2011-01-03 10:10:49Paper on correlation between CO2 and temperature
John Cook

john@skepticalscience...
60.231.58.57

A few people have emailed me about this new skeptic paper:

Warming Power of CO2 and H2O: Correlations with Temperature Changes

Full Text (PDF, 1794KB)

I haven't read the full paper yet (although it's freely available) but the abstract seems to indicate there's no correlation between CO2 and temperature thus indicating the CO2 greenhouse effect is saturated and water vapor is the major driving greenhouse effect.

Lastly, is it just me or do skeptic peer-reviewed papers always seem to be freely available online while many papers giving evidence for AGW are hidden behind paywalls?

2011-01-03 10:49:46not a good paper
dana1981
Dana Nuccitelli
dana1981@yahoo...
71.140.20.164

It's a really difficult to follow paper.  Part of the problem is probably language, as the author is Brazilian.  There are quite a few typos that never got caught during the review process, and there's a lot of unnecessary historical/background rambling, in my opinion.  Like he babbled about the UHI effect, but then just plotted data from NOAA and GISS which have removed UHI, so what's the point in talking about it?  Kind of frustrating, because I don't have time to read the paper in detail right now, so I want the guy to get to the damn point.

It seems as though he's arguing based on short-term data, atmospheric CO2 is lagging behind temperature increases, and thus the CO2 increase is due to ocean outgassing.  Of course we know this is wrong, because of the many many lines of evidence that the atmospheric CO2 increase is anthropogenic.  So, assuming I'm understanding the argument correctly, the conclusion cannot possibly be correct.  There's a total throw-away statement in there about solar irradiance increasing, with no source or specification as to the referenced timeframe.  While looking the paper over, I had several "WTF are you talking about?" moments.

Overall this seems like just a really poor quality paper.  And it seems like he could have cut out a bunch of the unnecessary crap and made it about one-third as long.  But maybe I'm missing something...?

2011-01-03 11:56:56
Rob Painting
Rob
paintingskeri@vodafone.co...
118.93.252.210
Dana, had a quick read. Boy it sure is garbled. It seems littered with fallacies, rising co2 is from ocean outgassing, cooling 70's, water vapor is causing global warming etc. Global temperature doesn't immediately respond to a change in forcing?. Get outta here!, really?.  
2011-01-03 12:52:15Like all things in life, you get what you pay for...
Daniel Bailey
Daniel Bailey
yooper49855@hotmail...
68.188.192.170

That one's been making the rounds for some time now.  Dunno when it first got published (I recall glancing through it some time ago; all I remember is the author's name, go figure), but that publisher (SCIRP) is known for republishing papers previously published elsewhere without attribution.

WTFUWT has a current post up on it now:

http://wattsupwiththat.com/2011/01/01/new-peer-reviewed-paper-absence-of-correlation-between-temperature-changes-and-co2/ 

A reader commented on it over at RC and Ray Ladbury remarked on it in general here:

http://www.realclimate.org/index.php/archives/2011/01/blog-updates-and-suggestions/#comment-196477

"About this particular incident of dumber-than-owl-droppings idiocy, Rattus Norwegicus vectored me to this nice, little evisceration of the publisher:"

http://improbable.com/2009/12/22/strangest-academic-journals/

'The publisher, Scientific Research Publishing, has other journals, as well. Some of them (Applied Mathematics, Journal of Modern Physics, Biosciences*, Journal of Cancer Therapy**, etc.) also appear to follow the “publish things that were already published, but don’t mention that” principle.

On the organization’s web site, we found barely any identifying or location information. The contact page says “Name: Scientific Research Publishing, Inc. USA” and lists an email address —  but we have not found any such corporation in the USA, and email sent to that address has produced no reply.  The web site is registered to an organization in Wuhan, China.'

2011-01-03 14:42:04Peer review
John Cook

john@skepticalscience...
60.231.58.57
So its safe to say it's not a peer-review journal? I saved that paper to the database as a peer-reviewed paper but perhaps I should change it to online article.
2011-01-03 18:15:49
Ari Jokimäki

arijmaki@yahoo...
192.100.112.203

There are of course plenty of other studies on this issue:

http://agwobserver.wordpress.com/2010/02/18/papers-on-co2-temperature-correlation/

2011-01-04 04:00:23nice
dana1981
Dana Nuccitelli
dana1981@yahoo...
38.223.231.252

Nice list there, Ari.

According to the PDF, it was published in the 'International Journal of Geosciences".  But when I did a Google search for that name, all I could come up with was "Published/Hosted by Scientific Research Publishing."

"International Journal of Geosciences is a peer reviewed journal dedicated to the latest advancement of geosciences. The goal of this journal is to keep a record of the state-of-the-art research and to promote study, research and improvement within its various specialties."

A Google Scholar search for the paper comes up empty.

Similarly a search for the author (PC Soares) comes up with a few hits, but mostly in Portuguese, and not for this particular paper.

Seems pretty sketchy to me, John.  Considering how shady SCIRP seems to be, and the low quality of the paper, I would say it's not really peer-reviewed.

Oh an in the comments to my recent 'scientists predicting global cooling' post, somebody linked to a paper by Douglass which cherrypicked ocean heat content data terribly to try to show ocean cooling.  The paper was published by SCIRP in the International Journal of Geosciences.  I see a pattern emerging.  Perhaps SCIRP is combining republishing old papers and publishing new 'skeptic' papers.  Perhaps the new Energy & Environment?

2011-01-04 05:53:52
nealjking

nealjking@gmail...
84.151.52.141
Someday, when the rebuttals are more or less complete, it might be worthwhile to search out and attack such error-filled articles.
2011-01-04 06:15:38
Ari Jokimäki

arijmaki@yahoo...
91.154.99.160

SCIRP has been bombing my e-mail for a while now with call for papers for this and that journal. They seem to be desperate to get their hands on something they can publish. In a situation like that they are likely to publish lot of rubbish. Here's the International Journal of Geosciences:

http://www.scirp.org/journal/ijg/

Latest issue has papers by Soares and by Knox & Douglass. Editorial board doesn't have any familiar names to me.

2011-01-04 09:14:45
MarkR
Mark Richardson
m.t.richardson2@gmail...
134.225.187.80

I plan to read it properly tomorrow if I have time...

 

But it seems a pretty retarded way of going about it. Errors that spring to mind are because water vapour follows temperature quickly, whilst temperature follows CO2 slowly and temperature is not only affected by CO2.

 If it's actually a published paper then I'm pretty confident that we have the know-how to eviscerate it in a peer reviewed response.

2011-01-06 08:56:52Warming power of CO2...
Klaus Flemløse

KlausFlemloese@mail.tele...
62.243.102.250

 

I am looking forward to an analysis of this paper.
The starting point of the paper is a statistical analysis of the relation between temperature and CO2.
Due to the random variation is difficult it to verify or falsity anything. 
He makes several assertions about LIA, MWP, UHI – probably not correct.
Several graphs do not contain relevant information in respect of the objectives of the paper.
Question:
How can we verify the correlation between CO2 and temperature?  
This paper will be used on the skeptical blog again and again. Therefore a reaction is needed.
2011-01-06 21:55:49
MarkR
Mark Richardson
m.t.richardson2@gmail...
192.171.166.144

Errors/chery picking/etc:

1) CO2 claimed as not human caused (well, duh, 30bn tons pumped out, atmosphere going up by 15bn tons. Can ANYONE explain how that is possible if not human caused? Especially as oceanic pH is falling)

2)  temperatures only taken from tropical northern ocean, not global. Justification is use of CO2 measurements from there but CO2 is largely well mixed. I'm pretty sure that choices of different areas would change his data significantly.

"The volume of data and information is fantastic and one
may unwarily select partial data and show bias results." - indeed!

3) Ignores radiative forcing, only looks at CO2. Allows aerosols to mess with results.

4)  does a McLean, looks at temperature change versus CO2 change, then plots a graph. Fig. 8 can easily be explained by aerosols and noise.

 5) Ignores sources of other signals where to account for them would destroy his conclusion.

 

Might be worth emailing the guy. Always a chance that he's genuinely ignorant of a lot of the data. Or he could just be being intentionally dumb to get media attention or oil funding.

2011-01-07 08:32:04Rebuttal of 'no correlation'
John Cook

john@skepticalscience...
60.231.58.57
Mark and Klaus, would you 2 be interested in writing a joint blog post? It could be used as a rebuttal to the argument "there's no correlation between T and CO2".
2011-01-10 01:21:18
MarkR
Mark Richardson
m.t.richardson2@gmail...
134.225.187.80

My contribution would have to be relatively short. I'm falling behind with my real work at the moment!

 

But sure, I think we can get a few simple 'take home' points from this.

2011-01-11 03:04:22
MarkR
Mark Richardson
m.t.richardson2@gmail...
134.225.187.80

This is really, really hard.

 

The paper makes so many fundamentally stupid mistakes I really don't know where to start or what to concentrate on. It's given me ideas for multiple posts.

 

1) Correlation of 'cause' vs 'effect' is difficult when you have multiple causes. Uses CO2 only when you should consider total radiative forcing.

2) The McLean and deFrietas deception of comparing differentials. I might do something akin to my 1998 DIY statistics post from ages ago showing how you can make CO2 caused global warming disappear (well, be very hard to see) with this particular sleight of maths.

3) Confuses feedbacks and forcings (and gets difficult because of the CO2 feedback too. Warmer years see less dissolution in the oceans, obviously)

4) Makes stupid unphysical arguments contrary to real world measurements and physics based on naive correlation analysis.

2011-01-11 04:30:10I hear that
dana1981
Dana Nuccitelli
dana1981@yahoo...
38.223.231.252
I often get that feeling when a 'skeptic' writes a long post containing a bunch of stupid statements.  It's hard to even know where to start refuting them.
2011-01-11 22:14:28
Riccardo

riccardoreitano@tiscali...
192.84.150.209

Actually the paper is easily dismissed. While it claims no correlation between CO2 and T, it shows something else; nowhere in the paper the correlation between the two quantities is analized. I mean, something like this.


2011-01-12 03:10:04
MarkR
Mark Richardson
m.t.richardson2@gmail...
192.171.166.144

It's easy to dismiss, because if you understand it it's obviously stupid and it's absolutely astonishing if this got past proper peer review.

 

But I think we could do a good job of explaining where it goes wrong so that more people can see how stupid it is. I've got some planning notes written down but I'm quite busy and it really is hard to work out where to start - there's so much wrong that picking a coherent collection of points is difficult!

 

 

 

I wanted to make sure it wasn't me being the idiot though so I've emailed Soares. Will give him a bit of time to reply whilst  I try and put together a post or two.

2011-01-12 03:11:03
MarkR
Mark Richardson
m.t.richardson2@gmail...
192.171.166.144
Also, quantities are analysed, but mainly in terms of annual changes. Which is a clever way of hiding global warming that McLean, deFrietas and Carter used (along with hiding a 0.2 C drop from splicing data).
2011-01-13 03:57:31
MarkR
Mark Richardson
m.t.richardson2@gmail...
192.171.166.144

The way I've been running with it is to cover a couple of things first and then 'what if CO2 warming was real... would this method detect it?' (the answer is no, but other methods would. And these have been used and it has been detected).


 

I need an analogy to do with a measurement that is inappropriate for detecting a long term change but can find a short one. Maybe about how if you measure the change of height on Earth you conclude the Earth is flat rather than round? 

The basic mistake he's made is to compare short term differentials that are clouded by noise: he's looking for a signal of <0.02 C/yr with data that has a range of 0.5 C/yr, and some of the 'noise' is actually anti-correlated to the CO2 for some of the time as well!

2011-01-13 04:29:40casinos
dana1981
Dana Nuccitelli
dana1981@yahoo...
38.223.231.252
What about gambling?  You gamble in a casino for a day and at the end count up your chips.  You won $500, therefore in the long run, surely you'll make a fortune!  Not a good analogy for the measurement aspect though.  That's a tough one.  Your analogy with the flat Earth is pretty good.  Your speedometer tells you instantaneous speed, but it's difficult to use it to determine a long-term average speed.
2011-01-13 06:55:27Good one from Scott Mandia
John Cook

john@skepticalscience...
121.222.100.112
If I measure the height of my six year old at the end of every year he is definitely exhibiting a growing trend. If I measure him at the end of every day for one week I might see no trend at all. I certainly cannot tell the world that my son has stopped growing!

BTW, this kind of discussion is why I started the climate metaphors thread.

2011-01-14 03:43:21
MarkR
Mark Richardson
m.t.richardson2@gmail...
192.171.166.144

No reply from Soares yet. Either I've missed something major or his conclusions are so utterly stupid that I can't believe a serious journal would publish them.

 This is what I mocked up so far (minus the pictures. I include his Fig.8, a heat flow diagram and the climate change attribution figure from Meehl et al 2004). This is a rough draft, significant editing will be done!

 

 

-----------

 

 

 

A recent paper (Soares, 2010) used correlations between temperatures and CO2 concentrations to conclude that;

“The absence of immediate relation between CO2 and temperature is evidence that rising its mix ratio in the atmosphere will not imply more absorption and time residence of energy over the Earth surface. This is explained because band absorption is nearly all done with historic CO2 values.”

Soares looks at the correlations between the change in CO2 and the change in temperature for periods of a month to a few years. He finds a strong correlation between temperatures in a region and the CO2 amount there a few months later, but not the other way around, which suggests temperatures cause the short term CO2 rise. He looks particularly at the change in CO2 and the change in temperature.

We want to work out whether we live in Soares’ world where CO2 doesn’t cause warming, or in the world inhabited by mainstream physics where theory & direct measurements of increased CO2 warming from satellites say that it does.

What would Soares find if he used his techniques and a real CO2 warming effect exists?

 

We expect the temperature of the atmosphere to change regularly: July has a higher mean global temperature than December, for example. And year to year we have known natural cycles like El Nino which transfer heat between the oceans & atmosphere and change cloud patterns. What we see is a very strong relationship between El Nino and warmer years, and La Nina and cooler years – if there were no global warming nor cooling then temperature changes of 0.3 or 0.4 C in a year would happen, which is what we see in the graph: with a range from about -0.3 to +0.3 or, in total, 0.6 C.

 

In a world where the CO2 global warming effect is real, the expected global warming from CO2 for the past century has been less than 0.02 C/year, so on Soares’ graph, if the CO2 warming effect is real, you expect an average value for the temperature change in a year to be positive, but no more than 0.02 C. And you still expect the natural transfer of heat from the atmosphere to the oceans and back again – so you expect something like 0.02 C  0.60 C.

Soares is looking for a signal that is tens of times smaller than the noise and, with a relatively small number of data-points he claims he hasn’t found it!

What we actually expect is accelerating global warming going with our accelerating CO2 emissions (because our emissions are constantly increasing), so what you expect to measure on that graph is a positive average temperature change with a slowly increasing gradient, which is what Soares finds, although amongst too much noise to find a statistically rigorous correlation.

But importantly you expect a very small signal amongst noise many times larger than it.

 

Another important point is that Soares has assumed that there is no correlation between any potential CO2 heating and other heating. Which is wrong: when CO2 emissions were low at the beginning of the century the Sun warmed up a bit. When CO2 emissions increased at a middling rate, we also pumped out aerosols which cause ‘global dimming’ and reflected heat faster than CO2 could trap it.