How exceptional is this year within the UK?

[ukmoose]

The year has been interesting in some ways but not surprising and I believe it is an indication of what is to come.

The summer period was long (lasting from May to October) and largely very warm indeed. This pattern will continue - warm or very warm summers starting early and lasting well in to the autumn season with the summer heat reluctant to move and difficult to dislodge. Generally very dry with occasional thunderstorms breaking the drought.

The winter season was interesting in that there were several attempts made to introduce cold air to our shores. It almost made it but never quite did and largely failed to deliver and we ended up with nothing much. The significance is that that is as cold as it will get in these shores in the foeseeable future. Last winter was 'cold' by modern standards. Don't expect a repeat.

The only difficult thing to tell at the moment is whether our winters will be wet or dry. There is every indication now that this winter will turn out to be drier than first thought. Maybe not as dry as last winter but dry nonetheless and mild. No months will be below the long term average.

I know this will not be a popular post but there it is.

Sorry.

Moose

[Snowly does it]

I know this will not be a popular post but there it is.

Sorry.

Moose

Have you been playing with your time machine again Moose?

You may be right of course, but you might also be wrong.

Its exceptional this year the sheer depression people are showing for winter after a below average start to November, I know that much.

[Stratos Ferric]

Have you been playing with your time machine again Moose?

You may be right of course, but you might also be wrong.

Its exceptional this year the sheer depression people are showing for winter after a below average start to November, I know that much.

SM,

If you go see the chart I posted most recently on here (probably a page or two back now) you will see that one very clear interpretation of that chart is exactly what Moose has just said. Unless we do get a switch back in the warming trend (I'm not saying we can't, but equally I can't see where it's going to come from so personally I have to doubt it) then it really is the case that last winter is currently about as bad as it can get. In the current version of the even larger teapot a freak along the lines of 1963 relative to its time would, nowadays, probably yield no worse than a 1996. And with each passing winter without a reversal the baseline nudges ever higher.

There's a bit of a danger that we get like snooker commentators, where "every frame is vital" to the result of the match, but after last year's (by recent standards) cool winter, what happens this year is going to be of very great interest.

[Snowly does it]

SM,

If you go see the chart I posted most recently on here (probably a page or two back now) you will see that one very clear interpretation of that chart is exactly what Moose has just said. Unless we do get a switch back in the warming trend (I'm not saying we can't, but equally I can't see where it's going to come from so personally I have to doubt it) then it really is the case that last winter is currently about as bad as it can get. In the current version of the even larger teapot a freak along the lines of 1963 relative to its time would, nowadays, probably yield no worse than a 1996. And with each passing winter without a reversal the baseline nudges ever higher.

There's a bit of a danger that we get like snooker commentators, where "every frame is vital" to the result of the match, but after last year's (by recent standards) cool winter, what happens this year is going to be of very great interest.

I'm sorry Stratos but I don't buy into it, last winter is by no means as bad as it can get. I'll buy as bad as its LIKELY to get. I don't expect a cold winter, I don't expect regular cold winters but I do have objections to statements like 'no month will be below average'. We have seen that is patently not a 'rule' as in Nov 05, Mar 06 etc.

I have never argued that the world is not getting warmer, and that the UK is not also feeling this effect, but a colder winter than last year is quite possible. A 47 or a 63 is out of reach imo.

[Stratos Ferric]

I'm sorry Stratos but I don't buy into it, last winter is by no means as bad as it can get. I'll buy as bad as its LIKELY to get. I don't expect a cold winter, I don't expect regular cold winters but I do have objections to statements like 'no month will be below average'. We have seen that is patently not a 'rule' as in Nov 05, Mar 06 etc.

I have never argued that the world is not getting warmer, and that the UK is not also feeling this effect, but a colder winter than last year is quite possible. A 47 or a 63 is out of reach imo.

SM, I don't disagree, and have certainly never said we can't get below average months; the key points though are how often, how far, and - and this is the vital bit just now - which average? It's more a point of semantic pedantry than real substance to debate whether it's as bad as "it can" rather than "is likely" to get. For what it's worth you're quite correct re the latter, but only "for now".

I don't know whether you've actually looked at the chart or not, if you haven't it's worth a look. For sure, the baseline is drawn "by eye", but only once since 1960 is that line breached, and the upward gradient on it is severe by any measure. Yes, there's around a 1:20 - 1:30 chance of a month dipping below the line, and within that short cold spells are still possible even in an average month. I think most people's frame of reference (witness the CET guess thread) is the 30 year mean - in a warming trend a long average is flawed unfortunately, and the longer it is the more flawed it becomes, and doubly so if the rate of warming is significant and / or increasing. The current ten year averages for winter months are around 1C warmer (at least) than the 30 year means - which means around 2C warmer than the start of the series, and 1-2C in winter is a big leap in a regime like ours where much cold weather is very marginal.

We don't really notice 5C of warming at the pole, because it gets down to -35 / -40C quite frequently, and below freezing is below freezing, but here in the UK it's of the essence.

The reality is that there are still a lot of people whose notions of a cold winter, or even a cold winter month, are already almost certainly the stuff of the ether.

[Snowly does it]

I have looked at the chart yes, and stark reading it is.

I don't know what to add really. Your comment on the other thread about an upward shift of the classic Bartlett interests me greatly, I wonder what short term changes in 'classic' synoptics might do as a trend within the general trend. Some of them might presumably provide false hopes of a reversal?

I guess the worst thing that could happen right now would be a Pinatubo or worse masking where this is and ends up.

[Stratos Ferric]

I have looked at the chart yes, and stark reading it is.

I don't know what to add really. Your comment on the other thread about an upward shift of the classic Bartlett interests me greatly, I wonder what short term changes in 'classic' synoptics might do as a trend within the general trend. Some of them might presumably provide false hopes of a reversal?

I guess the worst thing that could happen right now would be a Pinatubo or worse masking where this is and ends up.

Dunno; a big volcano would help, and we're probably overdue one.

There is also the possibility that there is an accelerated effect locally in the warming, and that it really is just a case of "bad (and 'unlucky') synoptics" as opposed to bad but technically what might be expected synoptics. My proposition is rational, but it isn't the only one - for sure.

The other factor is that there is a realistic base above which winters, for now at least, cannot go. In amongst the amusing spat with AFF on here there is a critical point about winter which rather undermines ANY theory regarding "rebound". Above all else the climate is driven by inbound solar radiation; once the sun dips below the horizon / gets low in the sky insolation falls so low that surface and air cool. This lack of warming means that there has to be a point above which winter cannot warm, unless or until the oceans start to boil. On my chart the baseline cannot continue upwards for ever.

I know Philip E subscribes to a twenty-odd year cycle, and we're bearing down on it now. Last chance saloon it may be in some senses, but be reassured that we haven't that cycle yet. When we do, and if we then get nothing remotely cold, then we should be very concerned for winter. We can't categorically write it off yet though, for all that I personally am holding a more bullish view (this, by the way, is in part a case of "expect the worse, hope for the best").

[The Enforcer]

It's been said before that winter months are becoming homogenous, at least in terms of CET and that chart certainly doesn't disprove that assertion. One could be forgiven for thinking in 20 years every winter month without fail will come in at around 5.5-6C.

[SMU]

Dunno; a big volcano would help, and we're probably overdue one.

There is also the possibility that there is an accelerated effect locally in the warming, and that it really is just a case of "bad (and 'unlucky') synoptics" as opposed to bad but technically what might be expected synoptics. My proposition is rational, but it isn't the only one - for sure.

The other factor is that there is a realistic base above which winters, for now at least, cannot go. In amongst the amusing spat with AFF on here there is a critical point about winter which rather undermines ANY theory regarding "rebound". Above all else the climate is driven by inbound solar radiation; once the sun dips below the horizon / gets low in the sky insolation falls so low that surface and air cool. This lack of warming means that there has to be a point above which winter cannot warm, unless or until the oceans start to boil. On my chart the baseline cannot continue upwards for ever.

I know Philip E subscribes to a twenty-odd year cycle, and we're bearing down on it now. Last chance saloon it may be in some senses, but be reassured that we haven't that cycle yet. When we do, and if we then get nothing remotely cold, then we should be very concerned for winter. We can't categorically write it off yet though, for all that I personally am holding a more bullish view (this, by the way, is in part a case of "expect the worse, hope for the best").

SF-

With this summer delivering the sort of CET results that it has, based on the H5 & H85 anomalies that we have tapped into the ol' Maritime climate comes under threat, to be replaced by a possible 'Continental Climate'-

Remember that overall this year the jet has been Slower, the mid lattitude blocking has been more significant-

Besed on that 'assumption' it is likely that the Winter could take a similar continental type flow- I sure you mentioned Warm & Dry- but It could also be cool/Cold & dry depending on how the Winter feedbacks work-

Remember should your jet move into 65/70N teritory then there should still be a still be a tongue of polar air coming under and Mid lattitude block....

S

Edited November 7, 2006 by Steve Murr

[AtlanticFlamethrower]

Stratos Ferric

But March wasn't very cold. It was only about 1C below normal. Cool, or slightly cold, but not very cold.

Even more strange you would consider March cold enough to call it an "outlier."

March was NOT exceptional. I the all time ranking it was around 219th as I recall. I.e. in the middle third of all time, mid table, around average. Certainly NOT exceptional for its cold.

It was exceptional for recent times - as was the whole winter, which showed a "classic" winter is still possible synoptically as it was 40 years ago.

I said it was an outlier, and explained precisely why based on relative ranking.

March happened. It was not an outlier on that fact alone. Nothing that happens can be an outlier. It happened.

I think we have repeated several times that the entire thread is about the UK.

You -

I'm sorry to say, it might be winter that's the anomaly in the christmas pudding.

But it isn't remarkable. The charts suggest it is all normal. Had there been a huge meltdown at the pole this summer fair enough, but so far as I recall it was as warm up there as it has been in recent years. Do you have any data to suggest that the pole experienced the excessive warmth that the UK did, other than what I am having to take as your own presumption at present? And in any case, did this cause the ice to retreat to a lower extent than normal. It's "no", "no" and "no" however you cut it.

You have claimed the UK had an exception summer and that "winter is the anomaly in the christmas pudding."

Are you saying you only meant this phrase in the context of UK? That this phrase "winter is an anomaly in the christmas pudding" had no bearing on the other 510 million km² of Earth?

Have you devolved some of the more important decision making powers your finger tips? Because what I think you mean to say and what you type are two different things.

Yes March did happen, but to suggest, as you seem to be doing, that we can only assess a moment in the context of what preceded it, and not what followed, is quite simply stupid.

You think so?

Unless your best friend is Watson or your surname is Marple, to understand the cause of something you need to look what happen before the event because that explains what came after.

So no, I don't think it's stupid to say the heat of July or September in any way impacts our assessment of March's CET 4 and 6 months earlier.

Compared to the hot months March was much colder. March also was at a different time of year.

I want to know what you're trying to prove calling March an "outlier." It makes no sense to me.

I've posted - and you're certainly not providing any FACTS of your own - March was the coldest month this year in relative terms.

Why should I trouble you with more facts when we can't even agree on the significance of "cause and effect"?

You seem to be arguing that overall March was normal, and that at least five of the other months in the year are outliers (actually, you're both arguing normal and exceptional at the same time, which is not unusual so far - quite how you arrive at this I do not know: it was slightly cooler than average - in the middle third; in a year that has registered two hottest ever months, and one third hottest, I am at a loss to understand how even a village idiot could arrive at this conclusion).

I'm not saying any month is an outlier. It's a meaningless phrase. How can something that happened be an outlier? There's variation. There's extremes. There's average. But an outlier implies March was an anomaly which even you admit wasn't.

Analysis of the RANK of each month shows that nowadays the majority of months are in the warm half. Statistically, if we weren't warming, this sequence would be so unlikely as to be impossible. Clearly we are warming, and as we do so months as relatively cold as March will become increasingly infrequent.

Warming is happening. It's measured. That's all we know. You additionally want to claim this means March will become "increasingly infrequent." Might summer not be a warm cycle within a warm cycle within a warm cycle? Might not winter 2006 be a cold cycle within a warm cycle - in which case if the cold cycle within a warm cycle continues cold Marches will get more frequent, as long as the cycle persists?

You are making quick, easy assumptions. Why? Why must March be hidden in a box of "outliers"? Can't March's CET have its own cause? Must there be one monolithic cause?

Hence my final musings. You don't have to agree, but you aren't putting up any compelling argument to the contrary I'm afraid other than a sense of not wanting to accept or believe in the likely consequences IF the current warming continues.

You're making assumptions about the causes of temperatures in UK that lie outside the nautical boundaries of the UK. "Current warming" is an incredibly leading phrase - you can be more precise.

You keep coming back to the UK and ice at the pole. The last time I checked we were, as I suggested previously, well over 1000 nm from any ice. AND you're still not explaining just what is remarkable about the change in sea ice this year which appears to me to be normal. Are you perhaps assuming that the pole this year was much warmer than last year? Do you have any facts? So far you are a factual equivalent of the Gobi Desert - rather cold and barren.

Off topic!

All you are doing is repeating the same old mantra, you're not actually progressing the argument or discussion here. You're not responding to the challenges presented other than in rhetorical and opinionated fashion. I'm all for argument and debate, but we'll make better progress if you can actually provide rational argument for your position, data and / or other evidence.

This is about your interpretation of facts. A Chef might put all the right ingredients on the table but if the gravy is over the strawberry mousse and the sticky toffee cake is blended with prawns...

FOOTNOTE:

Actually, stupid of me, I've just noticed the piccies on your avatar and signature. Now I understand. It's not that you can't understand, it's that you don't want.

Winter avatar means - big snow / cold is possible in UK. Only an idiot would deny this when major cold shots of recent years heve missed UK by a whisker. I accept warming is happening and there is less Arctic ice but exceptional temperatures we've had is not GW, rather of the sorts of a warm cycle within a cycle. I'm mean this is obvious unless you want to claim GW is more than 0.5C. Winters can still deliver big time with the right set up - even at +0.5C most of the historic great evens would still deliver snow.

[Mr Sleet]

I can say that - the fact is that in 2006 March (ok - and Feb) have been big outliers. The next coldest month was January, which was 106th warmest - well in the top third of all time. In the context in which I used it, and with the addition of February, my use was perfectly acceptable.

Not sure what the density of thermometers has to do with anything; you'd better elaborate for me.

As for the pattern of more blocked winters, one swallow does not a summer make, nor does one cold month a long trend make. Let's not get too carried away, it'll probably only end - as it often seems to on here - in tears.

It's sad to see you reduced to slinging insults at AFF.

You can't pick and choose which "outliers" suit your argument better, it just confirms that you have a bias toward wanting a milder outcome, and this colours your thinking.

I could say that the annual CET was "saved" from being lower than 2005 by unusual warmth in July,Sept and October. If it weren't for these months we would be looking at a lower CET than 2005, continuing the downward trend.

Edited November 8, 2006 by Mr Sleet

[Thundery wintry showers]

SF's graphs are perfectly reasonable and show a current trend towards homogeneous winter warmth. However, while the point is valid that the potential for cold extremes is declining markedly, I don't think that relationship can be continued forever; as Enforcer said, if it kept up, every month would be in the 5.5-6C range.

I don't think that February 2006 is the coldest it can get; for example, March 2006 was cold relative even to the 1961-90 average (5.0 as measured against a baseline of 5.7 CET). In view of the fact that 1-22 March was about 2.5-3C down on the 1961-90 average, I suggest that maybe 1-1.5C CET is the lowest we could possibly attain nowadays. Nonetheless, this is still a much greater rise in the minimum baseline (relative to, say, 1963) than that in the maximum baseline, so there is still some truth in what Stratos is saying.

What needs to stop, though, is the trend towards excessive warming of the Arctic interior; we usually struggle to get significantly below average temperatures when that occurs.

[weathernewbie]

SF's graphs are perfectly reasonable and show a current trend towards homogeneous winter warmth. However, while the point is valid that the potential for cold extremes is declining markedly, I don't think that relationship can be continued forever; as Enforcer said, if it kept up, every month would be in the 5.5-6C range.

I don't think that February 2006 is the coldest it can get; for example, March 2006 was cold relative even to the 1961-90 average (5.0 as measured against a baseline of 5.7 CET). In view of the fact that 1-22 March was about 2.5-3C down on the 1961-90 average, I suggest that maybe 1-1.5C CET is the lowest we could possibly attain nowadays. Nonetheless, this is still a much greater rise in the minimum baseline (relative to, say, 1963) than that in the maximum baseline, so there is still some truth in what Stratos is saying.

What needs to stop, though, is the trend towards excessive warming of the Arctic interior; we usually struggle to get significantly below average temperatures when that occurs.

TO settle the argument does anyone know of a graph of CET, showing from when records began or a sensible date, to the present day. That would clearly show what is being discussed. (if presented in a clear manner, i.e. pictorially)

Edited November 8, 2006 by weathernewbie

[Nick F]

TO settle the argument does anyone know of a graph of CET, showing from when records began or a sensible date, to the present day. That would clearly show what is being discussed. (if presented in a clear manner, i.e. pictorially)

http://www.metoffice.com/research/hadleyce...bsdata/cet.html

[Nick H]

It's sad to see you reduced to slinging insults at AFF.

You can't pick and choose which "outliers" suit your argument better, it just confirms that you have a bias toward wanting a milder outcome, and this colours your thinking.

I could say that the annual CET was "saved" from being lower than 2005 by unusual warmth in July,Sept and October. If it weren't for these months we would be looking at a lower CET than 2005, continuing the downward trend.

Totally agree, Mr S. I'm not sure why SF describes March as an "outlier". He correctly points out that it is 219th in the all-time list of Marches (the middle third), even describing it (equally correctly) as "not exceptionally cold", which makes his labelling of March as an outlier all the more puzzling. The OED defines an outlier as "a result differing greatly from others in a sample". It's difficult to see how March fits that description. If any month in 2006 has been an outlier, surely July is a better example than March.

The CET was 1.2C above par for this year at the end of October. Of that differential, exactly 50% is due to July, a percentage that increases if we add into the pot the first, chilly, week of November. Indeed, the running yearly CET was below par during virtually the entirety of the first half of this year, often by as much as one degree C.

Edited November 8, 2006 by Nick H

[Osbourne One-Nil]

I've been reading this, and if you don't understand why SF describes what he does how he does by now, you're NEVER going to. So how about we just let it go now? In fact, no....let it go now (no how about).

[Stratos Ferric]

It's sad to see you reduced to slinging insults at AFF.

You can't pick and choose which "outliers" suit your argument better, it just confirms that you have a bias toward wanting a milder outcome, and this colours your thinking.

I could say that the annual CET was "saved" from being lower than 2005 by unusual warmth in July,Sept and October. If it weren't for these months we would be looking at a lower CET than 2005, continuing the downward trend.

Mr S,

On the one hand we should both be ashamed of ourselves, but on the other I cannot help feeling that I had made it perfectly clear the sense in which I was descrbing my argument, and the bounds, and to that extent my response to hime was atually rather understated. That's not to say there aren't other arguments, but AFF was not presenting any (other than one that was not relevant, and which caanot in any case be substantiated by facts - relating to ice growth in the arctic), and he was completely failing to accept what was a perfectly sound rationale, in the context of 2006, for describing March as the outlier. ANd last nights post is pitted with contradictions; not least the argument that one cannot compare any occurence to what follows, whilst then going on to himself make an argument regarding March whilst referring to July. One could be forgiven for being exasperated at someone calling into question what is clearly (as OON suggests) a cogent argument whilst basing their challenge on inconsistent logic and irrelevant facts.

I do not have a bias towards warm, not in the sense that many on here do towards cold i.e. they WANT cold. I would like cold, but all I ever try to do is present facts / data and an interpretation thereof. The facts and data are there and all should feel free to interpret them alternatively and / or to present alternative data. Simply saying "no, it isn't so" is not compelling argument, and if nothing else speaks ill for modern education if that's the best we can produce by way of advocacy.

To this extent see TWS's response, which builds nicely on the points I was making. And I would concur with his correction / clarification; in absolute terms March 06 was not as cold as it can get in winter, but relative to norm ON THE BASIS OF THE SECOND PLOT I PROVIDED, it probably was about as cold. This variance of around 1-1.5C below norm if applied to Feb, would suggest a potential cold threshold of around 2.8C based on a ten year rolling mean, or 1.8C based on a thirty year mean. Given that we're warming the shorter mean is the more likley reference point at present.

I find it as unpalatable as anyone on here, but that is not going to allow me to deviate from a rational and objective assessment.

The value of these fora is greatly reduced if they are not open for argument and debate, but argement and debate does need to be based on certain principles if it is to be constructive.

[WhiteFox]

I'll try....

Here we have a graph showing the CET by month plotted against the difference from the 61-90 CET for that month:

As you can see, most of the months are above average. However, is this significant?

Take a look at the next chart:

This shows the average CET difference for the year, 0.125 and plots the monthly differences on the same graph. Based on a 95% confidence level, we can see that the Interval is between 0.41 to 2.11 above the mean figure for departure from CET for the year.

Based on this figure, March 2006 is outside the 95% confidence level and so can be said to be statistically significantly different from the mean. July, September and October are also significantly different from the mean. Based on these figures you could say that March was significantly below the average difference for the year, whereas July and Sep are exceptionally above.

If we go back over a two year period, from Nov 2004 to Oct 2006, we see a slightly different result.

The 95% confidence interval for this period is 0.63 to 1.53 degrees either side of the average departure from the CET for the period of 1.08oC.

The results are:

Jan - Apr are all outside the 95% confidence interval on the downside, with March being 4.16 intervals away from the mean. June, July, September and October are all outside the 95% interval to the upside, with Jul being 5.58 intervals away from the mean.

I hope this makes some sense and also goes to demonstrate to some degree just how exceptional this year has been for the UK.

[Stratos Ferric]

I'll try....

Here we have a graph showing the CET by month plotted against the difference from the 61-90 CET for that month:

As you can see, most of the months are above average. However, is this significant?

Take a look at the next chart:

This shows the average CET difference for the year, 0.125 and plots the monthly differences on the same graph. Based on a 95% confidence level, we can see that the Interval is between 0.41 to 2.11 above the mean figure for departure from CET for the year.

Based on this figure, March 2006 is outside the 95% confidence level and so can be said to be statistically significantly different from the mean. July, September and October are also significantly different from the mean. Based on these figures you could say that March was significantly below the average difference for the year, whereas July and Sep are exceptionally above.

If we go back over a two year period, from Nov 2004 to Oct 2006, we see a slightly different result.

The 95% confidence interval for this period is 0.63 to 1.53 degrees either side of the average departure from the CET for the period of 1.08oC.

The results are:

Jan - Apr are all outside the 95% confidence interval on the downside, with March being 4.16 intervals away from the mean. June, July, September and October are all outside the 95% interval to the upside, with Jul being 5.58 intervals away from the mean.

I hope this makes some sense and also goes to demonstrate to some degree just how exceptional this year has been for the UK.

WF,

Useful stuff. What data set have you used to calculate the confidence intervals? I only ask because with such a small data set, and assuming you've only used the ten months from this year as base data, the degrees of freedom are small and I'd expect the intervals for 95% confidence to be much wider than the ones you seem to be suggesting.

Can you clarify what you mean by

The 95% confidence interval for this period is 0.63 to 1.53 degrees either side of the average departure from the CET for the period of 1.08oC.

. 95% would normally be a line at the same value +/- the mean, though I think this assumes a normal distribution strictly speaking.

Also, I'm not sure that comparing each month to the average for the year is necessarily valid. Taking a silly extreme example, if we had one month at 12C above par, and all others at par, the mean would be +1C; on this basis there would be eleven cool months and one warm one. That's why I prefer always to take a number of cuts of data to see a pattern from several dimensions. Yours is a different treatment, and interesting and additive, so long as the use of the maths is technically correct.

The reason why I chose the analysis I did was to present each month in the context of its family for the entire CET; i.e. all Januarys compared, etc. This was the basis for saying March was the outlier, because it was the ONLY month this year (to date) that came in as lower half by all time rank.

It is fair to show that Feb also came in cooler than average, and on this basis a different view of the year can be taken, but as I recall it was not lower half by rank - which if nothing else shows that the distribution of mean temps for Feb is NOT normal.

[WhiteFox]

Can you clarify what you mean by

. 95% would normally be a line at the same value +/- the mean, though I think this assumes a normal distribution strictly speaking.

Also, I'm not sure that comparing each month to the average for the year is necessarily valid. Taking a silly extreme example, if we had one month at 12C above par, and all others at par, the mean would be +1C; on this basis there would be eleven cool months and one warm one. That's why I prefer always to take a number of cuts of data to see a pattern from several dimensions. Yours is a different treatment, and interesting and additive, so long as the use of the maths is technically correct.

The reason why I chose the analysis I did was to present each month in the context of its family for the entire CET; i.e. all Januarys compared, etc. This was the basis for saying March was the outlier, because it was the ONLY month this year (to date) that came in as lower half by all time rank.

It is fair to show that Feb also came in cooler than average, and on this basis a different view of the year can be taken, but as I recall it was not lower half by rank - which if nothing else shows that the distribution of mean temps for Feb is NOT normal.

I'm a bit rusty at Confidence analysis as it's been a long time! The figure for the 10 months of 2006 would be 1.25 +-0.85 where 1.25 is the average difference between actual CET and average CET (1961-90). This is only based on the ten months of the year as it is the topic of the thread. Obviously taking a longer period would produce different results.

I realise that using the departure from average CET difference is unusual, but as we were discussing whether or not March is an outlier in the context of this year, I thought it may be a useful comparison. Clearly March is an outlier (which proves the point) as are July, September and October (although I'm sure that surprises no-one). Of course, March may or may not be outside the 95% interval if we took the entire CET dataset; my hunch would be that it is not or very slightly.

[shuggee]

*Ventures into stats thread quietly*

The above graph for me, illustrates how out of step with the year as a whole August was. I think in the list of remarkable things to happen in 2006, the disappointing August would rank up there.

[ukmoose]

I'll try....

Here we have a graph showing the CET by month plotted against the difference from the 61-90 CET for that month:

As you can see, most of the months are above average. However, is this significant?

Take a look at the next chart:

This shows the average CET difference for the year, 0.125 and plots the monthly differences on the same graph. Based on a 95% confidence level, we can see that the Interval is between 0.41 to 2.11 above the mean figure for departure from CET for the year.

Based on this figure, March 2006 is outside the 95% confidence level and so can be said to be statistically significantly different from the mean. July, September and October are also significantly different from the mean. Based on these figures you could say that March was significantly below the average difference for the year, whereas July and Sep are exceptionally above.

If we go back over a two year period, from Nov 2004 to Oct 2006, we see a slightly different result.

The 95% confidence interval for this period is 0.63 to 1.53 degrees either side of the average departure from the CET for the period of 1.08oC.

The results are:

Jan - Apr are all outside the 95% confidence interval on the downside, with March being 4.16 intervals away from the mean. June, July, September and October are all outside the 95% interval to the upside, with Jul being 5.58 intervals away from the mean.

I hope this makes some sense and also goes to demonstrate to some degree just how exceptional this year has been for the UK.

Depends which CET values one looks at. There's a bit of a problem here as some sources say the winter months were below average and some don't.

Take the humble Beeb for instance. Accoording to their data: January was + 0.6 the 1961 to 1990 average; February +0.2 and March + 0.6 Most people seem to be under the impression that we had a 'cold' winter last year. By recent standards that's true, but by measuring agiinst long term trends, it isn't.

We have a problem if we can't agree whether last winter was above or below the long term mean.

Are temperatures rising throughput the year (all seasons)? Yes. Has this been a trend for a number of years? Yes. Is it statistically significant? Yes. Can we pretend otherwise? Yes.

Moose

[Snowly does it]

Depends which CET values one looks at. There's a bit of a problem here as some sources say the winter months were below average and some don't.

Take the humble Beeb for instance. Accoording to their data: January was + 0.6 the 1961 to 1990 average; February +0.2 and March + 0.6 Most people seem to be under the impression that we had a 'cold' winter last year. By recent standards that's true, but by measuring agiinst long term trends, it isn't.

Edit - meh, useless point which I have deleted.

61-90 the last winter was margainally warmer, 71-00 margainally cooler. Not that impressive a winter temperature wise, but in recent years not so bad. The world is warming, guess thats what you get.

Edited November 8, 2006 by snowmaiden

[ukmoose]

Edit - meh, useless point which I have deleted.

61-90 the last winter was margainally warmer, 71-00 margainally cooler. Not that impressive a winter temperature wise, but in recent years not so bad. The world is warming, guess thats what you get.

Sorry SM you've lost me with this post. What are you saying actually?It's warmer now than whenever...so...if it's not a really really mild winter, it's not so bad after all...?

Moose

[Snowly does it]

Sorry SM you've lost me with this post. What are you saying actually?It's warmer now than whenever...so...if it's not a really really mild winter, it's not so bad after all...?

Moose

I am not sure either, I'm rather tired tonight.

Ignore it.