General Climate Change Discussion Continued:

[Guest North Sea Snow Convection]

This isn't some kind or competition, its just various people giving various opinions on what they see/think/believe is happening.

For example, I go surfing in the sea off Cornwall and have done for many years, I can't say I've noticed the sea getting any warmer, nor the temperatures on shore likewise, ergo, from my point of view there is little to suggest this warming is taking place, indeed 12 years ago when I was taking my GCSEs it seemed a lot warmer than it has over those last 12 years, IMBY I know, but isn't that the point, MBY does make up part of this globe.

Quite. Couldn't agree more.

There are continual charges of not giving evidence to back up thoughts. The trouble is I already believe I have - as far back as last year or thereabouts I can't remember exactly when. I pasted some links and some pertinent texts about the possible problems with the assumed positive feedback and IPCC considerations as well as stratosphere and cloud feedbacks etc etc. Most of if was at the time ignored, ironically, having been requested at the time after being accused of the same thing as now!

It gets very tiresome having to go back, dig out, and/or repeat huge long posts of that sort just to satisfy someome who maybe wasn't privvy to see them the first time around. Or re-satisfy others all over again. And who would quite probably dismiss them anyway even if you did. That is not being funny, just being realistic. And it is a valid reason to feel resentful now when charged as having no back up case, having spent enough spare time providing such evidence previously (at least to the best of one's own ability) that could have been devoted to other life (that does and should go on) outside net weather. I am loathed, frankly, to go to the same effort all over again, just to get to the end of a pointless internet tennis match. Winning doesn't come into it - just being able to give an opinion without feeling obliged to run around in circles just to satisy others is all that is asked. As you say it isn't a competition. Or shouldn't be. Clearly making this a general discussion thread and having a separate technical thread doesn't make the slightest bit of difference at all. There are still caveats to what one can and can't discuss it seems and still demands for evidence that if supplied threefold would never be deemed enough anyway.

It is indeed like having to resit an exam on here again that you have passed already - even if it was just with a marginal C- grade.

[rob48]

Quite. Couldn't agree more.

There are continual charges of not giving evidence to back up thoughts. The trouble is I already believe I have - as far back as last year or thereabouts I can't remember exactly when. I pasted some links and some pertinent texts about the possible problems with the assumed positive feedback and IPCC considerations as well as stratosphere and cloud feedbacks etc etc. Most of if was at the time ignored, ironically, having been requested at the time after being accused of the same thing as now!

It gets very tiresome having to go back, dig out, and/or repeat huge long posts of that sort just to satisfy someome who maybe wasn't privvy to see them the first time around. Or re-satisfy others all over again. And who would quite probably dismiss them anyway even if you did. That is not being funny, just being realistic. And it is a valid reason to feel resentful now when charged as having no back up case, having spent enough spare time providing such evidence previously (at least to the best of one's own ability) that could have been devoted to other life (that does and should go on) outside net weather. I am loathed, frankly, to go to the same effort all over again, just to get to the end of a pointless internet tennis match. Winning doesn't come into it - just being able to give an opinion without feeling obliged to run around in circles just to satisy others is all that is asked. As you say it isn't a competition. Or shouldn't be. Clearly making this a general discussion thread and having a separate technical thread doesn't make the slightest bit of difference at all. There are still caveats to what one can and can't discuss it seems and still demands for evidence that if supplied threefold would never be deemed enough anyway.

It is indeed like having to resit an exam on here again that you have passed already - even if it was just with a marginal C- grade.

Good for you Tamara.

The AGW brigade's sense of entitlement is such that they automatically assume they alone are correct, occupy the moral high-ground, and throw out challenges to others as though they are some sort of deity.

Don't waste your time, they're hardly on solid ground themselves.

[Solar Cycles]

Yes, but opinion gets us nowhere, and runs the risk of seriously misinforming other people. Even some sort of evidence to back up some of the spurious claims of 'no warming' or, 'CO2 not responsible' would be nice, after all, people should have some solid basis for having their opinion. If your opinion is based on whether you feel cold when you jump into the sea, then please forgive me for not taking your opinions as seriously as others! Hopefully your opinion is based on more than just that.

For example, looking at:

http://www.cefas.co....ity-trends.aspx

http://www.cefas.co....4-weymouth.aspx

There's not much evidence for significant cooling post-1998, and the Weymouth data (the nearest one with a long enough timeseries) contradicts your 'feel'.

I think Edinburgh feels colder over the last few years, but I'm not taking into account overnight minima, or instrumental observations that are far superior to "it feels colder" or "it feels warmer". And of course Edinburgh's temperatures over a few years have a fractional effect on global climate. It may have been remarkably warm in Yellowknife, Borneo and Timbuktu, but I can't see them outside my kitchen window, but I understand that to get a picture of global climate, I need to know what's been going on in other regions of the world far beyond MBY.

sss

Arrogance comes before a fall sss! You asked me to show you proof that we haven't warmed in the last 13 years, I was going to post the evidence, but then thought, "what is the point"! You my dear friend are so entrenched in your way of thinking, that any evidence presented, will be cast aside like a pros knickers. Any how time to take some annual leave from the climate threads, there's a stench around here that I don't like!

[Dartmoor_Matt]

Also, the second link, showed, to my eyes that the mean from 1971-2000 was lower than the mean for the whole dataset on the majority of months, so what was it you were getting at?

On my second attempt to view it it was blocked as a potentially dangerous site by my anti virus...

On top of that, Weymouth is an awful long way from Polzeath. So I'm not sure it contradicts anything. It affirms one or two things though.

[Iceberg]

Blimey we are not still on the old chesnut that it globally hasn't warmed, over the last 2,10,20,30,40 etc years are we. ?

I know we go round in circles on this thead, but come off it.

[Gowon]

Blimey we are not still on the old chesnut that it globally hasn't warmed, over the last 2,10,20,30,40 etc years are we. ?

I know we go round in circles on this thead, but come off it.

What!? Putting words in our mouths again? We know it has warmed but not for the reasons you think.

[noggin]

http://www.guardian.co.uk/environment/2010/feb/18/yvo-de-boer-climate-change

Ship. Sinking. Rat. Deserting.

??????????

[Thundery wintry showers]

What!? Putting words in our mouths again? We know it has warmed but not for the reasons you think.

Solar Cycles referred to an absence of warming over the last 13 years (the conveniently cherry-picked start year of 1998 with the remarkable El Nino), so Iceberg isn't putting words into anyone else's mouths. The fact that we're going through a spell of months that are, globally, comparable to 1998 despite a weaker El Nino doesn't inspire confidence for global cooling or even a stall of the underlying warming trend- though it does suggest that as yet the warming trend is not accelerating, as it will need to in future to progress beyond the lower bound of the IPCC's projection range.

The Copenhagen summit is proving a failure- which I half-expected from the start- but it says nothing about AGW or the science behind it.

[sunny starry skies]

Arrogance comes before a fall sss! You asked me to show you proof that we haven't warmed in the last 13 years, I was going to post the evidence, but then thought, "what is the point"! You my dear friend are so entrenched in your way of thinking, that any evidence presented, will be cast aside like a pros knickers. Any how time to take some annual leave from the climate threads, there's a stench around here that I don't like!

I'm still waiting for the evidence, SC and NSSC. For those suggesting arrogance, I've presented quite a lot of evidence to support my position, and I am entirely willing to view opposing evidence. So far, all I've got is a "I can't be bothered to find it now" approach - why don't you try me (and others) with your evidence? Or are you afraid that it won't stand up to the same scrutiny you expect of all evidence supporting the AGW theory?

Are you going to suggest that the trends clearly identifiable in GISS, RSS, UAH and HADCRUT3 are all bunkum? Or any evidence to contradict my opinion the only way you can find a cooling trend is by cherry-picking a massive outlier as your start point:

http://tamino.wordpress.com/2009/12/07/riddle-me-this/

Nobody has responded to this one: take the trend from 1980 to 2000, project it forward for the next ten years, and what do you find? Any change? What's the slope on all graphs between 2000 and 2009 (albeit not significant as it's such a short dataset) - is it positive or negative?

Please, SC, all I'm looking for is good evidence to support your point of view, but the only way you can suggest the world has cooled is if you use the great 1998 El Nino as your cherry-picked start point. And even that is not the warmest year in all global records.

sss

[Thundery wintry showers]

Looking at those graphs, we've ended at a similar end point to where the IPCC expected, but this has been arrived at by more-than-expected warming up to 1998, and less-than-expected (indeed statistically insignificant) warming since then.

This is consistent with the suspicion that natural/cyclical factors were mostly in positive phases in the 1990s and contributed to the rapid warming of the 1990s, but then switched to mostly neutral or negative phases in the 2000s, offsetting the underlying warming trend.

[sunny starry skies]

http://tamino.wordpress.com/2009/12/15/how-long/

Speaking of statistical significance in the trend, here's a nice article by the same author discussing how you establish error estimates on a trendline for a temperature timeseries. Because the variability about the mean is autocorrelated (one data point is frequently related to the next), you need to treat the variability as red noise, rather than white noise, and so the standard error margins (1-sigma etc) are larger.

What this all means is that in order to determine a trend in the data, you need a bare minimum of 15 years, statistically, to identify your trend in the temperature datasets. Anything less (let alone cherry picking 1998, 2005 or whatever), and you are establishing the slope of the noise rather than the slope of the trend.

sss

[kold weather]

I'm currently running the Climate prediction model at the moment to see what happens, lets just say the very first day is a very wet day over the west of the UK with a classic +ve NAO, should be interesting to see how my model run ends up.

[VillagePlank]

http://tamino.wordpress.com/2009/12/15/how-long/

Speaking of statistical significance in the trend, here's a nice article by the same author discussing how you establish error estimates on a trendline for a temperature timeseries. Because the variability about the mean is autocorrelated (one data point is frequently related to the next), you need to treat the variability as red noise, rather than white noise, and so the standard error margins (1-sigma etc) are larger.

What this all means is that in order to determine a trend in the data, you need a bare minimum of 15 years, statistically, to identify your trend in the temperature datasets. Anything less (let alone cherry picking 1998, 2005 or whatever), and you are establishing the slope of the noise rather than the slope of the trend.

sss

Bump :lol:

You've got me confused, there, SSS.

If one is to use linear least squares (linear regression) then I thought that the following must be assumed:

1. There are no errors in the independent values
   
2. Errors in the dependent values are normally distributed with a mean of zero and a constant variance
   
3. Crucially, errors in the dependent value are uncorrelated
   

Now, 1 is clearly true (or, at least it is easily observable as true) since it is a timeline. But what of the other two? Specifically 3: autocorrelations is a technique for correlating the data set to itself, and, furthermore, as far as I can make out, if there are low changes (the codomain range is low) across a series, then autocorrelations will display a positive bias. Also, 2 is quite difficult, I think, to ascertain before a choice of using LLS is made with reference to geo-data.

Because of these problems is it not better to bin the whole lot, and change the whole perception of the argument such that trending with this data is, probably, a silly thing to do, and that, perhaps, comparison of time periods is possibly much better (I know the general public want trend lines!) such that we don't say:

'The temperature is trending upwards'

But we do say,

'The mean temperature of this five years is higher than the mean temperature of some five years hence'

I did say I was confused! I note that the link does mention that the error estimates tend to be too low (so 'corrections' tend to be too high) but I also note that autocorrelation techniques do work in other domains such as automatic pitch correction in music production (you didn't really think that those singers could actually sing that much in tune, did you?)

I think I've gone full circle now, and I don't think even I know what sort of point I was trying to make, so I'll hide back in my hole, and read more!

Edited February 19, 2010 by VillagePlank

[sunny starry skies]

Bump

You've got me confused, there, SSS.

If one is to use linear least squares (linear regression) then I thought that the following must be assumed:

1. There are no errors in the independent values
2. Errors in the dependent values are normally distributed with a mean of zero and a constant variance
3. Crucially, errors in the dependent value are uncorrelated

Now, 1 is clearly true (or, at least it is easily observable as true) since it is a timeline. But what of the other two? Specifically 3: autocorrelations is a technique for correlating the data set to itself, and, furthermore, as far as I can make out, if there are low changes (the codomain range is low) across a series, then autocorrelations will display a positive bias. Also, 2 is quite difficult, I think, to ascertain before a choice of using LLS is made with reference to geo-data.

Because of these problems is it not better to bin the whole lot, and change the whole perception of the argument such that trending with this data is, probably, a silly thing to do, and that, perhaps, comparison of time periods is possibly much better (I know the general public want trend lines!) such that we don't say:

'The temperature is trending upwards'

But we do say,

'The mean temperature of this five years is higher than the mean temperature of some five years hence'

I did say I was confused! I note that the link does mention that the error estimates tend to be too low (so 'corrections' tend to be too high) but I also note that autocorrelation techniques do work in other domains such as automatic pitch correction in music production (you didn't really think that those singers could actually sing that much in tune, did you?)

I think I've gone full circle now, and I don't think even I know what sort of point I was trying to make, so I'll hide back in my hole, and read more!

I'm afraid my statistical knowledge is far below Tamino's, and most likely yours in this matter, although what Tamino said appeared to make some sense. From Wiki:

"Brown [red] noise can be produced by integrating white noise. That is, whereas (digital) white noise can be produced by randomly choosing each sample independently, Brown noise can be produced by adding a random offset to each sample to obtain the next one." This is very like the climate signal.

I see your point (3) about linear least-squares regression, though have just read (again on wiki, but I need to be quick here I'm afraid!) that the assumption of "Statistical independence of the observations is not needed, although it can be exploited if it is known to hold."

So far as I'm aware (and I may be wrong, hey we all are learning aren't we!), you don't need the data to be entirely independent, but you need to allow for non-independence when calculating confidence limits associated with a trend. Your last point about using means I think is pretty good, though I think the aim of using trends in this case is to highlight the underlying signal of warming, showing that we're on a trajectory at present, and for now have not left that trajectory. In future, we may leave that trajectory in either direction of course.

sss

Edited February 19, 2010 by sunny starry skies

[Iceberg]

I think the main point is that I hope nobody is using linear trend analysis for trend predictions, but that this is being used to trend hopefully known values from the past.

Does the above make sense...?

This is why some areas of skeptic reasoning concerning IPCC projections of temperature and short term deviations are incorrect, the projections are made not using linear analysis projections, but applying a trend analysis to known outcomes derived from climate models (however good or bad those models are, which is a different debate). If you then applied a 95 percentile to those predictions from the models and draw an additional upper and lower line, current global temperatures are indeed within the bounds of IPCC predictions.

BTW have you tried this VP with your LI software i.e making very small adjustments to the variables to produce a range of outputs (I know you said that you have a very good correlation anyway and that cooling wouldn't come for another 25 years), but.....

[stewfox]

Not sure if this has been posted elsewhere but I found the attached interesting

http://news.bbc.co.uk/1/hi/sci/tech/8511670.stm

[VillagePlank]

BTW have you tried this VP with your LI software i.e making very small adjustments to the variables to produce a range of outputs

See, here, for my intended testing method

I think the main point is that I hope nobody is using linear trend analysis for trend predictions, but that this is being used to trend hopefully known values from the past

Indeed, such extrapolations would be sinful! But, with respect that's not point - the point is that historic analysis is using linear least squares. This may be OK (I don't know) but I'd like to know a good reason why clear assumptions can be laid to one side for certain classes of data.

[Iceberg]

See, here, for my intended testing method

Indeed, such extrapolations would be sinful! But, with respect that's not point - the point is that historic analysis is using linear least squares. This may be OK (I don't know) but I'd like to know a good reason why clear assumptions can be laid to one side for certain classes of data.

I think your OK, as what is being plotted is not an absolute value(temparature), but an anomaly.

[VillagePlank]

I think your OK, as what is being plotted is not an absolute value(temparature), but an anomaly.

That's surely for peer review and not for you or I?

[Iceberg]

Sorry it was a royal your, not specifically for you. If that makes sense in electronic form.

[Anti-Mild]

And the prize for the most flagrant shoehorning of the potential effects of the carbon footprint on the climate goes to:

http://news.bbc.co.uk/1/hi/world/asia-pacific/8538033.stm

Seriously, I'm at a loss for words!

No wait, I remember something I wanted to poke extra fun at! I let out a chuckle upon reading Dr Pershing's description of whales as the "forests of the sea"!!! :lol:

I have no idea what Dr Pershing looks like, but if he doesn't have long hair and wear beads and a kaftan I'll be devastated!!

Edited February 26, 2010 by Anti-Mild

[jethro]

A huge iceberg has broken off from Antarctica, there's speculation that it could disrupt the world's ocean currents and weather patterns leading to colder winters in the North Atlantic.

http://news.bbc.co.uk/1/hi/sci/tech/8538060.stm

[Devonian]

A huge iceberg has broken off from Antarctica, there's speculation that it could disrupt the world's ocean currents and weather patterns leading to colder winters in the North Atlantic.

http://news.bbc.co.uk/1/hi/sci/tech/8538060.stm

There was an article about another polyana in the Antarctic in last months 'Weather'. The flux of energy involved when off shore winds means water is persistently frozen and re opened despite temperatures way below freezing are very, very great.

Given that, and fwiw, I'm to be convinced a iceberg, even a big one, can have much effect on overall (rather than local) Antarctic deep water formation and then climate. But, I will read further articles on this with interest.

[Solar Cycles]

There was an article about another polyana in the Antarctic in last months 'Weather'. The flux of energy involved when off shore winds means water is persistently frozen and re opened despite temperatures way below freezing are very, very great.

Given that, and fwiw, I'm to be convinced a iceberg, even a big one, can have much effect on overall (rather than local) Antarctic deep water formation and then climate. But, I will read further articles on this with interest.

Have to agree with you Dev, I can't see an iceberg, no matter how big, having much of a global impact.

[jethro]

I think the effect (if any) has nothing to do with energy flux.

Impacts may happen if the volume of fresh water stored in the iceberg is sufficient enough and in the right position to interrupt the deep currents by reducing the salinity. It's been speculated that the last ice age was caused by a vast volume of fresh water bursting it's dam and interrupting the ocean currents. No, I'm not saying an ice age is on the way, just demonstrating the proposed mechanism and dangers involved with this adrift iceberg.