Comments barry schwarz has made

Global warming stopped in 2003

The linear trend from 1998 to 2007 is 0.1C. For that particular period, the globe has been warming.

Run a linear trend from 1998 to July 2008 (latest data at the time of this post), and we still have a warming trend. This despite choosing the highest anomaly ever as a starting point and a very low anomaly for the end point.

But if you run a linear trend from 2001 - 2007 (or mid-2008), the trend is cool or flat (depending which record you use).

Does this mean warming stopped in 2001?

Look at the numbers.

Hadley temperature anomalies in degrees C:

1988 - 0.174
1989 - 0.109
1990 - 0.247
1991 - 0.203
1992 - 0.070
1993 - 0.104
1994 - 0.169
1995 - 0.270
1996 - 0.138
1997 - 0.347
1998 - 0.526
1999 - 0.302
2000 - 0.277
2001 - 0.406
2002 - 0.455
2003 - 0.465
2004 - 0.444
2005 - 0.475
2006 - 0.421
2007 - 0.399
2008 - 0.280 (average of current monthly data for the year)

Clearly, the decade following the very high 1998 is warmer than the previous. 1998 was an exceptional year, but the linear trend from that year to present is still warming.

If we run linear trend lines (with Hadley numbers) from after 1998 to present, we get downwards trends. But does that mean the 2001 start year for those trends mark the year global warming "stopped"? Not at all.

If we treat 1998 (which warm spike includes the latter months of the year before) as an extraordinary anomaly - caused by a very strong el Nino - we still have warming up to about 2004 - 2005.

I ran polynomial trend lines in Excel from 2002, 2001, 2000... 1988 to present to get a curve on the change in temperature. Every one of the curves 'peaked' around 2003 - 2004 or after. The shorter the time frame, the earlier the peak. Even eyeballing the numbers you can see that the warming continued, with a spike at 1998 (not the highest jump in the record), until 2003 - 2005.

So which is more correct? To say warming stopped in 2001 (or 1999 or 2000) because these are start years for a downward trend to present? Surely not. Look at the figures again from 1999.

1999 - 0.302
2000 - 0.277
2001 - 0.406
2002 - 0.455
2003 - 0.465
2004 - 0.444
2005 - 0.475
2006 - 0.421
2007 - 0.399
2008 - 0.280

The five years following 2001 were all hotter. How then can it be said global warming "stopped" in 2001? The peak temperature is at 2005, but I nominate 2003 because the polynomial trend curves from 1988+ to present all peak around 2003 - 2004.

The globe has been cooling for 5 years at the most, possibly for 3. 2001 can't be the year global cooling started because the globe kept getting warmer beyond that until 2005.On 'Global warming stopped in 1998'--Only if you flagrantly cherry pick posted 1 year, 3 months ago 170 Responses

references

And you should quote IPCC directly. They say:

Cite: For the next two decades a warming of about 0.2°C per decade is projected for a range of SRES
emission scenarios.

That's a 20 year window starting from 2006. Bit early to be talking about that projection mid-2008. Unless, of course, you believe temperature should increase monotonically with CO2 rise. I doubt you make that mistake yourself, but I asked above to make sure.On 'Global warming stopped in 1998'--Only if you flagrantly cherry pick posted 1 year, 4 months ago 170 Responses

numbers

Max, you must be congratulated for showing how to fiddle figures.

Let's see.

1906 - 1945 = 0.54C

1946 - 1976 = 0.07C (period showing flat or cooling trend - depending on which years are picked - this period is the one most commonly referenced)

1977 - 2005 = 0.48C

Now, there's something wrong here. These linear trends add up to 1.09C! And the ratio of warming for the two warming periods is 52% for the first period, and 47% for the second.

Can you explain why?

(Please run a linear regression for the periods and see if I got the figures right)

Now, let's express the math a little differently.

The 1st period of warming was 0.54C over 40 years

The 2nd period of warming was 0.48C over 29 years.

Extrapolating to decadal trends...

The decadal trend for the 1st warming period is 0.135C

The decadal trend for the 2nd warming period is 0.165C

Therefore, the rate of warming has increased.

Fun with numbers. Just with Hadley numbers, though.

You seem to be surprised that Hadley adjust their data. You should not be.

However, your concern does make one wonder why Hadley makes the mistake of giving data at ANY TIME that doesn't work in their best interests. I mean, if Hadley are so rotten, why don't they just fabricate their data from the outset? Why would they give ANY fuel to 'rational skeptics'? Why not just make every month for 2008 hotter than any previous month? Then they wouldn't have to bother with this 2008 'problem', eh?On 'Global warming stopped in 1998'--Only if you flagrantly cherry pick posted 1 year, 4 months ago 170 Responses

Max

"Global warming stopped in 1998" - not accepted.

The annual trend from 1998 - 2007 is upwards.

The trend to May 2008 is invalid, since you are starting with a highly anomalous year and ending with a highly anomalous few months.

If I accept this cherry-picked trend, and agreeing that it shows cooling, the statement "global warming stopped between 1998 and May 2008" is not accepted.

"The period 1998 - May 2008 shows cooling with linear regression analysis" - accepted.

"The period 2001 - May 2008 shows cooling with linear regression analysis" - accepted.

This is not an artefact of annual analysis, this is an artefact of monthly analysis. The noise outweighs the trend. If you check your coefficient of determination (R-squared), you will find, as I did, that it has a low value (between 0 and 1), representing a poor fit to data with the simple least squares methodology. The variance over the short period is too large to admit a useful calculation using linear regression. If you are a whizz at statistical analysis, you should know this. You need to deploy a different methodology to get a better result.

I learned how to plot a trend on Excel. That should have been obvious from some of my later posts over the last few days. I learned quite a lot about statistical analysis. It seems you are unwilling to face facts on the paucity of your methodology.

As you say, the cooling trend (from a linear regression) doesn't tell us much over the short period, certainly not that "global warming stopped in 1998.

I notice that the global warming trend "stopped" many times during the last hundred years. Clearly, it didn't for the last 100 (or 106) years.

As you know, mainstream science does not predict monotonous warming, and various AR4 projections posited future cooling or flat trends of 8, 10 and even 20 years duration, owing to the influence of natural forces (weather), still finding a general trend upwards over the long term. Thus, the recent trend, even when calculated with the low-significance least squares method, is unsurprising. The signature of radiative forcing remains both in the instrumental record, and in projection to 2100.

I will, like you, watch with interest the coming year's temperatures with great interest. If the recent cooling/flat trend lasts another 12 years, I will be very happy to hope that mainstream AGW theory is seriously flawed. What I expect, however, is that the trend will resume to warming, and that the high temperature of 1998 will be exceeded in all temperature records within the next 8 years.

I agree that there's no use continuing. Thanks again for the conversation.On 'Global warming stopped in 1998'--Only if you flagrantly cherry pick posted 1 year, 4 months ago 170 Responses

Max

Please explain why you begin your trend analysis at 2001 and not 2000 or earlier.

As you have said, statistics can be used for various and nefarious purposes. I agree. I want to know why you have chosen the particular time period you have, not just because it's up to date, but why you haven't begun with, say, 2000 or 1995, or 1990.

I agree that the period you have picked shows a cooling trend. Please explain the validity and significance of the period you have chosen (and are fixed on). It does not match any other period discussed here. What's your point regarding it?On 'Global warming stopped in 1998'--Only if you flagrantly cherry pick posted 1 year, 4 months ago 170 Responses

Max

Looks like the Hadley assessment ignores the last 5 months of data (January through May 2008).

My analysis does not.

Can this be the cause for the discrepancy?

Certainly. You have selected a different time series; 2001 - May 2008.

Please explain why you have chosen this period.On 'Global warming stopped in 1998'--Only if you flagrantly cherry pick posted 1 year, 4 months ago 170 Responses

FYI Max

I received an email reply from Hadley on how they calculated the 1998 - 2007 temperature trend.

Dear [Xxxxx],

The trend was calculated from annually-averaged data using a simple least squares fit of a first-order polynomial.

I imagine they got a better coefficient of determination from this method. They used annual rather than monthly data.

"Simple" for some! Way over my head.

A linear equation (y = b0 + b1x) is called a first order polynomial

http://www.cohort.com/costatpolynomialregression.htmlOn 'Global warming stopped in 1998'--Only if you flagrantly cherry pick posted 1 year, 4 months ago 170 Responses

Max

You said;

Watts does an excellent job of substantiating his statements with the actual records he cites.

Both UAH and RSS records do, indeed, show a cooling trend since 2001

I... have... confirmed to my satisfaction that Watts' statements are absolutely correct on this cooling trend since 2001.

Clearly, you did not read the post that Black Wallaby cited, and which I was responding to. Here it is again.

http://wattsupwiththat.wordpress.com/2008/07/02/what-a-di ...

Watts isn't talking about the 2001 - 2008 trend.

Cheers,

barry.On 'Global warming stopped in 1998'--Only if you flagrantly cherry pick posted 1 year, 4 months ago 170 Responses

Max

First of all, it's not "my" statement. I am looking at this objectively. I am not a spokesperson for the IPCC.

You said;

So it is not correct to say that "the world's surface temperature has continued to increase since the TAR".  It would be more correct to say: "Clearly, the world's surface temperature has NOT continued to increase since the TAR", (since the records all show that this is what happened).

As HadCRU give a trend for 1901 - 2000 at 0.6C, and for 1901 to 2005 at 0.71C, the AR4 comment is consistent with Hadley's analyses (and GISS and NCDC, I believe).

You said;

The trend from 2001 (when TAR was published) until now has been one of slight cooling

The AR4 was published in 2007, displaying conclusions up to 2005. Pitting that against the trend "to now" is invalid if you wish to test the propriety of the AR4 conclusions. If you want to compare properly, run a trend line from 2001 to 2005, and if you want to compare with AR4 specifically, run a trend for GISS, NCDC and HadCRU and compare, as was done in AR4.

BTW, as your trend analysis for the last decade is at odds with Hadley's (+0.1C), presumably using the same data, to what do you attribute the discrepancy?On 'Global warming stopped in 1998'--Only if you flagrantly cherry pick posted 1 year, 4 months ago 170 Responses

After more reading, Max

You say

IPCC uses a more sound analytical method (linear regression over many years)

Reading the AR4, I see no single trend analysis methodology. TAR used optimal averaging for their final figure and both TAR and AR4 employed maximum likelihood linear trends as well as linear regression regarding the trend sets they work with (neither IPCC report confines its calculations to Hadley alone).

I've dabbled with Excel on my own (regression technique) and am close to learning what to do. I'll wait til my experimenting has been verified by someone with qualifications.

As I've been reading on the subject, I've learned a few things.

There are a number of methods for establishing trends in statistical analysis, and each are used depending on their utility regarding the population being assessed. Least squares does not always provide the most valid methodology.

An important factor to consider is the coefficient of determination (which lies between 0 and 1 - the soundest fit being close to 1). It would appear that TAR found better efficiency using optimal averaging. (Out of curiosity, what was the coefficient for your trend analysis?)

Both TAR and AR4 mention various methodologies employed by different analyses, depending on the population being assessed.

These comments and figures might interest you.

A HadCRUT3 linear trend over the 1906 to 2005 period yields a warming of 0.74°C ± 0.18°C, but this rate almost doubles for the last
50 years (0.64°C ± 0.13°C for 1956 to 2005...

Clearly, the changes are not linear and can also be characterised as level prior to about 1915, a warming to about 1945, levelling out or even a slight decrease until the 1970s, and a fairly linear upward trend since then (Figure 3.6 and FAQ 3.1). Considered this way, the overall warming from the average of the first 50-year period (1850-1899) through 2001 to 2005 is 0.76°C ±
0.19°C. Clearly, the world's surface temperature has continued to increase since the TAR and the trend when computed in the same way as in the TAR remains 0.6°C over the 20th century. In view of Section 3.2.2.2 and the dominance of the globe
by ocean, the influence of urbanisation on these estimates is estimated to be very small. The last 12 complete years (1995- 2006) now contain 11 of the 12 warmest years since 1850, the earliest year for which comparable records are available....

The HadCRUT3 surface warming trend over 1979 to 2005 was more than 0.16°C per decade, that is, a total warming of 0.44°C ± 0.12°C (the error bars overlap those of NCDC and GISS). During 2001 to 2005, the global average temperature anomaly has been 0.44°C above the 1961-1990 average. The value for 2006 is close to the 2001 to 2005 average.

http://ipcc-wg1.ucar.edu/wg1/Report/AR4WG1_Print_Ch03.pdf ...

AR4 gives a figure of "slightly more than 0.65C" for the period 1901 - 2005. This comes from combining NCDC, GISS and HadCRU, the American trend sets finding a lower increase than Hadley over the period. The linear trend given by Hadley for 1901 - 2005 is 0.71C (table 3.3 - page 248).

Where I'm at so far...

The word "therefore", then, is quantitatively misleading in that it does not account for an increased trend of 0.09C derived from losing the period 1901 - 1905 from the start period in the new (not cherry-picked) "updated 100-year trend", but not qualitatively misleading in that the trend has increased, whether the analysis begins from 1901 or 1905. If we base the calcs on HadCRU alone, the trend is bumped up 0.03C from the loss of the 1901 - 1905 period.

The 1901 - 2000 trend remains 0.6C using the optimal averaging methodology, which is presumably more robust than least squares (higher coefficient of determination) for the population then being assessed. There is nothing inherently wrong with shifting the period when the apples being compared are 100-year trends. AR4 gives figures in the body of the report for the 1901 - 2005 period. To the best of my understanding I agree that the language is SPM is misleading, and find that this is a quantitative, not qualitative digression. The basic conclusion is sound.

I hope this makes sense and look forward to your comments.On 'Global warming stopped in 1998'--Only if you flagrantly cherry pick posted 1 year, 4 months ago 170 Responses

Following you perfectly

You're not following me.

It may be that Hadley have not significantly revised their 20th century figure. I don't know. Neither do you. We're both working under the assumption that your Excel results are sound. I have no way yet of verifying that. What I've found instead is that the TAR did not screw up Hadley's figure, as you vouched. This does not encourage confidence in you. Nevertheless, I'm not remotely interested in using that error as a way to dismiss your other points. We move on.

You've not seen fit to explain how I might verify your result (except to urge me to do something I don't know how to do). That's ok. It's not your job to teach me how to run a trend analysis in Excel. I've an acquaintance who is an economist who has moved here from Britain and is trying to get Excel program for his shonky computer.

Having clarified, I feel I need to reiterate to save misunderstanding, this time quoting you.

According to you (based on a report by Jones in 2001, which you cite) Hadley later revised their record for 1901-2000 from 0.6 to 0.65C, which is exactly what their current record shows.

Jones et al's figure is 0.6C. The 0.65 figure comes entirely from you. I've not seen it corroborated anywhere else. Until I am able to run a trend analysis myself, I've only your word to go on. I have never said Hadley revised their figure. I have said that Hadley have adjusted the data, and on the assumption that your figure is correct, speculated that this may account for the discrepancy. Nothing more assertive than that has come from me.

I've pondered it openly but offered nothing conclusive save to verify that Hadley's figure for 20th century was indeed 0.6, It may still be.

Is that clear?

Trust but verify (speculate along the way). That's how I'm operating right now. That's how I always operate.

In other words, Hadley corrected the record "ex post facto" to show 0.65C warming rather than 0.6C warming over the 20th century period from 1901 to 2000.

I have stated very clearly that I can find no corroboration from Hadley for your 0.65 figure. I have asked you if you have been able to verify it at source (not by running your own trend off the data, but a direct cite for the figure from the UK Met Office - I can't find one).

I've twice made clear that I find the language in AR4 ("therefore") questionable and potentially misleading. I do not know why you repeat the assertion when I've provisionally accepted it.

Are you following so far?

I'll repeat this. I agree the language is dubious. Until I can run my own trend analyses, I am going along with your posit.

Surely you can understand that it is entirely reasonable for me to trust a reputable institution (Hadley) at least as equally as an anonymous internet poster (you). Until I can do the math, this is the fairest position to take. If I simply swallowed what you said hook line and sinker, without the skill to verify, I wouldn't be a very good skeptic, would I?

You are making the firm assertions here, not I. I am considering your points pending my own education on the matter. I can't think of a better way to go about it. You appear to want me to believe, and are impatient because I don't have complete, unquestioning faith in your abilities. Surely you're not asking me to take what you say on faith? Isn't that what 'alarmists' do regarding the IPCC?On 'Global warming stopped in 1998'--Only if you flagrantly cherry pick posted 1 year, 4 months ago 170 Responses

Kup

Hope you don't mind me replying.

There are plenty of good reasons outside of AGW to transfer to other energy sources. Eventual peak oil is one (and rising prices make it economically feasible), and another is to wean ourselves off the necessary relationships with tyrannical regimes. If there's anything to the war-for-oil hpyothesis, there's another good reason.

A free-market rationalist might argue that there's a bonanza in a new energy market.

We made the shift away from CFCs without too much pain. We reduced sulfur and other pollutants without going belly up, and new industries were spawned. Weaning ourselves off fossil fuel is a much more difficult prospect, but there are potential market benefits for doing so for smart entrepreneurs.On 'Global warming stopped in 1998'--Only if you flagrantly cherry pick posted 1 year, 4 months ago 170 Responses

Black Wallaby

My word Barry, if you don't know a bit more about Phil Jones et al, (like Briffa and Osborn) at UEA Climate Research, then you must be very new to this scene!

18 months of fairly intense study and discussion via the semi-popular literature, IPCC, all the sites we've mentioned and many more, and a bunch of actual studies.

I think I get your drift. The IPCC community and endorsing scientists are incompetent/corrupt/snared by groupthink, and McIntyre, Pielke's Snr and Jnr, Motl, Soon, Baliunas, Allende, ICECAP, Spencer, Christy, Carter, Monckton and the rest (on Ifnohe's list, for example) are 'realists'. I'm an alarmist. You're a realist. There are two sides in this 'controversy'. There is no consensus.

Am I close?

Oh, there I am getting involved with the political angle. This should help us get down to brass tacks. We'll be writing each other off in no time!

seem to be avoiding quoting him of late. (just as with Al Gore)

Anyone citing Al Gore on the science needn't be considered seriously. Discussing the merits or otherwise of his film is a different matter. For the record, I think it's based on sound science, but with a propagandistic twist. I think the UK judge got it right.

Presumably if The Great Global Warming Swindle is brought up in a court case on its presence in schools, the ruling to have it accompanied by additional learning materials would apply also, under the same rubric.

Concerning your long discussion on validating the date of your reference from Hadley

I don't do soundbytes. The truth isn't arrived at via that mechanism on such a complex issue. When someone makes blanket statements with an air of certainty, it is a strong indication their POV is influenced by something other than science.

In any case, my "your long discussion on validating the date of your reference from Hadley" amounted to one sentence. I have no idea what you're talking about.

It does not surprise me that your latest Press Office release dated 2008 includes the older 2007 figure, because the 2008 figure would be rather contradictory to their spin.

What 2008 'figure'? When presenting an annual temperature plot on a question that deals with annual temperatures, it is best to wait until the last year is over before including it.

I'm still interested in why you think the recent profile is a better match with 1940 than the other years I posited. And what your general point might be on that.On 'Global warming stopped in 1998'--Only if you flagrantly cherry pick posted 1 year, 4 months ago 170 Responses

Barry; sucked-in by Phil Jones et al?

I might ask the same question of you. If you think Jones is a problem, why do you give credence to his data sets? He's partly responsible for the numbers you use.

Let's make it clear who Phil Jones is.

Philip D. Jones (1952-) is a climatologist at the University of East Anglia, notable for maintaining of the time series of the instrumental temperature record [1]; this work figured prominently in the IPCC TAR SPM [2]. He is director of the Climatic Research Unit and a Professor in the School of Environmental Sciences at the University of East Anglia in Norwich.

http://en.wikipedia.org/wiki/Phil_Jones

The Met Office (Hadley) produce their temperature series in conjunction with the Climatic Research Unit. CRU. The CRU in HadCRU. Phil Jones, CRU director, is responsible for your data sets. Jones et al 2001 was the latest study available when TAR was written, authored by the director of the Hadley (HadCRU) temperature record (this is more for Max's benefit). It isn't some study cherry-picked by the IPCC - this was from the guy who does the Hadley records (in co-operation with others).

If I'm sucked in, we all are.

As I said, I give equal credence to the three main temperature reconstructions. They have excellent,  if not perfect, agreement. If you're going to call Phil Jones into question, what temperature record do you have left? What will you run your analyses from?

Barry's Hadley reference does not appear to have a date, but it is notable that the year 2007 bar is coloured green, which means in Hadley lingo that it was an estimate back then

There is a link on that page to a pdf version. At the bottom of the pdf version, there is this;

Produced by the Met Office
FitzRoy Road, Exeter
Fax: 0870 900 5050
© Crown copyright 2008 08/0112

http://www.metoffice.gov.uk/corporate/pressoffice/myths/b ...

When was the MSU reanalysis done? (Above is the fax number if you want to check with Hadley on when the 1998 - 2007 trend was done. Max may wish to fax them a query on their analysis)

Barry, should also be well aware that there is a typical "correction" after any up or down spike, and that the sharp reversal of T in 1999 & 2000 after the high of 1998 is thus not unusual.

I do not know what you are trying to imply by 'correction'. I don't see it as at all unusual that after a temperature spike the temp returns to some kind of base, roughly. In the case of an ENSO event, the temperature 'normalises' when the ENSO event finishes. This seems pretty straightforward to me. I've not understood why you're so fascinated with it. I imagine such spikes occur in any plot with noise. My bank balance does this every pay day, and, unfortunately, there is often a sharp 'correction' when I withdraw what I need for the week (old habits die hard). But I've been saving... :-)

Barry should also be aware that the later Hadley presentation which is part of the comparisons in the link I post above, shows a break-over in the smoothed curve with a suggestive broken blue line, ending at 2006. So why did they stop short at 2006?

I don't know. 2007 data too recent to include in a decent analysis? What was their methodology?

And weren't you telling me that ten years is too short a time to establish a trend? Now you want it to, what, the last month?

Well maybe because the raw data suggests a sharp down-trend from thereon that they don't like to see

2007 is 0.025C less than 2006. Is this a 'sharp' downturn?

somewhat similar to what happened around 1940?

I put this to you before: there are other down curves in the temperature record that show better fit to the recent 'downturn'. I see a better fit with 1869, 1960 and  1970. 1940 is much sharper than those and the current profile. See for yourself. Why are you fixated on that period?

I'd be interested in a response on this - why you favour the 1940 fit over the other periods?

It occurs to me that global warming stopped in 1869, 1880, 1900, 1940, 1960, 1969 and 1978. Except it didn't.On 'Global warming stopped in 1998'--Only if you flagrantly cherry pick posted 1 year, 5 months ago 170 Responses

Max

Thank you for confirming that you made the graphs, and that your methodology was "I simply downloaded the raw data as published by Hadley (source cited) into Excel, plotted the graph and 'put in' a linear trend line, and calculated the equation for the decadal trend".

I agree that the language in the SPM bit you cited is questionable.

In 2001, Hadley's trend for the 20th Century stood at 0.6C. I've corroborated that from the source study, and also from a Hadley newsletter, both from 2001. The TAR stated the trend as Hadley had put it then.

The methodology used to arrive at that figure is probably a bit more complicated than whatever you're doing in Excel. Check the study, and read here on Linear regression.

http://en.wikipedia.org/wiki/Linear_regression

You reckon that's the Excel methodology?

Hadley has revised the dataset many times since TAR. I do not know if your Excel result uses the proper methodology, but if it does, the discrepancy may be an artefact of reanalysis and changes to the data.

I have good cause to question your methodology. You say there has been a flat trend in the last 10 years. Hadley says the trend has been 0.1C. Check this update from Hadley.

http://www.metoffice.gov.uk/corporate/pressoffice/myths/2 ...

How do you explain the discrepancy between your figure and Hadley's? It seems you are happy to give their data credibility, but not their analysis. Can you explain the problem with their trend methodology?

I will come back to see what you've posted, and will continue the conversation if there is something substantive therein.

Cheers,

barry.On 'Global warming stopped in 1998'--Only if you flagrantly cherry pick posted 1 year, 5 months ago 170 Responses

I think we're done Max

The insults are unwelcome.

I don't know how to make any clearer that I already understood the data source (Hadley), from the first time I looked at the 'farm' graphs. The information is right there on the page. I've said twice now that I knew this.

But that is not what I asked you for.

I asked you who made those graphs, and for some information on the methodology used.

Who made the graphs? Who did the trend analysis? What was the method used?

You have failed to provide this information after repeated requests. I assume you don't know or are unwilling to say.

If these graphs are applied using "the same" methodology as Hadley, why does Hadley give different results? Is Hadley in error?

You claim that Hadley has said: ""A simple mathematical calculation of the temperature change over the latest decade (1998-2007) alone shows a continued warming of 0.1 °C per decade."

This is not my claim. This is a direct quote, word for word, from Hadley, which I hyperlinked for you. Here is the full URL.

http://www.metoffice.gov.uk/corporate/pressoffice/myths/2 ...

Either Hadley is wrong or you are.

I don't know how to do a trend analysis myself. I suspect you won't be interested in helping me understand how to do it.

The point was that IPCC SPM 2007 insinuated with the use of the word "therefore" that the increase from 0.6C to 0.74C which resulted from the five-year shift was a result of the increased warming trend at the end of the period (2001-2005) rather that from eliminating a cooling trend from 1901-1905 from the record (which was the actual cause for the increase).  Did you understand that, or is this too complex for you?

I did indeed understand it. I will quote my last post.

I see what you mean, but owing to the complexity of trend analysis, it's difficult to take on face value (looking at curves in the plot) that the 0.14C increase between TAR and AR4 is an artefact of removing the 1901 - 1906 period rather than an artefact of the last few years of the updated record. With respect to that, I do agree that the language in the IPCC ("therefore") is confusing and may not be based on something as clear cut as what it purports (recent 12 years highest in record). It may even be quite misleading. I am investigating.

I've essentially restated your thesis here while putting forward my own thoughts on it.

You have not dealt with your claim that IPCC called the 1905 - 2006 period the "20th Century". I see now that you will not.

Thank you for the conversation.On 'Global warming stopped in 1998'--Only if you flagrantly cherry pick posted 1 year, 5 months ago 170 Responses

Still thinking about it, Max

I think you raise some interesting points, Max. I will attempt to reply to them within my limited capacities.

Firstly, clearing up some possible misunderstandings and making myself clearer;

In the TAR (2001) IPCC made reference to 20th century warming, which it defined as the period 1901-2000 (i.e. the "20th century, as we know it).

Ok. TAR evaluates the 'last 100 years', which is the 20th Century.

In your previous post, you said,

IPCC stated that "second half 20th century (redefined to exclude 5 years of cooling from 1951-1955)...

[...]

Taking a look at the IPCC's cherry-picked "20th Century", starting with 1906

I got the impression that you were saying that IPCC had arbitrarily (or conveniently) picked the years 1906 - 2005 and called it "the 20th Century". When I checked the source (IPCC 2007), the period 1905 - 2006 was called "the last 100 years", not the 20th Century, as I believed you were alleging. Your reply is to refer me to the IPCC 2001 report, which does refer to the "20th Century", that being the last 100 years prior to 2001.

Unless I've misunderstood you, you are conflating the two different reports' time series and erroneously concluding a bit of fudging. The period 1906 - 2005 is not described as "the 20th Century" anywhere in the IPCC (AFAIK).

But I do get what you're saying about the change in periods in your following commentary (see further below).

Another minor issue, but one which pops up a few times in your commentary.

It [TAR] referred to the Hadley surface temperature anomaly record in saying that this showed a linear warming of 0.6C over the century.  Now, if one actually checks the linear increase over the period 1901-2000 according to the Hadley record, one sees that this was actually 0.65C, rather than 0.6C as reported by IPCC.  A minor error.

(And one that suggests the IPCC either rounded down the trend, or calculated it as less than that given by Hadley (which you've said elsewhere in your post that you prefer to GISS and NCDC). Responding to your suspicion of the IPCC, doesn't this tidbit indicate a more conservative reckoning by the IPCC than you'd expect, or is your thesis here that the IPCC is incompetent, rather than deliberately misleading?)

I could not find a reference at the UK Met office for the 1901 - 2000 series used in TAR having a warming trend of 0.65C, per your advice to check it. I went to the source study Hadley used at the time for the period (Jones et al 2001 - cited in TAR), where I found this;

The 20th century experienced the strongest warming trend of the millennium (about 0.6C per century).

I could find no reference to 0.65 for that period (1901 - 2000) from Hadley or anywhere else. Could you please provide a clearly referenced cite re Hadley 1901 - 2000 temperature trend of 0.65C to corroborate your claim that IPCC got it wrong?

Now to the latest IPCC SPM 2007 report...

The increase of 0.09°C over the 100-year period (not 0.14°C  as erroneously reported by IPCC) came from removing a cooling trend from 1901-1906, and NOT from adding a period of rapid warming from 2001-2006, as suggested by IPCC by inserting the word "THEREFORE".  In fact, the linear trend over the period 2001-2006 was essentially "flat".

I see what you mean, but owing to the complexity of trend analysis, it's difficult to take on face value (looking at curves in the plot) that the 0.14C increase between TAR and AR4 is an artefact of removing the 1901 - 1906 period rather than an artefact of the last few years of the updated record. With respect to that, I do agree that the language in the IPCC ("therefore") is confusing and may not be based on something as clear cut as what it purports (recent 12 years highest in record). It may even be quite misleading. I am investigating.

If you would like to see the graphs of the Hadley record confirming all of what I have said above, please let me know.

I would appreciate that, as well as attributing authorship, methodology and/or, if applicable, any studies or links to explanations. I do know the source material is Hadley data for your other graphs, but that's not the question I'm asking. 'Who' and 'how' is what I want to know. It may be that whoever drew up those trends is not qualified/mistaken or has used inappropriate methodology. How am I to check that with such scant information?

Now to your point: "I note also that the IPCC averages from the 3 main instrumental temperature records, not just Hadley."

This is not correct. If you take the time to download the Hadley record as is it reported, you will see that it is exactly the same as the IPCC record over the entire 20th century.

You may be right. Ch. 3 AR4 refers to a variety of temperature records, including GISS and NCDC, but doesn't categorically state which series it adheres to, or if the SPM figures above come from a combination of them. Hadley-like and actual Hadley graphs are used in the chapter, but NCDC and GISS overlays appear within some of them and they are all discussed. I'm checking on it.

OTOH, have you now jumped back to TAR (I assume you meant AR4)? It's important we both know what is being referenced as we move through each point.

With Hansen's shift from "objective scientist" to "AGW political activist"

This is politics. Let's stick to data and analysis.

plus the recent disclosures of errors in the USA GISS record, I have a bit more faith in Hadley than GISS today.

"Recent disclosures of errors"? Errors are regularly found in all temp records, and Hadley continually updates its temperature record (example, example, example), as the others do, when errors are discerned or better information comes along. If there was no adjusting going on or errors being admitted, then I would become suspicious.

The satellite records (UAH, RSS) also check more closely with Hadley trends than they do with GISS.

Counterintuitively, the satellite record (which measure temps by inference, not directly) may be more susceptible to error than the surface record. The 2 main records (RSS, UAH) have some significant discrepancies between them, and both constructions rely on the same data. That either or both of them more closely match the Met Office record is not inevitably a validation of HadCRU. For the record, I take it that all three main temp surface records are robust, and that they all have close agreement on trends, even if individual years are slightly different.

We have seen from the above example that "what it actually says" must be expanded to include "what it actually implies by saying what it says".  By adding the word "THEREFORE" the clear implication is something totally different from the actual facts.

I will reserve judgment til I've investigated further, but I have to say that the use of the word "facts" is a red flag to me, particularly when dealing with temperature reconstructions and complicated statistical analyses - if there's one thing I've learned, the trend analyses applied to the topic at hand vary and the results are all approximate only. There are no "facts" on this, just degrees of validity.

When IPCC tells us (see quote above) that the linear trend of the past 50 years (1956-2005) is steeper than that of the full 100 years (1906-2005) it is doing exactly this [cherry-picking].

One could do the same in reverse (a reverse "cherry pick") by saying the linear trend from 1910 to 1944 was steeper than the linear trend from 1910 to 2005, thereby insinuating that warming has slowed down over the century.  This would be as misleading as the IPCC statement hinting at the opposite conclusion....

....I will just attach one curve from IPCC AR4 Chapter 3 FAQ 3.1 Figure 1, which is an extreme example of this "smoke and mirrors" approach.
http://farm4.static.flickr.com/3221/2534926749_f2be35e86f ...

The question posed by IPCC is 'have we warmed any faster over the last 150 years'?

The methodology employed seems valid - run a trend line for 150, 100, 50 and 25 years and compare the results.

This is not "cherry picking". This is a valid approach to answer a question.

The question that your cherry-picking example would be answering is, 'has there been a period of warming in the last 150 years greater in trend than the last 50 or 25?' A different question. Selecting the 1910 - 1940 period is cherry-picking (we both agree on this much) within the context of the question being asked.

Can you see the implication in this curve that things are getting worse as we move along in the 20th century?  It is clearly there (if only implicitly stated).  But if we start at the front of the overall curve and make linear trend lines of ever increasing increments always starting from the front, we will see exactly the opposite trend.

Then, if you are able to gin up graphs, let's see the result of plotting 25, 50, 100 and 150 linear trend lines all starting at 1850 going forward. I am very doubtful we will see a mirror of those trends in the IPCC graph. The sine curve goes up on one side - it doesn't have a mirror profile on the other.

I propose that you have the sine curve comment backwards. The temperature plot is not designed to be a sine curve, but plotting shorter periods reveals that it is over the long-term, or has been so far. You are putting the cart before the horse.

I know you WANT to believe the IPCC claims

Actually, I'd rather the IPCC claims were all wrong and that we are not headed for global warming. But my views are not based on what I prefer to believe.

so it is reasonable to assume that you will automatically think that it is more reasonable to trust IPCC trends over someone's rationally critical appraisal of these IPCC   claims, even if the critical appraisal is backed up by the same source data used by IPCC.

I will constrain myself to Hadley references, as you appear to give that source credence. You may know that Hadley endorses the findings of the IPCC.

And you're right - I do think it more reasonable to credit the validity of the IPCC (or Hadley, GISS, NCDC) over a mysterious "someone", who remains anonymous after several requests for authorship. You are asking me to take things on faith. I am still open to giving those plots credence, but you've given me no reason for such confidence. Effectively all you've said is "you can believe in them, trust me".

Who made those graphs?

I'll get back to you on your erroneous belief (against the recorded and published temperature records) that "the past 10 years have shown warming".  It's just not so, Barry and even the IPCC Chairman, Pachauri has publicly stated that he is aware of this "plateau".

Hadley says;

"A simple mathematical calculation of the temperature change over the latest decade (1998-2007) alone shows a continued warming of 0.1 °C per decade."

So does NASA, so does NCDC, so does the Australian Bureau of Meteorology and reputable scource that has looked at this (with slight variations in value). Graphs with no author or methodology attributed may have been, for all I know, drawn up by someone with an ax to grind and are erroneous. Surely you can understand the necessity in giving this information (not just pointing out that the source data is Hadley, which I got the first time I looked at them) in order for anyone to have any confidence in them?

Still my mind is open. My view is based on what I've learned so far and provisional on better knowledge, well-supported and explained. Let us both try to avoid making assumptions about the other, eh?

barry.On 'Global warming stopped in 1998'--Only if you flagrantly cherry pick posted 1 year, 5 months ago 170 Responses

IPCC facts and figures

I often find that values attributed to the IPCC are at odds with what it actually says.

You wrote, Max:

IPCC stated that "second half 20th century (redefined to exclude 5 years of cooling from 1951-1955)

[...]

Taking a look at the IPCC's cherry-picked "20th Century", starting with 1906

The very first thing said in the IPCC chapter on Observations of Climate Change is this;

"Global mean surface temperatures have risen by 0.74°C ± 0.18°C when estimated by a linear trend over the last 100 years (1906-2005). The rate of warming over the last 50 years is almost double that over the last 100 years (0.13°C ± 0.03°C vs. 0.07°C ± 0.02°C per decade). Global mean temperatures averaged over land and ocean surfaces, from three different estimates, each of which has been independently adjusted for various homogeneity issues, are consistent within uncertainty estimates over the period 1901 to 2005 and show similar rates of increase in recent decades. The trend is not linear, and the warming from the first 50 years of instrumental record (1850-1899) to the last 5 years (2001-2005) is 0.76°C ± .19°C."

(PAGE 237, first comment - pdf)

IPCC states "last 100 years" here, not "20th Century", and measures 1906 - 2005.

If you know where in the IPCC it says what you state, could you please provide a link and page number?

I note also that the IPCC averages from the 3 main instrumental temperature records, not just Hadley. I get skeptical when Hadley is the sole source for skeptical commentary. It has a slightly lower trend than the other two, and it is the only one of the three showing 1998 as the warmest year on record. Is there a sound reason why this profile is preferred? It is fit for the topic of this thread, but I don't know why it is the sole source for the particular topic you've brought up. Has the same analysis you iterated been performed using the other two profiles?

Black Wallaby and yourself post graphics from http://farm3.static.flickr.com. When I go to that site, this is all that appears on the screen.

photocache313.flickr.sp1

So it's an image bucket. There are no references to whoever made the plots, and neither you or BW have attributed authorship. You both present them as if they're authoritative. Can you understand that it's hard to give them credence?

You've stated upthread that you use IPCC trend methodology. If I assume this is the case with the graphs you have referenced, how am I to reconcile the first 25 or 50 year higher trend with this;

"The rate of warming over the last 50 years is almost double that over the last 100 years (0.13°C ± 0.03°C vs. 0.07°C ± 0.02°C per decade)."

Either you, your source, or the IPCC is in error. I don't have the skills to do any analysis to see for myself whether this

Obviously, this analysis is just as phoney as the one made by IPCC

is true. But I think it's more reasonable to trust the IPCC trends over anonymous graphs, and not reasonable to credit claims that are demonstrably at odds with the source referenced (IPCC - but I'll wait for direction from you re IPCC statements that match your assertions).

If you have more information on these two items I will follow the trail to better understand.On 'Global warming stopped in 1998'--Only if you flagrantly cherry pick posted 1 year, 5 months ago 170 Responses

Max

Pity this site doesn't have an edit function. I feel that the comment I made in parentheses at the bottom of my post to you was far too hasty. Apologies.On 'Global warming stopped in 1998'--Only if you flagrantly cherry pick posted 1 year, 5 months ago 170 Responses

Black Wallaby and Max

Black Wallaby, Max has reiterated the meme that is the title of this thread, constraining the trend calculations to ten years - less actually, as you can see by the graphs with the seven year trend that have a trend line showing cooling. This is what Coby (and I) are responding to. It's fortuitous that Max has come along to substantiate that this trope is alive and kicking.

I checked out the website you linked to - it's not my cup of tea. Where people are calling each other 'deniers' and 'alarmists', the talk is mostly political in nature, not scientific. There is a bit of science attempted, but it is rooted in agenda and not very sophisticated. Too much focus on 'sides'. I much prefer McIntyre's main site where the numbers are crunched and very little name-calling goes on.

Max, I refer you to both Black Wallaby's and my comments saying that a ten year period is too short to register a climate trend. Black Wallaby has left off the conversation, but you'll see he agrees upthread.

But even within the constraints you are working with (trend analysis from 7/10 years of data), I think the trend lines may be wrong. If you'll go upthread to my first post, I wonder what you make of my simple methodology for establishing a trend. It shows warming whether you use GISS, Hadley or NCDC.

Actually, I'll save you the trouble and spell it out again.

For the last ten years alone, a simple trend line can be established by averaging the first and then the last five years. If the average of the last five years is greater than the first, global temperatures have risen over the decade, no?

Please show in detail how you arrived at the cooling trend line. What's your math?

(I will treat your comments outside this paradigm of 'sides'. I will endeavour to address the substance of your posts, not dismiss you as being driven by some sort of agenda or ideology. If you can return the respect, we can have a conversation)On 'Global warming stopped in 1998'--Only if you flagrantly cherry pick posted 1 year, 5 months ago 170 Responses

reply to Black Wallaby

Meant to say, my brother lives in Melbourne. I've visited and worked there a lot and love the city. If only it wasn't quite so cold in winter, I'd probably live there (bring on the global warming!).

It is actually quite difficult to find a website that supports a moderate view of the AGW debate...

Are you saying that there is no site representing the mainstream view that I can reference that you would approve of? That doesn't leave me with much.

I don't have a problem with realclimate. I think the tone is sometimes too dismissive. I prefer Pielke's site for reasonable discourse, but I'm in no position to judge authoritatively the quality of the science of any of them. I am, however interested in learning and testing what I've learned - why I go to pro and anti sites. The science is what matters. Realclimate presents science mostly in line with the IPCC. Climate Audit and Climate Science do not so much.

I for one believe that there IS some warming caused by anthro effects, but that it is rather trivial.

This is a common skeptical viewpoint, one I like to see explained so that I can test it against mainstream theory.

For instance, there is clearly a regional variation in the Arctic which is warming, but then at the same time there is an opposite regional variation in the Antarctic.

Yes. But that is the beginning, not the end of discussion. I would point out in a discussion on it that the warming signature in the Arctic region far exceeds the cooling signature in the Antarctic, and that the Arctic cooling (almost flat-lining, really) is not unanticipated. That's not the end of the story either.

Regarding your comment on the IPCC being thrown at you, how much of it have you read?

I've read about 60% in the last year as I've researched and discussed various topics.

One bit from something you quoted caught my eye.

On GHGs raising temps/not;

there is obviously no clear empirical evidence for either position

I think that's flat out wrong. First, definitions;

Empirical

1.

a. Relying on or derived from observation or experiment: empirical results that supported the hypothesis.
b. Verifiable or provable by means of observation or experiment: empirical laws.

2. Guided by practical experience and not theory, especially in medicine.

There most certainly is empirical evidence for AGW. The premise of the whole theory is based on an experiment that can be done in any well-equipped school laboratory. Shine a beam of (infrared) radiation through a volume of atmosphere. Record the amount that is emitted through the other side. Add CO2 to the volume. Perform the same test. Unless the parameters are unrealistic, less radiance passes out the other end of the volume when more CO2 is added. Therefore, increasing CO2 in a volume of atmosphere traps more radiance or heat.

This is the empirical basis of AGW theory. It is an empirical observation, not a theory, a test performed many times.

How this translates to the actual atmosphere is another matter, in which case the author might say that there is some a priori evidence that renders AGW theory questionable, but it is definitely not correct to say there is no empirical evidence.

There is an immediate political response to this fairly neutral opening

Seeing as the contributor concludes his remarks with this...

So what do we do now? Act irrationally, or study more and wait? I say, study more and wait.

...giving an unqualified opinion, how on earth do you credit the post as 'neutral'?

Just read the first page of that link and it seems that jae originally said that "world-renowned" physicists disagree that "GHGs will increase temperatures". Jae reduced 'world-renowned' to reputable and added the word "significantly" (reduce temperatures).

It's hard to credit that whoever wrote this is a "scientist". Judging by their misuse of the word 'empirical' I'd hazard a guess that they are not a critical thinker of any kind.

If you don't want me to cite realclimate, fine. Please reciprocate by citing only reputable science sources, not this amateurish stuff.

Just so you know, I can spot the difference between politics and science easily. Your post above is mostly political. Do you realize that?

The main blog at ClimateAudit is a different thing, and Steve McIntyre asks people to shuffle-off to the BB if things are going too much off topic or too elementary whatever

I have very much enjoyed the concentrated effort to assess the US temperature plot under the rubric of contaminated/non-contaminated sites. I recommended that particular thread elsewhere on this site.

Barry, concerning your or Tamino's (?) idea of excising spikes

Please, I'm not interested in politics. I'd appreciate it if you didn't try to align me with someone else. I post my own thoughts based on what I read. If I had picked up this theme from Tamino, I would have said so and probably cited him. Can we drop the name-dropping?

OTOH, if Tamino has posted something along the lines I've been dabbling with, I'd be grateful for a link.

I'll  be blunt;  there is no scientific or statistical basis for doing it.

Sure. The statistically valid way of removing noise (weather) is to smooth or remove well-identified signatures (noise). I do not know if the ENSO signature is sufficiently well-identified to feasibly remove from the trend. I posted the graph on it upthread (not from Tamino's site AFAIK) because it was somewhat in line with what I was thinking - an example of sorts. I'm keeping an open mind on it for now.

If you know any reputable scientific source that lays claim that 1998 El Nino contributed 0.2 degrees, I would like to see the theory and maths behind it.

To be frank, I am surprised that you are unfamiliar with this well-known factoid. I've never seen it generate a skeptical response, but kudos to you for not accepting it on faith.

Here's a page with seasonal stats:

http://www.cpc.ncep.noaa.gov/products/analysis_monitoring ...

Check 1997-98.

Another with yearly departures from norm (includes links to studies):

http://www.cdc.noaa.gov/people/klaus.wolter/MEI/#ElNino

Detailed SST anomalies (per day)

http://www.bom.gov.au/bmrc/ocean/results/pastanal.htm

An alternative methodology resulting in similar 0.2C anomaly for 1997/98 ENSO event (Multivariant ENSO - MEI)

http://www.cdc.noaa.gov/people/klaus.wolter/MEI/table.htm ...

And a few other studies and analysis with links provided. you will find some statistics to work the math on some the of these references.

http://www.cru.uea.ac.uk/cru/press/2006-12-cet-records/fu ...
http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid ...
http://www.ncdc.noaa.gov/oa/climate/research/1998/enso/10 ...
http://cat.inist.fr/?aModele=afficheN&cpsidt=1941004
http://www.cgd.ucar.edu/cas/catalog/climind/Nino_3_3.4_in ...
http://www.cgd.ucar.edu/cas/catalog/climind/TNI_N34/index ...

Other reading.

http://physicsworld.com/cws/article/print/1459

Let's see how you go with these two thingies before I respond to your other stuff.

This is the second time you've sent me on a quest with the promise of responding to my earlier points on my return. If you don't mind, I'll wait until you've responded to my old points before looking at the links you posted. Once bitten twice shy and all that.

The points outstanding that I'm interested in are;

1. If we are constrained to measure a trend over ten years, is the methodology I employed satisfactory?

(I agree that 10 years is not ordinarily sufficient - I'm looking for an answer within the bounds set by the thread topic, and one commonly set my the skeptical milieu- "The earth has been cooling since 1998". This particular question gets us back on topic)

2. What do you mean when you say temps have 'plateaued' recently? Do you mean the long-term trend or something shorter?

3. Do you agree that the two year correction following a spike does not happen "almost every time"? (Or will you post a value range for spikes so that I can test your thesis?)

I will attend to your other points and references  when you've responded to these.

Hope it's warmer where you are. It's getting mighty chilly in Bondi.On 'Global warming stopped in 1998'--Only if you flagrantly cherry pick posted 1 year, 5 months ago 170 Responses

In case you're wondering

I chose the 0.1C value and greater because it was easiest to work with.On 'Global warming stopped in 1998'--Only if you flagrantly cherry pick posted 1 year, 5 months ago 170 Responses

two cents

Not a simple issue to my mind. True enough that the US (don't forget Australia) has a huge per capita emissions output as well as having a much larger historical contribution that is present in the atmosphere today, but China is the biggest gross emitter and growing (faster than any other country, [perhaps?). On the one hand there's the nation to nation perspective, but on the other there is the total global gross output - the atmosphere isn't going to make any concessions regarding the distinctions we've framed here.

What I'd like to know is what would be a fair proposal for Kyoto 2012 regarding China's emissions. Should we tally up the historical contributions and per capita emissions and targets and work strictly according to that, or should we hedge on a strict accounting and restrict the allowance for China to some degree in order to close the gap on targets? And how would the math work for that?

A balance has to be struck that is reasonable, and, perhaps unfortunately, politically viable. I think Kyoto 2012 should include some restrictions for China, even if only to assuage the people who decry China's immunity. The US will be harder pressed top cry foul and will perhaps more likely sign up.

AFAIK, the US has taken some serious steps to reduce emissions outside the UN framework. But then, China's energy policies and practises appear to be progressive also. I don't know what the formula should be. There's a lot of extenuating circumstances to factor in - like, as mentioned upthread, production of goods and who benefits from who.

Has the issue of future targets for China been addressed anywhere I can check out on the web? Any proposed formulas for establishing emissions targets for Kyoto 2012?On China's carbon emissions highest in the world last year, study says posted 1 year, 5 months ago 7 Responses

reply to Black Wallaby

As a matter of principle, I do not open links to SkepticalScience or RealClimate (or Tamino/RayPierreH references), because in the first instance, I hold little credence for what they claim and manipulate.  Furthermore, the last thing I want to do is give them any benefit by increasing their website hit-count.

Ok.

Out of curiosity, where do you go to get a scientific understanding of the mainstream view? Are there sites you do visit that are not climateaudit, Pielke's site, or any of the skeptical sites? For balance, I mean. I go to Pielke Snrs site and climateaudit to get a skeptical view that seems reasonable. Where do you go for the mainstream view?

How much do you excise and under what rules?

Oh I'm no expert, but I'll dabble at an answer.

I guess you could pick a certain value, one that is a significant swing. Let's say we excise all annual temperature swings over 0.16C (+ or -). I'm guessing it's an odd way to do 'smoothing', but it would reduce the amount of noise and reveal the underlying trend, I suppose. This is the result of excising ENSO from the series. ENSO events produce periodic swings in the data. I've read that the 1998 el Nino added 0.2C to that year's temperature. As long as you take out the upswings and the downswings to the same value, I can't see how you'd be corrupting the underlying trend, just revealing it more.

Regardless, you cannot "legally" do that unless you also excise the typical two year reversals which follow such spikes.

If you think that it is needful to excise this two-year downswing, then fine by me. Perhaps you could try that. If you're a dab hand with graphics, maybe you could post the results.

How do you do that? Add a bit on? How much for year 1 and year 2?

I don't know. My method is simpler (and naive). I'd just excise them completely if the downswings are an artefact of the upswing. Not my thesis, but I say go for it if you think it's important.

I'd be happy for you to explain more about your thesis.

Please study the Hadley graph more carefully, and you should be able to see that the repetition of this internal "correction" over the 150 years is so remarkable and frequent, that 1999 and 2000 is just more of the same, and 1998 is just a typical spike.

To my eye, there are periods as you describe, as well as many other changes that don't fit that profile; spikes followed by a much smaller swing in the same direction, spikes followed by a retrogade change and then the next year reversing again, spikes followed by more than two years of the opposite direction in temps. It seems somewhat random. The only constant in the general trend upwards.

But let me come along with you a bit more. If there are regular periods where a big spike is followed by two years opposite swing, what is the minimum value of the initial upswing that gets the trio into your set? If you give me a figure, I'll check out the Hadley graph again and see for myself. (eg, is it an upswing of +0.15C or something else that qualifies as a contender for the profile you're talking about?)

As far as I am aware, Hadley and others issue the raw data and apply smoothing techniques without excising anything.

Yep. I certainly wasn't suggesting otherwise. I was following my own line of thinking - well, not my own, really. I thought of it before I came across examples on the web.

Why do you think that Hadley and even GISS, and others are not smart enough to do the excising that you discuss?

I don't understand the question. I never meant to imply that Hadley have/could/should knock any of the big swings out or saying anything about their ability to do so.... reading back on my posts, I see that I haven't implied that at all.

BTW, I posted @ 7:33 PM blog-time, and you responded @t 8:05 PM.   Even if you saw my post instantaneously, which would be "lucky", that would give you a max of 32 minutes to digest the information I gave, go and find your reference, compose a reply, and post it.

I didn't look for a reference. I looked at the graph for about five minutes, looked at some of the down trends, compared the recent downturn with others that appear after spikes and concluded that the 1940 spike was a sharper profile than the recent swing, and that there were better fits elsewhere in that record, and that's what I told you. You hadn't posted about your two-year response to spikes at that time.

Please demonstrate a little more care-time if you would like me to continue explaining things to you.

Read your first post to me and see if there was something I should have got from looking at the graph that was made clear in that first post. We may be talking at cross purposes. I refer you to my initial post. I've been assuming we were talking about trends and how they are worked out. Are you more interested in talking about this two-year drop hypothesis? That's cool. I'll try to follow, but I'd like to know if you're going to reply to the questions you invited me to ask.

It took me about half an hour to do my quickie post

I read your post twice - about four minutes, checked the graph, five minutes, and then composed my reply - ten minutes, including preview checks. About 20 minutes in all, I guess.

While I'm all for taking time and care with reading and replies, I see no profit in comparing our response times. Hopefully you'll see from my replies that I try to think carefully about the matters at hand.

From your previous post:

Look carefully and you will see that almost every time that there is an up or down spike, it is followed by a two-year correction, or reversal.

I don't see that happening "almost every time". I see it happening occasionally. But I'd need to get a better fix on how big the spike is you're working with. Let me know and I'll check it out and report back.

Meanwhile, if I haven't disappointed you yet, I'd be grateful for any thoughts on the questions I raised.On 'Global warming stopped in 1998'--Only if you flagrantly cherry pick posted 1 year, 5 months ago 170 Responses

logarithmic

I'd like to know what models assume with regard to the relationship between radiation transmission through CO2 and its concentration in the atmosphere.  Is it a linear decrease with ppm, exponential?

Logarithmic.

http://www.realclimate.org/index.php/archives/2007/06/a-s ...On 'CO2 doesn't lead, it lags'--Turns out CO2 rise is both a cause and an effect of warming posted 1 year, 5 months ago 43 Responses

plateau?

it is statistically significant to say that the spike of 1998 is not important

Elsewhere I put it that 1998 is an anomaly and it would be statistically valid to excise it (and other spikes downward or upward similar in magnitude) from the profile in order to get the underlying trend. For the same purpose, in some reconstructions the ENSO signal has been removed - el Nino and la Nina - to reduce the amount of noise from 'weather'.

http://www.skepticalscience.com/global-warming-stopped-in ...

(3rd graph)

How do you arrive at the conclusion that we've plateaued? Perhaps you are not speaking of trends but of recent movement in interannual variability. I hope to learn something more of statistical analysis. Eyeballing graphs is ok for starters, but no good for getting down to brass tacks.On 'Global warming stopped in 1998'--Only if you flagrantly cherry pick posted 1 year, 5 months ago 170 Responses

Hi Black Wallaby

Hi Bob, thanks for the welcome and reply.

No argument that the period is too short. I was working from the meme this thread is about, replying, as it were, to a common skeptical POV in the semi-popular literature. Quoting from the top post; "Global temperatures have been trending down since 1998. Global warming is over." Coby has already dealt with that well in the top post. I was interested in finding a way to work out a trend over the ten years in line with the skeptical constraints. I'm not aware of a more sophisticated (skeptical) approach dealing with the trend over the longer term.

Is there a particular reason why the 1940 down-curve was chosen in comparison? Why not the 1875-1880 curve, or the downturn just before 1900, or the one around 1960, which appear, at least to the naked eye from that Hadley chart, to better match the curve of the current trend?

If I get the gist of that chart correctly, it is being suggested that the 1940 downturn is something that might be occurring now, but I see no reason for that. The comparison seems arbitrary, perhaps even a little opportunistic if whoever put that graphic together is trying to suggest that we may headed for a mid-century-like cooling period.

I'm afraid my skills at statistical analysis are about as sophisticated as my initial attempt demonstrates. My maths is poor, but I'm pretty good at grasping things conceptually. If you can educate me in line with my strengths, I'm very much up for it.

Questions:

If we do constrain the period to that as put by the skeptical meme (last ten years), is my attempt satisfactory?

If the last five years is on average hotter than the previous five, should not a smoothed analysis recommend we're still warming? I believe this applies for any number of years larger than five to present, whether decadal, dodecadal (I may have made up a word here) etc.

Can you explain (without arcane math symbols) how the downward trend is arrived at in the last few years of the Hadley trend?

I realize this is all probably statistics 101.00001, but if you've the patience, I'll commit to applying myself.

From your user name; you're an Aussie? I am, posting from wintery Bondi, Sydney.

Cheers, mate!On 'Global warming stopped in 1998'--Only if you flagrantly cherry pick posted 1 year, 5 months ago 170 Responses

hard to drill down to facts on line

[Transferred and amended from my post at illconsidered blogspot]

Hi all,

I've been debating this particular issue for about 6 months, and I won't festoon this post with links, because there are few I've found on the web that get to the kernel of the matter - radiative absoprtion - and certainly not broadly enough to satisfy an honest inquiry.

[This is the best single study I've found (PDF). I believe Coby has linked to it in the top post]

The way greenhouse gases retain heat is by absorbing, re-emitting, and reabsorbing and re-emitting infrared radiation welling up from the Earth's surface. If the GHGs weren't there, the infrared radiation would escape to space and we'd be living (or dead, rather) on a very cold planet.

CO2 absorbs upwelling infrared radiation (IR) mainly (but not only) in the 15 micron wavelength range. But water vapour (WV) absorbs in many bands, and significantly part of the 15 micron band. If you looked at an absorption graph, you'd see a 'spike' for CO2 in the 15 micron range, and a 'slope' cutting across the 15+ micron range, being the WV absorption. WV also absorbs at lower wavelengths, as does CO2.

15 micron absorption of IR by CO2 is practically saturated in the middle of the 15 micron range, but there is more absorption possible either side. Adding more CO2 'fattens' the spike - a bit like painting a wide black stripe across a window, adding more CO2 is like adding a fine line of paint either side of the wide stripe, which prevents more light (IR) passing through the window. WV's absorption profile laid over this would be like painting a diagonal stripe that cut across the original one. If the wavelength values are increasing to the right, WV would cover the whole right hand side of the window and slope down to the left across the CO2 stripe. If you can picture this, that leaves mainly the left hand slope of the CO2 spike (very steep) available for absorption. Here is where most of the action is.

In the actual atmosphere, WV fluctuates dramatically. In arid areas, over deserts and the poles, the WV absorption profile is less pronounced. The CO2 15 micron 'spike' stands out more, and therefore the absorption of IR by CO2 is more pronounced.

At this point in the debate, honest skeptics want hard data from spectral absorption studies. There are hundreds, stretching back to the 1950s, when the US airforce began measuring the upper atmosphere for military purposes (thereby discovering that the atmosphere has many layers, and is not a single slab, as was thought until then). Early studies were ground-based at high altitude, from weather balloons and high-flying aircraft. Satellites were later used for spectral analysis, measuring from the top of the atmosphere down, and perpendicular to the surface of the Earth at altitude. Time series studies exist showing a rise in CO2 levels, and changes in thermal activity. Here is some of the hard data honest skeptics are looking for - at this resolution.

There is a database (HITRAN) that gives spectral absorption data for various gases. As with WV and CO2, there is some overlap in which gases absorb what wavelength of radiation. In Earth's atmosphere, there is a 'window' between the 8 and 12 micron band where there is virtually no absorption going on, and through which upwellng infrared radiation passes out from the surface to space virtually unimpeded. Adding more CO2 impinges on this clear part of the spectrum, at the 12 micron end.

However, the study is so complex, with instruments usually only capable of studying a few bandwidths at once, and the layers of the atmosphere needing to be taken into account, that you'd need to read a mountain of studies to get a whole picture. At Spencer Weart's website on the history of climate science, he recommends that one would need to go to a library and read at least one fat book on atmospheric radiative transfer to get the kind of data and understanding needed to get a fix on the drill-down 'facts'. He also points out that much of the spectral analysis was done 40 years ago, and many of the studies have not been transferred to the internet. IOW, to get to the guts of this issue, at the resolution of radiative transfer and absorption profiles, including times series and under different humidities, is probably impossible via the web. Not to mention that most studies that are available cost money to view.

There is a lot of pseudo-science on the topic. When a paper is cited, check the author's credentials. There are quite a number floating around written by (for example) neuroscientists (TJ Nelson), computer engineers (Petschauer), and experts in geomagnetics-regarding-military-weapons (Slade somebody-or-other, via the late John Daly's website).

I guess I want to warn you that absolute resolution on this, if pursued honestly, rigorously and to the enth degree, will leave the querant unsatisfied if they restrict their travails to the web. And I thought I'd point this out to you, Coby, because it's well to acknowledge limitations when acting as a conduit for information.

My take - I think it is unreasonable to think that mainstream climate theory is ignoring the role of water vapour in the atmosphere. I searched for the phrase (with the European/Australian spelling of 'vapuor') in the pdf files of the 2001 and 2007 IPCC reports. I got hundreds of hits, but the IPCC reports do not detail the spectral properties of WV and CO2 - that is too fine a resolution for the purposes of the report (if that kind of information was included, the document would be a hundred times larger and quite impractical). For that information you not only have to go to the studies listed in the relevant chapters, but also the studies those studies reference, and the studies those studies reference and so on, and pretty much all of these are pay per view and many wouldn't be found on line.

You have to be extremely bloody dedicated or slightly mad to get to the absolute bottom of this. I haven't made up my mind which category I fall in.

Hope this post was useful, or at least, that I spare some sincere sojourner the months of googling I went through. On the way, though, I learned a lot of vicarious arcana and useful stuff to understand what goes into the study of this particular subject.

May good tidings and intellectual honesty make a home in your hearts and minds,

barry.On Water vapor is indeed a powerful greenhouse gas, but there is plenty of room for CO2 to play a role posted 1 year, 5 months ago 29 Responses

water vapour discussed at length 2001 2007 IPCC

[Transferred in part (and slightly amended) from my post on the subject at illconsidered blogspot]

Water vapour gets a lot of mention in the IPCC documents (Intergovernmental Panel on Climate Change).

If you want to corroborate this, you can open the pdf files for each of the IPCC chapters at the links below, type 'vapour' (note the European/Australian spelling) in the search box, and click the appropriate button to search through each document (different versions of pdf have different search buttons).

IPCC 2001:

http://www.grida.no/climate/ipcc_tar/wg1/index.htm

Don't bother with the HTML pages on the left of that web page. They're just summaries of the chapters. Open the pdf files on the right hand side of that page and have a good search.

IPCC 2007:

http://ipcc-wg1.ucar.edu/wg1/wg1-report.html

I chose 3 chapters for 2007, the three I thought most likely would discuss water vapour (2, 3 and 8, from memory).

I also ran a search on 'carbon dioxide' and 'CO2' and compared the results.

Again from memory, there was 273 hits for 'water vapour', and 295 for 'CO2' and 'carbon dioxide' from those three chapters, including the studies referenced at the bottom of the page.

I chose 3 chapters from 2001 and did the same thing. There were tons of hits for water vapour in that one, too. I then checked all the chapters but didn't bother to count, as it was obvious that water vapour was discussed again and again. Some chapters only mention it a little.

Also, both the 2001 and 2007 reports had sections specifically dedicated to water vapour.

This says nothing about the focus or quality of the science, of course, but it completely dispels the myth behind this simple statement:

"Climate scientists dodge the subject of water vapor"On 'Climate scientists dodge the subject of water vapor'--No, they really don't posted 1 year, 5 months ago 4 Responses

oops

"Uranus may be undergoing internal processes that cause it to warm?"

Should have been "causing it to cool".On 'Mars and Pluto are warming too'--No they aren't -- and what if they were? posted 1 year, 5 months ago 24 Responses

skeptics working on it

Anthony Watts at surfacestations.org has been looking at US surface stations. With a team of dedicated skeptics he's been collecting photos and data on weather stations that are likely biased by local conditions (biased upwards temps), on the premise that the US temp record fails to properly account for the UHI effect.

Watts and his team have collected a good number of sites they think are of good standard. At the time of this post [transferred from illconsidered blogspot - posted late February 2008], they're about a third of the way through the project.

They have plotted the US temperature profile using only the sites they think sound, and have compared with the profile as given by the US Climatology Network (http://cdiac.ornl.gov/epubs/ndp/ushcn/newushcn.html), which is used by GISSTEMP and HADCRU.

So far, the results provide a very good fit.

http://yaleclimatemediaforum.org/features/1007_surfacetem ...

That article links to climateaudit, a well-known skeptical site run by Stephen McIntyre, which has been working on the data from the 'good' weather stations as determined by Watts' team. In the thread doing the math on the preferred data stations, this is the post (below) where one of the skeptics, John V, compares the surfacestations.org data with USCHN.

http://www.climateaudit.org/?p=2124#comment-147569

Even the skeptics are getting the same result as the mainstream. If anyone is sincerely interested in this subject, the surfacestations/climateaudit project now spans three threads (the post above is from the second one, which is 300 posts long), and it will be interesting to see the results when they complete their project. [If the results have been updated, I would be interested to know]

As a side note, I consider the discussion going on in this climateaudit thread to be the very best example of the skeptical community putting their money where their mouthes are. It is a substantive, polite investigation, and while they have begun with their conclusion (not very scientific), they are genuine in their efforts. When so much skepticism (and advocacy, for that matter), is couched in ignorance and vitriol, this project champions the best of the skeptical community. Let it be a standard-setter for all sincere discussion on climate change.

And, the US temperature profile looks almost exactly the same when all the urban data is excluded, which is one thing that is done to test for UHI.On 'Warming is due to the Urban Heat Island effect' -- No, it isn't posted 1 year, 5 months ago 25 Responses

collating papers from 70s

Recent (early 2008) article on the 1970s cooling meme.

The '70s was an unusually cold decade. Newsweek, Time, The New York Times and National Geographic published articles at the time speculating on the causes of the unusual cold and about the possibility of a new ice age.

But Thomas Peterson of the National Climatic Data Center surveyed dozens of peer-reviewed scientific articles from 1965 to 1979 and found that only seven supported global cooling, while 44 predicted warming. Peterson says 20 others were neutral in their assessments of climate trends.

The study reports, "There was no scientific consensus in the 1970s that the Earth was headed into an imminent ice age.

"A review of the literature suggests that, to the contrary, greenhouse warming even then dominated scientists' thinking about the most important forces shaping Earth's climate on human time scales."

http://www.usatoday.com/weather/climate/globalwarming/200 ...

A has been eloquently expressed above, it doesn't follow that because science includes hypotheses that turn out to be invalid, science is therefore unreliable.On 'They predicted global cooling in the 70s'--But that didn't even remotely resemble today's consensus posted 1 year, 5 months ago 29 Responses

Consensus list

Wacki - there have been some suggestions for additions to your logicalscience page (you complied it? Well done!) on consensus. Get on back there and update, dude. There's another Australian science institute to add, Exxon-Mobil quote (I posted those two), and Shell and a couple of others.

Here's the place;

http://logicalscience.blogspot.com/2007/07/consensus-on-c ...

On yer bike, lazybones!On 'There is no consensus'--If this is not consensus, what would consensus look like? posted 1 year, 8 months ago 109 Responses