As always your material is invaluable to lay people like myself who care to get educated. The tragedy right now is the scientific illiteracy that is prevalent in this country. Nuances are far to much for a lot of the media to grasp. And the forces of global warming denial are expert at playing on the public’s inability to grasp the the basic science behind global warming.
Sadly, as climate change influences weather more and more people will increasingly “get it”. I hope we take action soon.
Why does Maximum Latewood Density avoid the “divergence problem”?
[Response: You seem a bit confused. The divergence problem primarily impacts Maximum Latewood Density (MXD). (there are similar, but less pronounced issues with TRW). This paper purports to have largely gotten around the traditional divergence problem seen in past studies using MXD to reconstruct temperatures. -Mike]
It seems to me that the warming/cooling trend is even more complex.
As mankind grows and utilizes more of the fossil fuels, the mass of the earth decreases. However, the earth is collecting space dust in massive amounts as it moves throught its orbit around the sun. some years it accumulates very large amounts of dust, thus increasing the earth’s mass.
This loss of mass from burning fossil fuels and the accumulation of space dust varies from year to year.
Thus when we pass through a “clean” area of space, and at the same time burn condsiderable fossil fuels, the Earth mass has a sum decrease. This allows the Gravatational pull of the Sun to pull the earth just a litle closer in its orbit….thus Global warming. As mankind population grows and we utilize more and more fossil fuels, we must get warmer.
A potential solution is to quickly develop more solar energy systems, wind, and clean thorium atomic energy systems; instead OF BURNING MORE OF THE COAL, OIL, AND GAS.
Hopefully we can solve the Warming, so that our civilization can survive.
If the Earth has a mass decrease, the Sun will pull on the Earth less, not more. The force of gravity is essentially…
Pull of Sun on Earth = Mass of Sun * Mass of Earth / Sun-Earth Distance Squared
Less mass for Earth = Less pull of Sun on Earth
Mass & Energy are two sides of the same coin and can be converted between one another. When we burn fossil fuels, yes the exhaust has less mass, but the energy in put back into the Earth system, it is not lost to Space. I believe the main drivers of Mass/Energy loss to Space are the blackbody radiation, and evaporation (so to speak) of very light elements at the top of the atmosphere as individual molecules gain escape velocity. I can’t think of other ways Earth is losing mass or energy, but there might be some.
You’re conclusion is right, we need to develop renewable sources of energy, but the science is a mixed up.
…, but show some divergence in their lower frequency trends. [Emphasis added.]
Lower frequency than a multidecadacal timescale is, say, centennial or millennial timescale. I think it should read higher frquency, such as seasonal timescale.
Or I am completely lost.
[Response: No, you are reading it correctly. The sediment record only resolves multidecadal variations, and on that timescale the two records tend to agree. It is only at the lower frequencies (i.e. longer timescales), namely centennial and longer, where the records diverge. But I added some additional wording to clear up any potential confusion on this point. Thanks! -Mike]
Harry Francis (#4) and Unsettled Scientist (#5) – No. Combustion of fossil fuels does not reduce the mass of the earth, if we assume that the atmosphere is part of the earth. Combustion does NOT convert mass into energy. Nuclear fission/fusion do, but chemical reactions do not.
Now, maybe you’re trying to be more subtle, and I’m missing something… but what you wrote is going to confuse the heck out of people!
Harry @ 4,
This is a joke isn’t it? I am a bit scared that anyone might actually think that burning anything reduces the mass of the earth.
The only energy source that will do that is nuclear and I can’t imaging a few kg will make much difference
Was this their ‘response’ to your feb/2012 paper, or is that somewhere else? I’ve been curious about that.
Rob Wilson says:
8 Feb 2012 at 12:03 PM
Your paper has certainly generated a lot of discussion over the last few days between some dendroclimatologists.
You would appreciate that we are somewhat sceptical of your hypothesis and analyses and are drafting an appropriate measured response to your work.
With kind regards,
Rob, Rosanne, Ed, Kevin, Keith, Tom, Jan, Dave, Ulf, Brian, Håkan and Paul.
[Response: Thanks Rob. Looking forward to your alternative explanation of why these trees are missing/greatly underestimating all of the largest eruptions of the past millennium. I’d be open as anyone to a better hypothesis, and a rigorous reanalysis of the chronologies to insure beyond doubt that the effect we hypothesize is not present. This is the way science works, and I’m looking forward to a constructive discussion. As I’ve stated before, if the paper generates nothing more than that, then I would consider it a positive development. Note also–as I alluded to in an earlier coment above–that with all of the emphasis on the post-volcanic cooling, many observers are missing an equally important conclusion that we draw in the paper (and mentioned in the press release): that if the models we compared to have things about right, then RCS is doing a remarkably good job of reconstructing the long-term variability. This is a positive implication of our study that many are glossing over in their emphasis only on the thing that the reconstruction doesn’t appear to be getting right (large post-volcanic coolings). -mike]
Marcus, hmm. not sure what JefflD has commented on previously so I don’t know what to make of that. But Mike’s comment seems cryptic… Was there an official response from those scientists? And if so, I would very much appreciate a link to it.
[Response: Didn’t intend to be cryptic, just requesting patience. There is no published response at this time. These things need to work their way through the system (i.e. peer review), and it is possible that a comment/reply will ultimately be published. Nature journals have strict embargoes, and there is little more that I can say right now than that. -Mike]
Combustion does NOT convert mass into energy. Nuclear fission/fusion do, but chemical reactions do not.
Actually neither type of reaction does… mass is not CONVERTED to energy, it IS energy (times c^-2).
Chemical reactions release on the order of an eV per atom, atoms weighing several GeV apiece. That makes the mass loss, if the energy is radiated out to space, on the order of 10^-9 of the mass of the fuel burned. Uranium fission releases some 200 MeV per nucleus, which weighs over 200 GeV. Mass loss order 10^-3 of the mass of the fuel.
… and “Harry Francis” surely is a Poe… as the saying goes, you cannot make this up
“Harry Francis” reminds me of “Archimedes Plutonium” (see URL http://en.wikipedia.org/wiki/Usenet_celebrity ); the Earth’s mass is approx 5.9736×10^24 kg, so even if we could sequester the CO2 into space it would only remove the tiniest of fractions of this mass, some 20×10^12 kg per year, that’s of order 1ppt (parts per trillion). On a LinkedIn forum some similar comment wasted a lot of time & bandwidth – don’t let the same thing happen here (:-)
differences in seasonal sensitivity between the two proxies allow a more compete understanding of the climate system and likely explain disparities in inferred climate trends over centennial timescales.
I hate to nitpick, but shouldn’t it read ‘a more complete understanding of the climate system…’?
“Dendro for Dummies” request: Could you clarify what they did in order to get around the ‘divergence problem’? If this is the major advance, as you assess, the method should be of interest. I see that they state they found no evidence for divergence, but if there’s an account of why not, I’m missing it — is it simply the extent and replicatedness of the record? or the inclusion of living lakeshore trees along with the subfossil ones? or statistical smarts that I missed? or…?
Good stuff, but a wish,k maybe even for a short tutorial here at RC:
1) Reconstructions can differ by algorithms applied to same data.
2) Data selection can be land or land+ocean.
3) A reconstruction can claim to represent:
100% a) Earth (SH+NH)
50% b) SH
50% c) NH (0deg-90deg N)
or parts of NH:
30% d) 23.5deg N – 90deg N (Tropic of Cancer, I’ve seen that somewhere.
25% e) 30degN – 90degN
18% f) 40degN – 90degN
6.5% g) 60degN – 90degN [which after all has some interesting proxies)
4) Spaghetti diagrams can unintentionally mislead people , as I discussed on p.142 of Strange Scholarship inn the Wegman Report:
‘- Different reconstructions cover different geographies, and in particular, those focused on (land-dominated) NH extratropics are expected to vary more than the entire NH, which in turn varies more than global.
– Human eyes tend to notice the outer edges of the spaghetti graph more than the density of lines between.”
(I think that’s right, tell me if wrong, but these explain why I like the IPCC AR4 density graph as a display technique.)
Also, I’d add that people sometimes interpret the multiplicity of lines to think that researchers have no idea what was going on and must disagree (more than they do).
5) SO, I’D WISH:
a) For a short tutorial to set people’s expectations about expected differences among geographies.
b) For any listing of reconstructions or graph thereof to be explicit about land/ocean and geography claims.
Several people have already commented on #4 and #5, but even if the mass of the Earth did change (from solar dust, shooting probes to other planets, etc.) that would not effect it’s orbit–Galileo showed long ago that different objects fall at the same rate regardless of their mass–the same is true of objects orbiting the sun.
1. How accurately do density variations in tree rings reflect changes in solar insolation through time?
2. How do we know solar insolation changes over the last two millennia unquestionably result from long-term oscillations of orbital configurations?
3. Is there any other evidence to support the claim that the forcing of solar insolation changes over the last two millennia was up to four times as large as the net anthropogenic forcing since 1750?
4. Is there any way to know if the wood density data supports the general circulation models (to conclusively demonstrate summers have cooled substantially), or if the models were tweaked to match the data?
Mike & Gavin, comments please.
Comment by Louis Hooffstetter — 9 Jul 2012 @ 9:36 PM
Harry is wrong. The biggest problem is that we have epidemic obesity in the USA, greatly increasing the mass of the Earth, so it is being gravitationally attracted closer to the Sun, causing global warming. So we all need to go on a diet. (Sorry, couldn’t resist, even if it is not April 1.)
Michael Mann, thank you for your book. I’m working my way through it, increasing my understanding of the science, and posts like this make it clear that that you are real scientists interested in new information, that there is real science being done, still developing in some ways, separate from the political theater.
Let me add my agreement with what AIC said in #25. Your book was the first downloaded to my Kindle, and I found it a thoroughly satisfying read. I know you would rather immerse yourself in the science, but your public work (interviews, talks, articles, books) is very important also, so please keep it up. It’s been thrust upon you, but you’re bearing up fine! :)
About to N-Scan chart above, what do you say to someone who jumps on it and say “Aha! so it was warmer than today in Roman times!”? At least as high latitudes. Will the N-Scan results be incorporated into a global study?
[Response: Thanks Toby, appreciate the kind words and support. Re N-Scan, well sure cherry-picking and selective citation of individual records is a favorite past time of climate change deniers. Just have to call it out for what it is. -mike]
Actually what Galileo supposedly proved or expanded on was that two objects of different insignificant masses would fall at approximately the same rate in the presence of a significant gravitational field as long as the medium through which they were falling causes insignificant differences to the objects…because the main gravitational force would completely overshadow all the lesser forces working on the two masses differently. The “two masses fall at the same rate” idea is a rough rule of thumb that works good enough in most situations, but not exact or proper law of physics. In a vaccuum this rule of thumb is generally very close for objects without significant differences in mass or centrifugal force, because the next most significant force we usually experience is air resistance, so then we can for example make a feather fall as fast as a small sphere…however the magnitude of difference between the compared object’s masses affect the balance/difference of centrifugal force/gravitational pull of said objects between each other and any other nearby objects, such as between bodies orbiting in a solar system.
Because we are already off topic re the mass of the earth and orbits, I can’t help bringing up the fact that there is no such thing as a centrifugal force. It is a fiction of a rotating frame of reference. This is why it is referred to as a fictive force. Problems of understanding physical reality can result from this confusion.
Of course the Esper et al N-Scan reconstruction and accompanying chart have been all the rage across the denialosphere today, with many headlines trumpeting (for perhaps the thousandth time) that, finally, here is definitive “proof” that the planet is cooling, not warming.
I find it extremely telling that none will of course comment on the rapid post-1900 warming reflected in both the reconstruction and the instrumental record; “ignore the incline” is the new denialist motto, I suppose. But I’m nevertheless with most here in welcoming the article and having a chance to dig deeper into it, especially given that it seems it will help with both small-scale fidelity and the divergence problem. But on first read, Esper & gang do appear to be “extrapolating quite a bit”, since the long-term cooling trend they found is not reflected at all in proxy data such as those from Lake Tanganyika. A “stretch”, indeed.
It’s a shame that the authors of the Esper paper were so irresponsible in talking about the “radiative forcing” of orbital insolation changes. I already have found some people on the internet claiming that sensitivity is low because a ~6 W/m2 forcing only produced 0.3 C or so temperature change; some of the blame on that misunderstanding goes to the paper.
It was just a sound byte that takes away from an otherwise interesting and probably very useful study. Hopefully it helps out the dendro community!
[Response:I don’t think they were irresponsible at all Chris. The 6 W/m2 figure is not an estimate of theirs–they cite Berger & Loutre (1991) Quat. Sci. Rev. 10:297-317 for that. The main finding of the study is that ring widths and densities can return different long term trends. Three different types of evidence (their Fig. 3, Fig. S1, and considerations involving detrending specifics of the two ring variables), argue that it’s the density values that are most likely to be correct. Given that ring widths are far more commonly measured (and hence used in reconstructions), that’s potentially real important.–Jim]
[Response:As noted in our piece, where the authors can rightly be criticized is for vastly overstating the implications of their findings w.r.t. to multiproxy reconstructions of annual hemisphere-mean temperature. As we explain in the piece, that hypothesis appears to fall apart when you look at the actual reconstructions. Its disappointing the authors didn’t go the extra step that we did to check if their hypothesis holds up in that context. Indeed, this has already led to some misleading headlines e.g. New Scientist http://bit.ly/NjY6UY “Tree rings suggest Roman world…warmer than thought” (No it doesn’t!).” -Mike]
[Response:Since we are all chiming in, I agree with Chris on the forcing statements – the local/seasonal insolation forcing from orbital calculations shouldn’t be compared to the annual global mean radiative forcing over the 20th Century – it’s just not an apples-to-apples comparison – and as Chris noticed, it is easy to be mislead over what that means. – gavin]
> ring widths and densities can return different long term trends.
Jim, you’re the tree biologist, can you give or point to an 8th-grader-comprehensible take on what makes the differences people see, in terms of the actual tree’s life? Something that would help with questions like these that occur to me? (Pointer’s fine, I’m too busy for a while to try to find the best way to get into this whole area.)
What’s the individual tree doing internally when it’s making a thicker or thinner annual ring? What’s it doing when it’s making the denser or less dense ring? And if you compare trees growing together, how well correlated are they year by year.
[Response:To make a long and complex story overly simplistic, the tree is attempting to optimize water transport efficiency and mechanical strength/flexibility relative to carbon expenditure needed to do so. Much of this is driven by hydraulic considerations at very small scales, i.e. the internal diameter of a wood cell (i.e. conifer tracheid). As seasonal water stress increases, as it usually does over the growing season, the tree has to keep from having its cells cavitate (form an embolism) due to the enormous tension on the water columns in the xylem (a tension which typically shrinks the diameter of the tree during mid-day). Once cavitated, that cell is useless for water conduction, weakening the overall transpiration stream. If you decrease the internal diameter of the tracheids, the water column will not cavitate nearly as easily (exponential function there due to support from the cell wall adhesion), and increased cell wall width and lignification increase the mechanical strength so they don’t collapse. Therefore, maximum density culminates near the end of the growing season.]
How’s each tree allocating the resource available among making trunk, leaves, roots, new growing branches? And I’m guessing the “resource available” to the individual tree corresponds to what the climatologists want to derive — total and peak/minimum available light and water for photosynthesis; maybe wind, if the tree grows differently over years or decades in windy or more protected locations.
[Response:The question that keeps plant physiological ecologists employed! Trees are nothing if not the world’s greatest optimizers. For the purposes of dendro-climatology, the lower stem radial growth is on the hierarchy of a tree’s carbon allocation priorities, the better. That means it will be responsive to limitations in resources, which is what you want.]
And, long shot — seeing the new topic on Andrill — given that leaf waxes survive in sediments, is there anything in tree rings that correlates with leaf waxes? Anything you could say about leaf constitutents that would vary along with ring size or density, that might end up in annual layers near a site?
[Response:Don’t know about waxes. Leaf stomatal density would be the one that comes immediately to mind as the best possibility.–Jim]
Dave E is wrong is saying that “Combustion does NOT convert mass into energy”. Einstein’s equation E = mc^2 applies to chemical reactions just as much as to nuclear reactions. So the products of combustion do weigh less than the fuel + oxygen that went into them. It is just the the amount of mass lost is too small to be easily measurable in a chemical reaction.
The heat generated still has mass (weight) but assuming that eventually radiates into space, the earth will lose a small amount of mass due to burning fossil fuels.
However that will not alter the Earth’s orbit, because the mass of the Earth cancels out when calculating the acceleration caused by the sun’s gravity:
F = ma = GmM/r^2
(F = force, m = mass of earth, M = mass of sun, r = earth-sun distance, G = Newton’s constant). So:
a = GM/r^2
independent of the mass of the Earth. (The gravitational force gets less if the Earth loses mass, but the Earth’s inertia is less too, and the effects cancel, leaving the orbit unchanged. The interplanetary dust we pick up may or may not change the orbit, depending on whether it carries net momentum – I would have to check on that. But the increase in mass of itself does not).
And of course, none of the above has anything to do with global warming :-)
I’ve been involved in forest based science education for 15+ years and every forester and biologist I’ve asked has said trees are far more likely to be indicators of percipitation and soil nutrients than temperature.
Maybe it shouldn’t be surprising that we have papers utilizing tree ring data that indicate warming and other papers that indicate cooling.
[Response: Then apparently you were misinformed. There is literally decades of research behind the field of dendoclimatology, and it is well known and understood that one seeks different environments depending on what parameters might be of interest. When seeking to reconstruct temperature, one looks for environments (e.g. the boreal or alpine treeline) where temperature is indeed a limiting factor determining growth, and specific sites where stand dynamics etc. are not an issue. There is a wealth of literature on this that you might want to direct your colleagues to. In fact, there are books written on this. Start with Ray Bradley’s “Paleoclimatology” in the side bar of this site (or my book “The Hockey Stick & The Climate Wars” for that matter!). -mike]
“Aha! so it was warmer than today in Roman times!”?
Important to keep in mind, actually, since it completely scuppers another common denialist meme, namely, “Of course the planet is warming, we’re just coming out of an Ice Age!”
Of course, we’re not: AFAIK, the highest post-glacial temps were about 8 kyr back (updated info gladly welcomed.) The 1999 “hockey stick” showed cooling over the ‘handle,’ Kaufman 2009 showed 2 kyr of cooling in the Arctic (prior to the ‘blade’, of course), to name just two examples of studies showing/suggesting ‘elevated temps in Roman times.’
Folks who make these types of arguments aren’t always much concerned with logical consistency, of course, but if they try to make both sides of the argument in the same conversation, that too becomes persuasive–though not necessarily in the way they wish.
When seeking to reconstruct temperature, one looks for environments (e.g. the boreal or alpine treeline) where temperature is indeed a limiting factor determining growth, and specific sites where stand dynamics etc. are not an issue.
Perhaps part of the problem here is that foresters aren’t trying to grow commercial timber near the boreal or alpine treeline, and therefore are ignorant of the fact that in such environments temperature can be the limiting factor.
Got that part. And that’s why the paleo/dendro work isn’t done in the forest plantations but at the edge — edge of an ice cap, latitude or elevation or both — where they know the temperature was the limiting factor.
This is why it’s easy to find tree records that _don’t_ show temperature effects; because there are large areas of the world where they won’t because the temperature isn’t what’s making the tree hit its limit.
I know this is an area where there are lots of different possible causes/limiting factors affecting what’s observed (rings, wax in sediments).
And every one a likely PhD topic :-0
Science is rugby, not a lovefest, as Peter Watts says.
[Response:I have a lot to say on the topic Hank, but no time to say it. That goes for this Esper et al paper, also. I hope to get some time to elaborate fully on why their detection of a long term trend difference between density and width data is real important, why it arises, and what it implies for tree ring analysis. I have a very different take on this study than Mike/Gavin/Eric do. It will require a separate post to do so.–Jim]
Journalists should only be partially blamed for the bad coverage of the latest Jan Esper paper. Some of them wrote stories without interviewing the authors, which is wrong, but the press release issued by JG University in Mainz helps the denialist fringe by including a couple of odd quotes from Esper himself. Take a look at what he says:
“We found that previous estimates of historical temperatures during the Roman era and the Middle Ages were too low,” says Esper. “Such findings are also significant with regard to climate policy, as they will influence the way today’s climate changes are seen in context of historical warm periods.”
[Response: This line should have made it clearer that this was a single location, and I’m not sure what previous estimates he is referencing. The impact on climate policy is a stretch, though this certainly isn’t the only press release that has over-egged that pudding. – gavin]
“This figure we calculated may not seem particularly significant,” says Esper. “However, it is also not negligible when compared to global warming, which up to now has been less than 1°C. Our results suggest that the large-scale climate reconstruction shown by the Intergovernmental Panel on Climate Change (IPCC) likely underestimate this long-term cooling trend over the past few millennia.”
[Response: I think this is also a stretch. Is it ‘likely’ that any new NH reconstruction using N-Scan will have a big difference from the range of trends in the existing reconstructions? I’m not sure, and absent any actual test (which would be relatively easily do-able), I don’t know how he is so confident. – gavin]
[Response: The reference to “THE large-scale climate reconstruction shown by the Intergovernmental Panel on Climate Change (IPCC)..” is most peculiar indeed. IPCC AR4 report showed A DOZEN such reconstructions, each indicated that modern warmth is unprecedented as far back as they reconstructions go (some more than a millennium). In at least ONE case (Moberg et al, 2005) NO TREE RING DATA WAS USED to construct the long-term trend. So the statement in the Esper press release is not only vague but necessarily false! -mike]
[Response: I think Gavin is a bit too generous here. We’ve shown in the article that his hypothesis actually fails—it isn’t even consistent with the variation in the long-term cooling trend between studies that do and do not use tree-rings. Curiously, Esper cites two full hemisphere (not extratropical or high-latitude), annual (not summer), multiple-proxy (not tree-ring only) studies (Mann et al ’99 and Mann et al ’08) as examples of the impact this high-latitude, summer, tree-ring effect could have, and yet those two studies have the largest long-term cooling trend, and show no support for the effect he is talking about. Curiously unmentioned in the paper is the ironic fact that Esper’s own previous Northern Hemisphere temperature reconstruction (Esper et al ’02), which is indeed entirely based on tree-rings, and reflective of extratropical summer temperatures–shows among the least cooling of all the reconstructions, and is perhaps most likely to have suffered from the very bias mentioned in this study. Isn’t it surprising that this isn’t mentioned in the paper? The number of speculative, unsubstantiated—and indeed easily falsifiable–claims in the paper is rather stunning, and detracts greatly from what otherwise looks like a decent contribution. -mike]
I wonder if you guys could please comment on this press release, because it’s very hard for journalists to deal with such vague statements. Do you really think Esper is advocating lowering the tone of the IPCC reports?
[Response: I have no idea. I’d say it was more related to emphasizing the potential implications of one’s own work over anyone else’s – a frequent occurrence in press releases. I generally find it prudent to wait for the work on the implications to be done (for instance). – gavin]
[Response: Gavin is again quite generous. It would appear that Esper’s misleading statements and overstatement of larger implications directly fed the sort of denialist frame represented in the Daily Mail article. It is of course impossible, and unwise, to guess at whether or not that was his intent. -mike]
“Moberg et al (2006): -0.06 ºC/1000yr (0-1900)
Esper et al (2002): -0.11 ºC/1000yr (831-1900)
Hegerl et al (2007): -0.14 ºC/1000yr (558-1900, 30º-90ºN land, Chblend-dark)
Ljungqvist (2010): -0.18 ºC/1000yr (0-1900, 30º-90ºN)
Mann et al (1999): -0.19 ºC/1000yr (1000-1900)
Mann et al (2008): -0.23 ºC/1000yr (300-1900, nhcru_eiv_composite):
This can be loosely compared to the -0.31 ºC/1000yr estimate derived for N-Scan and trends of -0.10 and -0.19 ºC/1000yr at that latitude in summer seen in two model estimates discussed – though note that the model simulations will have smaller trends for the whole hemisphere and for the annual mean.”
Not trying to nitpick but showing different trends from different time periods takes away from your point there. How do we know that any of the shorter series would have similar trends without even covering the same length as the trend they reported.
[Response: Fair point. The trends above were for the maximum length of the reconstructions. If I take the 901-1900 and 601-1900 periods, then I get: Moberg: -0.50/-0.28ºC/1000yr, Ljungqvist: -0.53/-0.31ºC/1000yr, Mann08: -0.62/-0.38ºC/1000yr, Hegerl: -0.28/-0.14ºC/1000yr. All quite negative. I can’t find the N-scan numbers anywhere so I can’t do the trend for these exact periods for that yet (I’ll update if I find them) but note that these are for wider regions than just the N-scan location. The Supp Info has a good graph (fig. S13) showing the differences you get in the long-term orbitally-driven trend (land-only?) for different latitude bands in the models they used. For 30-90N, the trend is expected to be smaller that for 60-70N – by roughly half in ECHO-G and by even more in ECHAM5 (eyeballing, I estimate ~0.1ºC/1000yr and -0.03ºC/1000yr respectively). For the NH as a whole, the modelled trends are smaller still, and actually slightly +ve in ECHAM5 (-0.04ºC/1000yr and +0.01ºC/1000yr). These are just an indication of the latitudinal issues though. Overall, the reconstructed trends are dependent on the exact period looked at, but I don’t see any obvious discrepancy with the N-scan or the modelling results. – gavin]
[Response: Gavin is once again being extremely charitable. The calculations above, frankly, *falsify* the claims by Esper et al regarding the supposed impact of their finding on Northern Hemisphere annual mean temperature reconstructions. There is neither any observational OR theoretical evidence to support their claim that there is a fact. Will Nature Climate Change require a retraction of the associated statements in the paper? -mike]
[Further Response: I looked at the trends for different time scales with the N-Scan data and get -0.48 and -0.52ºC/1000yr for 901-1900 and 601-1900 respectively. Given the model results on the latitudinal relationships, I don’t see that any contradiction between this one study and the larger-scale reconstructions. – gavin]
> Given that ring widths are far more commonly measured
> (and hence used in reconstructions), that’s potentially
> real important.–Jim]
Jim, how is the density information captured?
Is it possible to capture after the fact?
Can it be done from archived photographs, for example?
If you’re given a dendro sample, and you’re the grunt worker/postdoc/tech — how much more is involved in getting density along with width from what you’re given?
And, same as the first question really, if you want to go back and get density info, what do you need? If it’s done on actual wood in storage, does that change over time in storage?
[Response:X rays! You cut a thin transverse section (i.e. perpendicular to the cell axes), and X ray it. How the actual data reading and recording is done I have no idea. It’s a specialized set of procedures and equipment that most labs don’t have. Fritz Schweingruber, from the WSL-Birmensdorf was the one who really put density data on the map in a big way, through an incredible, northern hemisphere collection effort in the 1980s and 90s, about 500 sites total. I read a paper a couple of years ago by some folks who claimed that a blue staining technique combined with regular light produced good correlations with the x-ray data. If so, that makes such data much more accessible, although some of the high end software for image analysis is still exorbitantly expensive, though really powerful also, some of it capable of measuring cell and cell wall characteristics automatically. As for old photographs I’ve never heard of anyone doing that; I doubt if you could estimate density with them but I don’t know. Even with ring size, you could run into problems with optical distortion issues if you didn’t know the lens that was used. I don’t believe the density deteriorates with time as long as the wood is kept dry but not sure on that. The answer to your question is, it’s a LOT more work and money to measure density; you either set your lab up to handle it big time, or you don’t do it. It’s enough work just to do regular dendro especially if measuring early and latewood width in each ring. Also, gonna get to your earlier question in a second here–Jim]
> old photographs
Might be surprising. I recall years back, when digital astronomy was taking over, that a project was taking a huge archive of old astronomical photos — they’re negatives, big thick sheets of emulsion with three-dimensional arrays of silver where the photons went through them. And by doing some kind of scanning (X-rays? nuclear magnetic resonance? I dunno) it was possible to basically do a 3-D image of the photon tracks in the emulsion for each star on each photograph. Turns out those long exposures on old film had accumulated far more information than the positive prints (flat 2-D images). Dunno what became of it.
But it wouldn’t surprise me if old photos have info in them. Negatives are so often used, well, were so often used for science — and remembering those are thick slabs not flat planes.
[Response:Interesting. Old photos would definitely have information in them, I just don’t know how many exist that would really contain all the information you’d need to make them useful in the context where you want quantitative information from each ring. You’d need to know the site location, species, lens focal length, distance from camera to object, and also that the camera was held ~ perpendicular to the item. If you’re just doing demography (counting rings), then they’re potentially much more useful, because those last 3 items then don’t matter.–Jim]
It would be nice if you could put a summary in layperson terms. The denier rhetoric is fierce and taking whatever they can in bits and pieces to use to further the climate change denier movement. A more clear and concise summation of your findings would be nice.
35 physicist said that a = GM/r^2 . I’m sure you must’ve really meant that the acceleration of the sun towards the earth is a = Gm/r^2 and doesn’t depend on the mass of the sun and that therefore the orbit depends only on the Earth’s mass but not the sun’s.
(occasionally one or the other of you find my sense of humor to be lacking in the funniness stuff. this is another one of my feeble attempts at it).
Comment by John E. Pearson — 12 Jul 2012 @ 12:00 PM
Is there any place I can find a list of all the current tree ring data, including when and where the data was collected?
Is there anyone doing research on using high frequency ultrasound to measure tree ring density? Would anyone like to? IMHO an mechanically scanned ultrasound reflectometer should be straightforward to implement and less expensive than Xray technology.
Another thing about the authors’ press release: even if it’s true that the Roman warming globally (as opposed to their one regional sample) is of the same order as that of today or even bigger, their graph shows a cooling trend for the last 2000 years, which should have been emphasised in recent years by an unusually deep solar low. Yet the right hand side of the graph has an upward shift. This does not in any way suggest that anthropogenic warming is not a real or significant effect. On the contrary, it goes against a trend over 2000 years (“a cooling of -0.3°C per millennium”), and I don’t suppose they’ve shown a change in the sign of the solar forcing trend. Without that cooling trend, we would be 0.6°C warmer now and remember also that we have not stopped the underlying cause of warming so it will continue.
I’m curious as to why the error bars are no bigger the older the data gets. Even if techniques have improved since the early reconstructions, wouldn’t there be significantly more uncertainty with older measurements?
Mike, I can see why you may be a tad annoyed that they are making such unsupportable claims after all you’ve gone through. I doubt somehow they are going to be hauled before a political kangaroo court and accused of faking their results, scientific incompetence or dishonesty.
Oh, wait — compare the picture at the press release
(…. red dashed line doesn’t go all the way to the present day)
to their more detailed picture linked in the main post above: http://www.realclimate.org/images/N-Scan.jpg
Do both images come from the full paper?
Or did the press release ‘improve’ by leaving out the recent part?
Hank, the first graph you link to http://www.uni-mainz.de/eng/bilder_presse/09_geo_tree_ring_northern_europe_climate.jpg specifically states the trend which stops around 1900. They could have continued the red dash trend to the present (1900 onward) which of course would show a warm trend but I think their purpose was to show how the past trend looked which is somewhat different from past reconstructions. Have you read the paper and not just the press release?
I think the downward trend line presentation is a good one. Good to see where things were going before mankind became fully dedicated to burning fossil fuels. Adding the data after 1900, I think, would somewhat distort the picture.
Also, what does that trend line say about the pop theory that mysterious oodles of unknown origins are causing a recovery from the LIA.
I’m a physicist who admittedly knows little about climate change .. but I’m trying to learn. I wanted to thank the authors for the excellent article and the good discussion in the comments. It’s enormously helpful to me. I had the same question many others raised as well … a better understanding of the technique used by Esper et al. to overcome the divergence problem. I then read the Esper paper and found that the discussion at this site helped me to follow the arguments in the paper. They also have an excellent description of the Maximum Latewood Density technique. If you were confused by the discussion here, it’s explained well in the paper.
[Response:The idea that this paper was somehow an improvement on, or solution to, the divergence effect, or that the authors developed some kind of new approach to deal with it, is not supported by the paper or the larger collection of existing studies. They make nothing of that issue, barely even mentioning it in passing, and then never again. It’s likely a fortuitous result, possibly related to use of density data and possibly not, but not one that is new–there have been numerous studies in which divergence at decadal scales was weak or absent. If you look at their Fig 3 you can see that the difference between density and ring width data is not great in recent times–they track each other fairly well but diverge as you go back in time. Nor do I agree that the authors were really making any adamant claim regarding the global prevalence of this orbital forcing–though they could have been clearer on that point. But they do provide some good evidence from widely differing sources regarding trends in far N Scandinavia and the sub-Arctic more generally (e.g. their Fig S1), and that ring widths and densities give different trend estimates. Their larger point, that a lot of reconstructions include a lot of northern tree ring sites in them, and these could well be affected by any weaknesses in the ability of existing analytical methods to capture long term negative trends in ring widths. That’s important, because a large number of studies are based on ring widths rather than density data.–Jim]
I have a question that I’m really puzzled about. From the discussion above:
“The authors make much of the importance of long-term radiative forcing due to the changes in the earth’s orbit for millennial timescale temperature trends. They argue that TRW data which fail to record this forced long-term cooling might therefore underestimate variability on millennial timescales more generally, and potentially underestimate the warmth of past warm periods (e.g. medieval and Roman periods). Orbital forcing is indeed substantial on the millennial timescale for high-latitudes during the summer season …”
Can someone please explain to me .. or at least point me in the right direction … what change in the earth’s orbit are we talking about that occurs on the 1000 year time interval? Is this the accepted explanation for both the Roman and the medieval warm periods? Finally, what is this orbital forcing doing now? Thanks!
[Response: The orbital forcing being referred to is (mainly) the shift in precession over the last 8000 years. This is a cycle that has a periodicity of about 19,000 years and relates to the position of the perihelion (closest approach to the sun) with respect to the seasonal cycle. (There is a shift in obliquity (tilt) that also plays a role over this time period too). Currently, perihelion is Jan 3, while in the early Holocene it was in August. The expectation is that if you are closest to the sun during the summer, the summer will be hotter. Thus NH summers are expected to have got slightly cooler over that 8000 year period, and will be expected to start warming up again in a couple of thousand years. But note that these are very slow long-term trends, and cannot be explanations for the much more rapid ups and down around that trend. Thus the Esper et al claim is *not* that the Medieval period climate anomaly was caused by orbital forcing, but rather that with more accurate low frequency variability in tree rings, the medieval period might have been warmer than existing reconstructions show. – gavin]
I looked quickly for mention of other papers mentioning high latitude temperature, comparing that to other locations. Here’s one: http://www.agu.org/pubs/crossref/pip/2012PA002291.shtml
A precise search for drastic temperature shifts of the past 40,000 years in southeastern Europe
“… the geographic extension of temperature anomalies is largely uncontrolled due to the scarcity of quantitative records of sufficient time resolution on the European continent. Here, we propose, based on a recently developed temperature proxy (TEX86), a reconstruction of millennial-scale temperature variations in a Black Sea sediment archive for the last 40,000 years….. In notable contrast to observations from nearby archives, Heinrich events imprinted our glacial temperature record, consistent with a strong reorganization of oceanic circulation and a large spreading of the temperature anomaly from the North Atlantic toward the south-east. Furthermore, in contrast to high latitude records, our Black Sea record lacks of the signatures of Dansgaard-Oeschger interstadials suggesting a decreasing temperature gradient away from the North Atlantic.”
Gavin — Thank you very much for taking the time to clear that up for me. I completely misunderstood the context in which orbital forcing was being used, especially when the title of the paper is “Orbital forcing of tree-ring data”. This is an incredibly valuable service you are providing.
Comment by David B. Benson — 15 Jul 2012 @ 5:14 PM
Update 6/12/12: Media Matters comments on the latest misrepresentations of the Esper et al study discussed in our article: ‘Surprise: Fox News Fails Paleoclimatology’
Update 6/13/12: Further comment from Bob Ward of the Grantham Institute in Huffington Post UK “The World’s Most Visited Newspaper Website Continues to Regurgitate Nonsense from Climate Change ‘Sceptics’”
Just been on a 2 week holiday and although I followed some of the discussion on my phone, I was not able to make any comments. As Jan Esper and other co-authors have been silent, I feel it would help if I clarified a few issues:
1. It has not been a” tough few months for tree-rings” at all. This comment is likely related to Mike Mann’s NG paper earlier this year. A response is in review and I am sure more will be discussed on this topic over the coming weeks/months.
[Response: Yes–I’m looking forward to this, but no the comment wasn’t referring to that study so much as it was mainly referring to your paper and comments both in it and in press release issued about it that call into question the reliability of tree-ring widths (TRW) for all long-term reconstructions. Indeed, if it is true that the multi-century/millennial trends in TRW data are more generally not reliable, then it has deep implications for long-term drought reconstructions from TRW as well, and the discussion that was included in the RealClimate piece (in part, for this reason) about the recent Steinman et al PNAS article. The apparent discrepancies between the lake-based precipitation reconstructions presented in that study and tree-ring drought reconstructions becomes even more interesting and relevant in this context as, in part, the two agree well on decadal and multidecadal timescales and it is only the long-term trend which is fundamentally in disagreement (of course there are other factors such as seasonality that come into it). As Gavin has noted (and our piece makes quite clear) the claim in your paper that this finding has implications for long-term annual hemispheric temperature reconstructions however appears without support. You can expect some discussion of this in the peer-reviewed literature too. In any case, thanks for dropping by! -mike]
2. Many commentators have been getting their apples and oranges mixed up between regional and large scale reconstructions. The N-SCAN reconstruction represents a new regional reconstruction of JJA temperatures for the northern Scandinavian region and comparison to larger scale composites and assertions of global climate changes need to be made with extreme caution. The basic observation is this: MXD data, measured from tree samples from northern Scandinavia, when appropriately processed (using Regional Curve Standardisation) to capture trends longer than the mean length of the samples, portray a long term decline in values which agrees with the expected orbitally forced trend for this location which is also seen in other long-term non-tree-ring proxies for high latitudes (Fig S1). The RW data when similarly processed do not show this trend. This has potentially massive implications, as Jim rightly realises, for larger scale hemispheric proxy temperature composites which utilise high latitude RW series. So that is the hypothesis – simply put – if this MXD vs. RW bias exists for all high latitude regions, then all larger scale composite reconstructions which have utilised high latitude RW data may underestimate temperatures during earlier periods. As we clearly state in the paper, quantification of this potential bias is not really possible as there are (1) few high latitude well replicated tree-ring records which have both RW and MXD measured from the same samples; (2) different large scale composite series are calibrated against different temperature targets (full NH, extra-topical NH temps, summer vs. annual etc etc) and (3) large scale composite series normally utilise a mix of high vs. lower latitude proxies with different seasonal interpretation (summer vs. annual) – the weighting of the individual input proxy series is often unclear. Therefore, looking at long term linear trends in most large scale composite series is not really very useful and does not help to assess our hypothesis at all. We specifically only compared our series to the Kaufman et al. (2009) study for this reason (Figure 3). If the bias exists using high latitude RW data, it WILL have implications for attribution studies as a few tenths of a degree (which does not sound much) will have an effect over the Medieval period for example. Reference to the Hegerl (2006/07) studies is not really relevant as their attribution analysis does not even include the Medieval period. The fact of that matter is that prior to ~1250, large scale reconstructions and model output do not agree well at all (see AR4). Even Mike Mann only showed the post 1200 period in his recent NG article.
[Response: Rob, thanks for coming by. The difference between TRW and MXD is very interesting and it will be interesting to see how this plays out in the tree-ring reconstructions. However, for the large scale reconstructions, it is not at all obvious that the result automatically means that the very lowest frequencies are underestimated. As mentioned above, Moberg et al only uses tree rings for the high frequency component, and more recent reconstructions use a mix of sources. The orbital trends are expected to be most important in the northern high latitudes and in the summer, and any differently targeted reconstruction will have less of an effect. That they all have negative trends over this period – that are frequently larger than the modelled trends you showed – indicates that it might not be as severe a problem. – gavin]
3. The Divergence issue is not as big a problem as may be perceived. It is very much a bandwagon which many jump on to and there are different issues at both high and low frequencies and for different regions. Certainly in Europe, Divergence is hardly seen. In my opinion, the region where “divergence” is greatest is central/northern Alaska and the Yukon and recent studies have shown that divergence is expressed more in RW than MXD. The large scale studies of Briffa et al. which showed post 1960 divergence are slowly being re-addressed at regional scales by multiple groups and much of the divergence noted in these earlier larger scale studies can be minimised by targeting of an appropriate season (not the fixed season used by Briffa et al for all locations), reduction of detrending biases (see papers by Melvin and Briffa), and not compositing over too large regions. More on these results and updates will come out over the coming years.
4. Lastly, I want to highlight the importance of the N-SCAN record for dendroclimatology. This record is the new Gold Standard for TR-based temperature reconstruction and all dendroclimatologists need to aspire to attaining such high sample replication. This is of course not easy. Esper et al. state that the “record was developed over three years” – that is actually only relevant for the measurement of the MXD data. In actually fact, the sub-fossil samples for the study were collected by Finnish colleagues over ~20 years – a difficult time frame to keep funding turning over to acquire such replication. RCS really only works when replication is high (> ~30 series) and periods of low replication almost always behaves “oddly” when different detrending methods are used. The Tornetraesk record is a good case in point where multiple studies over the last 20-years have come up with differing estimates of Medieval temperatures. This is hardly surprising when replication prior to ~1300 is only about 10 series.
So – let’s not get too hung up on misinterpreting what our paper says. This is a regional study which shows a decreasing trend for almost 2000 years. The Roman, Medieval and present periods are above this long term trend – these trends are not relevant for southern Europe and are certainly not relevant for the globe as a whole. This record should not be used to debunk global warming and people should not get their knickers in a twist if the medieval is warmer than present in this regional record. This is exactly what would be expected from orbital forcing of summer temperatures at this latitude.
[Response: That would all be fine and well if Esper’s press release hadn’t extrapolated wildly from that narrow specific conclusion to calling into question the IPCC consensus about large-scale temperature reconstructions, etc. As Gavin has explained above, those arguments really don’t appear to hold up when you look at the various reconstructions, the trends in them, comparisons with model simulations (taking into account the greatly reduced orbital trend in annual hemispheric temperatures), and the actual effect that removal of tree rings from that data has on them—which apparently goes in the wrong direction relative to what your paper would predict. Its a pity that you didn’t do that analysis in the paper, because it puts some of the papers comments in a curious light. For example, the paper oddly singles out Mann et al (1999) and Mann et al (2008) for citing possible impacts on large-scale temperature reconstructions, though these reconstructions show among the largest long-term cooling trend. In fact the cooling trend in Mann et al (2008) is larger than the models predict for high-latitude summers! And it is actually the Esper et al (2002) reconstruction, which is based purely on tree-rings (from the summer and largely extratropics), that shows among the smallest long-term trends of all. It sort of surprising that this wasn’t mentioned in the paper, isn’t it? Thanks again for stopping by in any case! -mike]
I’d think this might be interesting to forestry/wood industry/woody agriculture people too, now that it’s been pointed out.
Has anyone tried forcing trees (bonsai science?) in lab conditions to see what correlates with making an annual ring wider, or denser? I wonder if something like isotopic tracers would show where cells were being formed for example.
Can you do really slow positron emission tomography for example, looking at where and when for biological activity — cell division? cell size? — as a tree ring is formed over a year?
Structurally in a tree trunk, what’s the difference between a trunk with wider or narrower rings, or denser or less dense rings? Stiffer or more flexible/stretchable in windy conditions for example?
Once again thanks for your amazing correspondence on all aspects of ACC/AGW. This article is timely and some of the answers from you guys in the thread makes me realise for many years now that if it was not for you lot here and a few other good websites then the medias approach to senstaionalism and hype would have won out over me and I would be questioning evrything wrongly.