Old, but untold. Trenberth treated us to a trick in his Halloween interview with Bill Sweet by changing the sign on his own most famous quote. As Trenberth now tells it:
One cherry-picked message saying we can’t account for current global warming and that this is a travesty went viral and got more than 100,000 hits online.
The email in question actually bemoaned how Trenberth couldn’t account for the LACK of global warming:
The fact is that we can’t account for the lack of warming at the moment and it is a travesty that we can’t.
Global warming… LACK of global warming. Hey, what’s the difference?
This is Trenberth’s answer to having his doubts exposed by the ClimateGate leak: just cover them back up. Pretend that the revealing email said the opposite of what it actually said and PROBLEM SOLVED. The guy’s a genius. No wonder he rose to the esteemed lead author position.
Of course he’s not fooling anyone who knows what he actually said. Add that lack of warming does have to do with the state of global warming, and most knowledgeable people will grant Trenberth the benefit of the doubt, but should they? Ignorant people will be fooled, and Trenberth has a habit of misleading the ignorant.
Here is Trenberth in a follow-up interview with Sweet (after Sweet was apparently inundated with comments and email calling Trenberth a liar and castigating Sweet for playing softball—yay WUWT):
Sweet: Can you say something about the widespread belief that solar activity somehow accounts for the temperature changes we’ve seen in recent decades?
Trenberth: That’s easily disproven. It’s nonsense. Since 1979 we’ve had spacecraft measuring total solar irradiance, and there’s been no change—if anything the sun has cooled slightly. There’s nothing in the record that indicates that the sun is responsible for any of the warming in this period.
Trenberth knows full well that “solar activity” refers primarily to solar-magnetic activity, which varies by an order of magnitude over the solar cycle, while total solar irradiance is almost invariant over the solar cycle (which is why it is called the solar constant). Does he really think he can disprove the theory that 20th century warming was caused by solar activity without looking at anything but the least active solar variable?
Again, the knowledgeable will not be fooled, but it is perfectly clear that Trenberth’s intent in this instance is to deceive the ignorant. He is also providing us with an example of what he was talking about in his original IEEE interview when he said:
Scientists almost always have to address problems in their data, exercising judgment about what might be defective and best disregarded.
That pesky data about solar-magnetic activity and earthly temperatures being highly correlated? (“The long term trends in solar data and in northern hemisphere temperatures have a correlation coefficient of about 0.7 – .8 at a 94% – 98% confidence level.”) “Best disregarded.”
And it is easily done. Just change out “solar activity” for the least active solar quantity and, voilà. As easy as replacing “lack of global warming” with “global warming.”
As J.R. Ewing put it, “once you give up integrity, the rest is easy.”
Any other “lumpies”? (Santa must have had anti-CO2 alarmists in mind when he chose coal for the bad. Like crosses for vampires.)
Discover more from Watts Up With That?
Subscribe to get the latest posts sent to your email.
tallbloke says:
December 27, 2010 at 12:17 pm
“The standard deviation of GSN over the interval 1749-1876 was 28. We think it should be 43.”
Well, it all depends on whether you are considering solar variability over the 130 years from 1749-1876 or solar variability over the entire period from 1749-2010.
For 1877-2010 the SD for GSN is 44, and for the [official] SSN 46, so no real difference with the earlier period.
might be seen by some as another step in the direction of massaging the sun into having less variability.
That is because those ‘some’ do not wish to study or even contemplate the evidence for the revision.
I’m happy for there to be two schools of thought on this
This should not be so, as the data should speak. And the data does not speak with forked tongue.
How is it going with extending the geomagnetic analysis to earlier times before 1876?
Solid back to 1836.
Joel Shore says:
December 27, 2010 at 10:33 am
(One way to incorporate this would be to have the “neutral level” itself be a function of the SSTs since the is effectively what is going to be the case…because as the SSTs rise and the atmospheric temperatures rise also, the amount of heat that the earth radiates back out into space is going to increase.)
Yes, I had given some thought as to how much the equilibrium value might alter with changing temperature. I came to the conclusion, having studied the relationship between OLR, TSI and the SOI, that it doesn’t make a big difference. Cloud variation caused by solar activity is a bigger issue.
Thanks for the feedback.
tallbloke says:
December 27, 2010 at 11:02 am
rather than a fact established by peer reviewed referenceable science.
Snide and disingenuous comments about peer-review and facts are two-edged swords, cutting sharply both ways.
Here is then a fact established by peer reviewed referenceable science:
http://www.leif.org/research/2009JA015069.pdf
showing that solar activity [measured by the sun’s magnetic field in the heliosphere] in the 1835-1875 interval was comparable to that in 1950-2000. See e.g. Figure 10.
Yet the GSN for the first interval was 43, for the second 72. Since the heliospheric magnetic fields [a ‘fact’ according to your logic] were the same, the GSN should have been the same, etc.
Bob Tisdale says that there should be :
He “eagerly awaits.” Dude, you can pick ANY temperature record and you can pick ANY measure of solar activity, and you will see that, just as I said:
Are those broad strokes the least bit contestable? Bob contests that they support a sun-climate link by doing what Trenberth did. Asked what he thought of the competing theory that temperatures were being driven by solar magnetic activity, not by CO2, Trenberth’s answer was to substitute the solar constant for solar activity, in effect pretending that solar magnetic variables and solar magnetic theories of warming don’t exist. Bob also fixes on TSI and pleads ignorance of solar-magnetic theories of warming, begging me to please tell him what other physical effects of the solar flux and other components of solar activity might need to be taken into account. In other words: what’s wrong with completely omitting all solar magnetic variables from his analysis?
What’s wrong with ignoring competing theories? Why not just pretend they don’t exist? Is this really the rock that Bob wants to defend?
Trenberth’s method for evading the sun-climate link was the original subject of this post. Bob discovered in the Usoskin paper I cited a second method of evasion that has been employed not just by Usoskin but also by Benestad, Frolich and several other solar scientists, so we started talking about it. This is something I have been writing about for years, how the solar scientists are all pretending that for a sun-climate link to be indicated, warming would have to be accompanied by rising levels of solar activity. If steady peak levels of solar activity coincide with warming, that is supposedly inconsistent with solar warming. Are they really that fruity, or are they stating this post-1980 exception in order to protect their funding?
Bob apparently thought his Usoskin find had caught me in some kind of mistake or faux pas and decided that it was illegitimate of me to start criticizing this second ploy for evading a sun-climate link. Having started down this road, it seems he can’t let it go. Well, if he wants to deny that warming, accompanied by steady peak levels of solar activity, is consistent with a sun-climate link, there are only two ways to go that I know of: there’s the Usoskin et al. evasion, and there’s the Trenberth evasion. (IPCC reports are built around both. They use the Usoskin evasion to help justify the exclusion of all solar variables but TSI from their models.)
Both of these evasions are completely untenable. They are not even close to legitimate. This is no place to make a stand.
As for any flaw that Bob wants to point out in my ocean oscillations post, I obviously welcome any rational input.
[snip – your comment is way over the top, Dr. Trenberth may have made mistakes, but let’s not go into this defamatory territory here – moderator]
Leif Svalgaard says:
December 27, 2010 at 1:39 pm
Snide and disingenuous comments about peer-review and facts are two-edged swords, cutting sharply both ways.
Take it easy Leif, I made the tongue in cheek comment about facts and publication because you didn’t acknowledge that the Kraviva et al paper had just been published (23rd Dec) in your first reply. I find all the work done on reconstructing the solar past fascinating and informative. It’s not my job to take sides, nor is it yours to force me to. I await with interest your further pushing back your study past the Dalton Minimum. I can’t remember how far back you said the usable magnetic records go. What are the prospects?
Alec Rawls said:
Yes…That is exactly right. If they have found a correlation between plots of solar activity (or some proxy of solar activity) and temperature over hundred of years, then temperature continuing to rise while solar activity remains constant is inconsistent with such a correlation. If instead they had found a correlation between graphs of something else, like solar activity being at a steady high level and temperature trends being positive then PERHAPS the last 30 years would not be inconsistent. However, that is not the correlation that they have reported finding and you have presented no evidence that such a correlation exists for the period before 1980 nor that, if it does, the correlation is such that the numerical rise in temperature seen since then is consistent with it. What Usokin et al. found was that the correlation that they had found between solar activity and temperature broke down in a big way after 1980, so much so that thought it better not to even show the data past then because it would look so lousy (although there are other plots out there where you can see how badly it does indeed break down).
You have also not explained why one expects temperatures to continue to rise so significantly for such a sustained period after the solar activity has reached a maximum, given that we know that as temperatures rise, the Stefan-Boltzmann Equation shows that the radiation from the earth and thus any forcing will decrease. I would just love to see the reaction around here if climate scientists used a model for the temperature rise from greenhouse gases increases that didn’t include the feedback implied by the Stefan-Boltzmann Equation! People would be yelling “fraud” and worse!
In other words, you’ve got nothing, and your continued contortions on this thread are frankly embarrassing.
Alec Rawls says:
December 27, 2010 at 1:46 pm
Asked what he thought of the competing theory that temperatures were being driven by solar magnetic activity, not by CO2, Trenberth’s answer was to substitute the solar constant for solar activity, in effect pretending that solar magnetic variables and solar magnetic theories of warming don’t exist.
Alec, have you read Trenberth’s paper on variation in insolation and its potential impact on future temperature?
http://tallbloke.files.wordpress.com/2010/12/trenberth-solar2009.pdf
Alec Rawls says: “Dude, you can pick ANY temperature record and you can pick ANY measure of solar activity, and you will see that, just as I said.”
Dude? No one in my lifetime has called me Dude, and I don’t expect to be called Dude again.
I presented a graph to you earlier using HADCRUT global temperature anomalies and scaled TSI:
http://i52.tinypic.com/2mq4xax.jpg
The TSI was scaled so that it represented about a 0.1 deg C variation for a change from solar min to solar max, which is what one would expect.
You then replied, “But TSI is not ‘solar activity.’ TSI is ‘the solar constant.'” I didn’t want to argue with you that the belief that TSI was constant departed with the advent of satellite-based solar irradiance measurement. So I asked you, Alec, to present the data as you would like it presented. It seemed logical to me then that since you are a self-proclaimed expert on the impacts of solar variations on global temperature that you must have created a multitude of graphs comparing solar datasets of your choice to global temperature anomaly datasets, also of your choice. All it would have taken was for you to copy a link from the numerous solar threads you now claim to have written for your blog. And of course you proved me wrong; you have failed to produce one comparison graph and discuss it so that all who are reading this thread could understand your point.
So in an attempt to put this to rest, I’ll use Sunspot Numbers, scaled so that the data represents about a 0.1 deg C variation for a change from solar min to solar max (for the last 3 solar cycles), and compare them to the same HADCRUT global temperature anomaly data. I’ve also “centered” the scaled sunspots for appearance purposes. I’ve start both datasets in January 1900, so that we get the long-term view. And I’ve added linear trends for those who might be interested:
http://i51.tinypic.com/4lo8ev.jpg
As is clearly visible, solar variations cannot explain the warming during the second half OR during the first half of the 20th century. I could switch back to TSI and show you the same thing.
Since you won’t present a graph of your own making, you’ll have to use that one, and based on it, please explain how solar variability could possibly have driven the global temperatures since 1900. If you disapprove of my scaling, please justify another scaling factor. But I’ve used the same scaling in other solar-related posts. Here’s a link to one that was cross posted here at WUWT:
http://wattsupwiththat.com/2009/03/05/ipcc-20th-century-simulations-get-a-boost-from-outdated-solar-forcings/
Note the comments by Leif Svalgaard, a solar physicist. If my scaling was out of line, it’s likely Leif would have commented. But maybe he missed it. Since he’s been active on this thread already, if my scaling is out of line, I’ll ask for his recommendation.
Here’s the sunspot data since the start of that dataset. Maybe you or another person on this thread can use it as part of the discussion. It’s also smoothed with an 11-year filter to help illustrate when the sunspot data peaked in the second half of the 20th century:
http://i52.tinypic.com/xbmut5.jpg
If you like, I can prepare a similar graph using TSI data.
You wrote, “Bob also fixes on TSI and pleads ignorance of solar-magnetic theories of warming, begging me to please tell him what other physical effects of the solar flux and other components of solar activity might need to be taken into account.”
Wrong. Quote me if you can where I pleaded “ignorance of solar-magnetic theories of warming”, and begged you to please tell me “what other physical effects of the solar flux and other components of solar activity might need to be taken into account.” The comment you are referring to is here:
http://wattsupwiththat.com/2010/12/24/lump-of-coal-award-to-ipcc-lead-author-kevin-trenberth-for-hiding-the-decline-or-the-lack-of-increase-in-global-temperatures/#comment-559781
I simply asked you, as have others on this thread, to produce one graph that compares the solar variable of your choice to a global anomaly dataset of your choice. Have you done so? No.
You wrote, “Bob apparently thought his Usoskin find had caught me in some kind of mistake or faux pas and decided that it was illegitimate of me to start criticizing this second ploy for evading a sun-climate link.”
Wrong. As I’ve repeatedly noted on this thread, my objection to Usoskin et al was their use of the two Mann et al hockey-stick reconstructions. It’s you who attempted to spin your inclusion of that paper in your post and redirect the discussion.
You asked, “Is this really the rock that Bob wants to defend?”
I’m defending what the data presents. As noted before, you, Alec, have been asked numerous times to present the data in a fashion that illustrates your argument, and you have failed to do so. There are a number of ways that the readers here at WUWT will take that failure on your part, and few of them will be in your favor.
You wrote, “As for any flaw that Bob wants to point out in my ocean oscillations post, I obviously welcome any rational input.”
About 50 of my posts have been cross posted here at WUWT, and many of them discuss the processes that drive the ocean oscillations and their impacts of global temperature. If my posts were not rational, Anthony would not continue to post them. We’ll eventually get to your mistake in your ocean cycle post.
Bob Tisdale says:
December 27, 2010 at 6:14 pm
If you disapprove of my scaling, please justify another scaling factor.
Bob, have you read Nir Shaviv’s JGR paper “using the oceans as a calorimeter”?
He has an easy read version on his site here:
http://www.sciencebits.com/calorimeter
Highly recommended for anyone interested in how the oceans accumulate heat.
Tallbloke says: “Bob, have you read Nir Shaviv’s JGR paper ‘using the oceans as a calorimeter’?”
Nir Shaviv’s JGR paper “using the oceans as a calorimeter” was discussed here at WUWT over a year and a half ago:
http://wattsupwiththat.com/2009/04/15/the-oceans-as-a-calorimeter/
As always, Leif’s comments are worth the read.
Tallbloke says: “Highly recommended for anyone interested in how the oceans accumulate heat.”
By breaking the ocean basins into tropical and extratropical subsets, I’ve shown that much of the OHC gains are driven by ENSO, SLP, and the AMO:
http://bobtisdale.blogspot.com/2009/09/enso-dominates-nodc-ocean-heat-content.html
And:
http://bobtisdale.blogspot.com/2009/12/north-pacific-ocean-heat-content-shift.html
And:
http://bobtisdale.blogspot.com/2009/10/north-atlantic-ocean-heat-content-0-700.html
Joel Shore seems to think that Usoskin’s correlation results imply (if they are valid at all) a mechanism by which changes in the level of solar activity are causing temperature to change. That misunderstands what a correlation study is. Usoskin did not hypothesize a particular physical mechanism and then estimate its parameters. Instead, he was taking a preliminary step: “let’s look first to see if there is a correlation.”
Turning to physical mechanisms that could account for a correlation between solar activity and temperature, it can only be the level of solar activity that affects temperature. It can’t be the fact that the level is changing. Come on.
With his appeals to Stefan–Boltzmann, Shore is also demanding to know at what temperature steady peak levels of solar activity would equilibrate. Obviously that is a complicated question, and the idea that somebody in a comment thread should be scorned for not answering it is absurd. In general, answering this question requires modeling the physical mechanism. For instance, if the GCR-cloud theory is correct, and we can nail down the parameters of how solar activity blocks GCR and reduces cloud formation, then a GCM that includes modeling of this sun-cloud behavior would be able to tell us at what temperature steady peak solar activity would equilibrate.
I too would like to know whether the warming effects of our recent 80 year grand maximum of solar activity managed to reach equilibrium before solar activity fell off in 2005, but that’s a final step. The first step–what we have now–is numerous correlation studies which indicate that there is some mechanism by which solar magnetic activity has a powerful effect on climate.
We need to incorporate this evidence for a sun-climate link into our projections, even if we don’t know the mechanism. If we fail to account the likelihood that 20th century warming was caused by high levels of solar activity then these solar-warming effects get misattributed to CO2, giving the eco-lunatics an excuse to unplug the modern world.
tallbloke says:
December 27, 2010 at 3:41 pm
Take it easy Leif, I made the tongue in cheek comment about facts and publication
Then mark it as in-cheek, although I think you somehow mean it, judging from the word ‘crucial’.
because you didn’t acknowledge that the Kraviva et al paper had just been published (23rd Dec) in your first reply.
since I think it is no good, I see no reason to comment on it.
What are the prospects?
1720s, 1740s, and 1760s
Bob Tisdale says:
December 27, 2010 at 6:14 pm
Here’s the sunspot data since the start of that dataset. Maybe you or another person on this thread can use it as part of the discussion. It’s also smoothed with an 11-year filter to help illustrate when the sunspot data peaked in the second half of the 20th century: http://i52.tinypic.com/xbmut5.jpg
add 20% before 1945 and you should be good. The rationale is here:
http://www.leif.org/research/SPD-2009.pdf
Alec Rawls says:
And what they claim to have found is that there is a correlation but that it completely breaks down for at least the last 30 years and that is the point that you simply refuse to accept with your hand-waving nonsense.
What they showed the correlation between was the LEVEL of solar activity and the temperature AT THE SAME TIME. You may want to believe that there is a lag or a cumulative effect that you get by integration or what-have-you but that is NOT what they showed the correlation to be. And that correlation breaks down since 1980 (or the 1970s…or whatever). You may think that some other measure correlates better since the 1970s but then you have to demonstrate that this is the case and that it also correlates well prior to the 1970s as they have at least tried to demonstrate for their purported correlation.
Just admit it, you screwed up. It is not that hard to come clean and you will be more respected for it. Otherwise you can just continue to argue yourself into embarrassment and oblivion.
Bob: I am glad to use your graph of HADCRUT temperature anomalies vs. sunspot numbers. As I said, any temperature record and any solar activity record would illustrate how “the planet warmed while the sun was in its grand maximum phase, then temperatures flattened out after the sun went quiet in 2005.” Your graph is no exception.
By Usoskin’s calculations, the recent grand maximum of solar activity started in 1920 and ended in 2000. Your overlay shows how temperatures increased in two-steps-forward one-step back fashion throughout this span. Why the fits and starts in warming? Because ocean oscillations add a noise signal that can easily outstrip the temperature trend in any decade.
Taken separately, the periods before and after this grand maximum show approximately no temperature trend, but that must be largely coincidence. Ocean oscillations could have easily created a surface temperature trend in either direction, regardless of whether planetary temperature (the temperature of the oceans as a whole) was rising or falling.
You look at this graphic and say:
On the contrary. The 80 year grand maximum of solar activity explains the whole century quite nicely. You seem to be denying the significance of ocean oscillations. I thought they were your “bag” baby. (Sorry, I forgot. You’re not a dude.)
Bob Tisdale says:
December 27, 2010 at 7:05 pm calorimeter’?”
Nir Shaviv’s JGR paper “using the oceans as a calorimeter” was discussed here at WUWT over a year and a half ago:
http://wattsupwiththat.com/2009/04/15/the-oceans-as-a-calorimeter/
As always, Leif’s comments are worth the read.
Nir Shaviv answered Leif’s complaints at the link I gave you.
On April 16th, 2009 Paolo M. says:
Solar physicist Leif Svalgaard state that you have tortured the data you used for your paper figure 2. Find his critics at
http://wattsupwiththat.com/2009/04/15/the-oceans-as-a-calorimeter/#comments
on 15.04.2009 at 9:48
Moreover Lean 2000 is an obsolete reconstrunction.
Would you provide an answer? Thank you.
On April 17th, 2009 shaviv says:
Yes, I pulled finger nails until the data said “I give up, I give up!”
o.k., now seriously.
In order to get the cleanest data I used the 24 tide gauges chosen by Douglas 1997 for different stringent criteria (e.g., in geologically stable locations, long records, consistent with other gauges nearby, etc). I used someone else’s tide gauges so that I could not be accused of cherry picking.
Secondly, because I am not interested in long term trends, but I am interested in short term derivatives, I treated the data differently than what other people do. Instead of averaging the station heights and then differentiating, I first differentiated the data for each station and then added the derivatives. The reason is that this way I avoid getting spurious jumps from the start or end of individual station data. Because it can give rise to spurious long term trends and because I don’t care about long term trends, I simply removed any linear trend from the data.
In the graph from 1870 that Lief Svalgaard points to, one cannot see the 11-year signal because the latter only amounts to a few cm amplitude (3.5 mm/yr!). It obviously drowns in the annual noise or the long term trends in Leif’s particular graph. Note that at least over the past 50 years, Holgate sees consistently the same 11-year variations in the data (e.g., referenced here). Of course, because he uses a lot of lower quality stations (177) and/or is not careful to first differentiate and then add the tidal gauge data, he sees somewhat different variations before 1950, than what I find. (Of course, this is not a problem because he does not care about 11-year variations). Anyway, did Holgate torture his data too?
Oh, and the fact that Lean 2000 is used for the TSI is totally meaningless. The correlation with any signal synchronized with the 11-year solar cycle would give the same result. Note that I removed any long term trends from the tide data and from the solar proxies (whether TSI or cosmic rays).
Perhaps Leif could use this opportunity to respond. I hope he does, because this is actually very important.
Leif Svalgaard says:
December 27, 2010 at 8:10 pm
What are the prospects?
1720s, 1740s, and 1760s
Do you mean to say there is good magnetic data in those decades, and then none until after the Dalton minimum?
tallbloke says:
December 28, 2010 at 3:04 am
Perhaps Leif could use this opportunity to respond.
What precisely am I supposed to respond to?
tallbloke says:
December 28, 2010 at 3:14 am
Do you mean to say there is good magnetic data in those decades, and then none until after the Dalton minimum?
No, those were just the earliest. There is good date from 1780 to 1804, then none until 1818, then good from 1818 on. So, unfortunately, there is no data from 1805-1817.
tallbloke says: “Nir Shaviv answered Leif’s complaints at the link I gave you.”
One needs to divide the ocean basins into tropical and extratropical subsets as I noted in my follow-up comment.
Regards
Leif Svalgaard says: “add 20% before 1945 and you should be good.”
Thanks. Here’s the result.
http://i51.tinypic.com/330cpb5.jpg
Leif Svalgaard says:
December 28, 2010 at 6:58 am
There is good date from 1780 to 1804, then none until 1818, then good from 1818 on. So, unfortunately, there is no data from 1805-1817.
Thanks Leif. SO is their any clue as to why there is no data from 1804 to 1817? Any written work from that time commenting on very low values making measurement difficult or anything? Sounds like an interesting project for a historian of science. Any pointers welcome.
Leif, please could you respond to Nir Shaviv’s reply to your objections to his JGR paper I quoted at:
http://wattsupwiththat.com/2010/12/24/lump-of-coal-award-to-ipcc-lead-author-kevin-trenberth-for-hiding-the-decline-or-the-lack-of-increase-in-global-temperatures/#comment-560662
Thanks
tallbloke says:
December 28, 2010 at 8:01 am
Thanks Leif. SO is their any clue as to why there is no data from 1804 to 1817?
War [Napoleon]
tallbloke says:
December 28, 2010 at 8:03 am
Leif, please could you respond to Nir Shaviv’s reply to your objections to his JGR paper I quoted at:
What precisely? The reply as stated is not specific enough [and just describes various forms of data torture].
Alec Rawls says: “By Usoskin’s calculations, the recent grand maximum of solar activity started in 1920 and ended in 2000. Your overlay shows how temperatures increased in two-steps-forward one-step back fashion throughout this span. Why the fits and starts in warming? Because ocean oscillations add a noise signal that can easily outstrip the temperature trend in any decade.”
The first problem you have is the actual magnitude of the solar signal that’s present in the global surface temperature data. It reflects a signal that’s about 0.1 deg C from solar min to max. Unless you can provide and explain an unknown amplifying mechanism that does not exist in the global temperature data, the variations in solar can only create the trend shown in the graph I provided, or about 0.002 deg C/decade, while global temperatures rose at a rate of 0.07 deg C/decade.
http://i51.tinypic.com/4lo8ev.jpg
Since you continue to rely on Usoskin et al, I’ve got a question for you. Did Usoskin et al fail to mention that they needed an amplifying mechanism to make their wiggle matching work, or did they include that note and you failed to recognize its importance?
The second problem you have is your continued reliance on the Mann hockey stick data used by Usoskin et al. Enough said on this thread about Mann et al.
That brings us to the ocean oscillations, about which you stated, “add a noise signal that can easily outstrip the temperature trend in any decade.” And you added, “You seem to be denying the significance of ocean oscillations.”
There’s no denial on my part. Your comment is based solely on your incorrect assumptions, not on data. Obviously you’ve never removed ocean oscillations from the global temperature record; otherwise you would not have written what you did. Removing them is relatively easy. Let’s use global SST data (it’s not as noisy) and start eliminating them one by one. We’ll remove the effects of the AMO first.
The AMO data is North Atlantic SST anomaly data that have been detrended. But since you might want to argue that the trend is also natural, we’ll leave in the trend and use the North Atlantic data. The following graph compares the North Atlantic SST anomalies (0-75N, 78W-10E) to the global data (90S-90N). The variations in the North Atlantic data are much greater than the global data, but the trends are remarkably similar.
http://i55.tinypic.com/2bxyqp.jpg
If we assume the Atlantic represents 30% of the global ocean surface area and that the North Atlantic represents half of that, then the North Atlantic data cover 15% of the surface area of the global oceans. We’ll scale the North Atlantic data by 0.15 and subtract it from the global data. I’ve smoothed the data with a 13-month filter so that you can see the before and after results in the next graph. Some of the multidecadal variability has left the data.
http://i55.tinypic.com/sn1bva.jpg
Next is the PDO. The Pacific Decadal Oscillation data is the leading principal component of North Pacific (north of 20N) SST anomaly data, where the data in each 5X5 deg grid has been detrended. The PDO is also standardized, and this exaggerates the PDO signal. The PDO is NOT, definitely not, the SST anomaly data of the North Pacific, north of 20N. It is a statistically created dataset that represents the pattern of the SST anomalies, but it does not represent the SST anomalies themselves. And that’s the mistake you made in your ocean oscillation post. You assumed the PDO represents the SST data of the North Pacific, north of 20N. (That and assuming your comparison of Sunspots and global temperature anomalies was valid.) The PDO and the SST anomalies of the North Pacific (north of 20N) are actually inversely related. The ever-observant Tallbloke noticed that. Here’s a link to the post that shows the relationship.
http://bobtisdale.blogspot.com/2010/09/inverse-relationship-between-pdo-and.html
The next graph compares the statistically created PDO data to the SST anomalies of the North Pacific north of 20N (20N-65N, 105E-105W) and the global data with the North Atlantic removed. It shows how out of scale the PDO data is with respect to the actually SST anomalies of the same area of the North Pacific.
http://i52.tinypic.com/2rcc0ld.jpg
Let’s remove the PDO from that graph because it limits the ability to see the relationship between the North Pacific and the modified global data. The North Pacific data does have greater variability (It’s actually very similar to the North Atlantic data because the SST anomalies of the KOE, through teleconnections, are one of the drivers of the AMO) and that variability is an aftereffect of ENSO as discussed in the post linked above.
http://i54.tinypic.com/2u8kl79.jpg
Let’s remove the North Pacific data north of 20N, and for argument’s sake, we won’t detrend it. That subset represents about 11% of the global data. We’ll scale that data by a factor of 0.11 and subtract it from the Global data that’s had the North Atlantic data removed. The next graph (smoothed with a 13-month filter) shows the result. The multidecadal variations have decreased more. But the only thing that’s happening is the global SST data is flattening.
http://i55.tinypic.com/vdnus3.jpg
Next is the ENSO signal. We’ll use a common proxy of ENSO, NINO3.4 SST anomalies, for that. In the following short-term graph, the NINO3.4 SST anomalies have been scaled by a factor of 0.09. The wiggles in the NINO3.4 data are approximately the same magnitude as the resulting wiggles in the global SST data. Note also that I’ve lagged the NINO3.4 data by 2 months to align the leading edges of the 1997/98 El Niño.
http://i52.tinypic.com/r76az9.jpg
Here’s the long-term comparison:
http://i54.tinypic.com/20krk1v.jpg
Next: subtract the scaled NINO3.4 data from what’s left over from the other steps and compare it to the original global SST anomaly data. The multidecadal variations are greatly reduced and so are the year-to-year variations caused by ENSO. And the trend dropped a bit.
http://i53.tinypic.com/25s7kar.jpg
That’s it for the three major ocean cycle signals.
Let’s look at the solar data again. I’ve adjusted the pre-1945 Sunspot data per Leif Svalgaard’s comment above:
http://wattsupwiththat.com/2010/12/24/lump-of-coal-award-to-ipcc-lead-author-kevin-trenberth-for-hiding-the-decline-or-the-lack-of-increase-in-global-temperatures/#comment-560504
Thanks, Leif.
The next graph compares the sunspot data to the adjusted sunspot data (per Leif). The higher pre-1945 sunspot data is not going to help your argument, Alec, because it will decrease the trend.
http://i51.tinypic.com/330cpb5.jpg
Here’s a graph of the adjusted and scaled (0.0008) sunspot data and the residual of all of the adjustments to global SST anomaly data. Removing the additional variability of the ocean cycles does not come close to accounting for the difference between the scaled solar signal and the adjusted global SST anomaly data. I’ve even given you the benefit of doubt, Alec, by not detrending the North Atlantic and North Pacific data. Based on a visual comparison, the scaling of the sunspot data appears a little high. But keep in mind we established the scaling for the land+sea surface data which has greater variability.
http://i55.tinypic.com/1xyqs6.jpg
A couple of notes: I am in no way representing that the residual data above is a product of Anthropogenic Greenhouse Gases. In fact, I argue against it at every opportunity I get. Much of the rise in global SST anomalies results from the fact that the SST anomalies of the East Indian and West Pacific Oceans (60S-65N, 80E-180, or about 25% of the global ocean surface area) can and do rise in response to both El Niño AND La Niña events. I’ve written posts about that for almost two years. Also, there is little to no evidence of anthropogenic warming in the Ocean Heat Content data from the NCDC (Levitus et al 2009). Most if not all can be explained as a product of natural variables: ENSO, AMO, and Sea Level Pressure in the form of the NAO and NPI.
You may try to argue that the oceans are somehow damping the effects of the variations in solar irradiance. But the land surface data is not governed by that damping and it also does not reflect the variations in the 11-year cycle that you would need to make the solar variations a major part of the rise in global land surface data.
The ball’s in your court, Alec.