing chance this month might be the one, that we still havent breached that sub-3C mark. Nor have we had a colder winter than 1995/96 yet.

Speaking of which, as it stands, this winter is a blip. That is unless we see some more colder winters in the next 10 years.

This takes us into climate change issues. I'm increasingly satisfied that this is no blip. That the CET shows a cooling trend, that the last 10 months have shown this to be accelerating, and that this winter is part of a trend, not a blip, seems to me to be becoming more and more evidential. The most one can say is that it's a trend which will iron itself out in time, but to dismiss it as a mere blip in the light of preceding evidence seems to me to border on the disengenuous. Incidentally, this winter synoptically connects to the preceding summer in terms of the jetstream pattern: in both cases the polar jet has been much further south than that to which we have been accustomed in the past 15 years. I would be fascinated to hear SF's comments on that point, because I can recall agreeing with him some winters back that this was one of the key factors to screw up cold winter prospects for the UK during those long mild years.

Tonight's model outputs suggest to me that sub-3C is very much 'on'

Edited February 9, 2009 by West is Best

[SteveB]

Just to add

It was the Summer of 2007 that the cooling trend really took hold. And the Summer of 2008 also turned out to be cool.

Now winter 2009.

[Snowly does it]

A further stat..

The last 20 days of February need to average 9.2 to prevent winter 08/09 being the third season in a row below the 71-00 average (if Jan is confirmed at 3.3) and 7.4 to prevent it being the second season below the 61-90 average in a row (again thats if Jan is confirmed at 3.3)

[kold weather]

Pretty much certain now we are going to come in below the 61-90 average from the looks of things SM given at least the next 5 days look pretty cold still.

I'd also say sub 3C is looking very possible indeed now, Jan which came in at 3.3C was at 1.4C by the 13th and whilst we will see a rise over the next day or two once that over we have at least another 4 days of average temps probably between 0-1C and thats if the most progressive idea is correct, its more then possible the cold will hold way longer then that...

[Duane S.]

I dont remember anyone saying 3C is impossible, nor a sub-10C year

I do B) No doubt worded with convenient get-out clauses though.

Errrrm, unless you think that language never signifies that which it denotes this comes as close as you'll get ...

This takes us into climate change issues. I'm increasingly satisfied that this is no blip. That the CET shows a cooling trend, that the last 10 months have shown this to be accelerating, and that this winter is part of a trend, not a blip, seems to me to be becoming more and more evidential. The most one can say is that it's a trend which will iron itself out in time, but to dismiss it as a mere blip in the light of preceding evidence seems to me to border on the disengenuous. Incidentally, this winter synoptically connects to the preceding summer in terms of the jetstream pattern: in both cases the polar jet has been much further south than that to which we have been accustomed in the past 15 years. I would be fascinated to hear SF's comments on that point, because I can recall agreeing with him some winters back that this was one of the key factors to screw up cold winter prospects for the UK during those long mild years.

Tonight's model outputs suggest to me that sub-3C is very much 'on'

Good post WIB.

One of the main factors always given out about why are winters have been so poor of late (aka the 'even larger teapot') is the northern extent of the jetsteam. This has been notably further south over the past couple of years. That's a big 'blip'.

[acbrixton]

An interesting post Richard.

Looking at the GFS 12z a CET of about 0.7c looks likely by 14th and even by the 2/3 mark on 19th 0.7c is quite achievable (subject to the usual FI caveats). If that transpires then the likely range for the month end is, say, 1.5c [requiring the last third to record about 2.8c] and 3.5c [about 7.7c for the last third]. My guess now is that a final figure will be perhaps between 2.0c and 2.5c, i.e. sub 3 looking probable. I have to admit that the prospect of a (comfortably) sub 3 month had seemed to me to be very unlikely though not impossible.

To beat 1985/6 or 1995/6 requires a February CET of 1.6c: possible but still quite unlikely.

You are right to make the point about the effect of the tendency to a southerly tracking jet in the last 18months plus but just as it is too early to dismiss that tendency as a blip it surely follows that it is too early to proclaim with any degree of certainty that a new trend or pattern has been established. Like Reef I would want to see if that trend was discernible over, say, a 10 year period.

kind regards

Andrew

[West is Best]

To beat 1985/6 or 1995/6 requires a February CET of 1.6c: possible but still quite unlikely.

Very interesting post back Andrew.

Just to check on your figure there: 1995/6 winter is 'beaten' by a 2.2C month this month. 1985/6 is 'beaten' by 1.8C.

I'm not sure 1985/6 is a worthwhile comparison in a way because it was really comprised of one cold spell from mid Jan to end of Feb. The -1.1C Feb figure is freakishly cold but bone dry, and such a big contrast with that December at a whopping 6.3C.

At the end of the day though, beating a winter is slightly arbitrary in some ways. I'm more interested to see if this sub 3C February can be done away with once and for all.