Monday
Jul212014
by 
Bishop Hill
Bishop Hill New paper from the Netherlands on C02 emissions
Jul 21, 2014 Climate: carbon budget 'A paper published today in Atmospheric Chemistry and Physics finds that only about 3.75% [15 ppm] of the CO2 in the lower atmosphere is man-made from the burning of fossil fuels, and thus, the vast remainder of the 400 ppm atmospheric CO2 is from land-use changes and natural sources such as ocean outgassing and plant respiration.'
Read all about it at Climate Depot and H/T to Swiss Bob in the BBC and Lawson comments.
Reader Comments (61)
I won't claim that I have read the paper in detail, but from a quick look, I don't think it is claiming what The Hockey Schtick somewhat breathlessly says it is. The quote highlighted there is "We find that the average gradients of fossil fuel CO2 in the lower 1200 meters of the atmosphere are close to 15 ppm". This does not mean that only 15 ppm of atmospheric carbon dioxide is of anthropogenic origin. I think this is saying that the difference in anthropogenic CO2 concentration between the ground and an altitude of 1200 m is of the order of 15 ppm (a 'gradient' would make more sense if it was in ppm per m or km, but I think I'm understanding correctly). At ground level near power plants or roads, CO2 concentration gets very high, well above 400 ppm, and it then mixes into the atmosphere. A change in concentration of 15 ppm between ground level and 1200 m altitude is nothing unexpected or inconsistent with a largely anthropogenic origin for the increased carbon dioxide concentration in the atmosphere since pre-industrial times.
The scientists behind this new paper must be mistaken. We have been told by the senior scientists in the IPCC and the Met Office that the science is settled. All the recent floods, storms, droughts, heat waves, cold spells, and polar sea-ice melting (and growing) are at least in part due to increased anthropogenic CO2 emissions. Indeed, Peter Stott has stated categorically that there is now a direct linear relationship between atmospheric CO2 emissions and global average temperatures. The consensus is quite clear on this matter, and no alternative viewpoints should be discussed, the BBC have said so.
Going with JH, here.
==========
The oceans are the largest producer of CO2, according to NOAA, volcanoes also produce more CO2 than fossol fuel burning, again from NOAA. A point missed by many is that volcanogenic CO2, from above sea level volcanoes, is injected into the mid to high troposphere whilst our production is at sea level and readily used by plants. ie. our CO2 might not spend any time in the atmosphere before use by photosynthesis. Oceanogenic CO2 is the excess over that used by plankton etc. so is outgassed as the water warms. Another point is that volcanogenic and arthropogenic CO2 are identical isotropically so how do you tell the difference?
This paper might not cover the whole story.
I think what the paper says is that in the region they have been modeling (Western Europe) fossil fuel CO2 levels can be 15 ppm higher than the background CO2 levels - which already include historic fossil fuel CO2.
The metric they use is % of Carbon 14 in CO2. Fossil fuels do not contain 14C. Background CO2 is a particular % of 14C CO2 and in densely populated/productive areas % of 14C CO2 drops because a greater proportion of the CO2 is from fossil fuels.
I basically agree with Jeremy's interpretation here, but it's long detailed technical stuff and would need a careful reading to be sure.
yeah, I they're just simulating local mixing.
Well this is rather embarrassing for Hockey Schtick which is always a dubious source of anything scientific. As the first commenter, Jeremey Harvey says, this paper does NOT say that only 15% of atmospheric CO2 is from human sources. Sorry folks!
The paper is simply describing a computer model whichis designed to produce ‘expected’ CO2 gradients across Europe.
JH's interpretation seems more credible.
If "...only about 3.75% [15 ppm] of the CO2 in the lower atmosphere is man-made from the burning of fossil fuels," then the word 'gradient' would be superfluous.
Best delete this post or someone will use it to ridicule you on TV.
Marginally off-yopic, but I don't know where else to post it:
Danny Fortson in The Sunday Times Business section reports on a statement by Gerard Mestrallet, head of GDF Suez (Europe's biggest independent power generator).
M. Mestrallet warns that, due to 'failed' green energy policies, Europe faces blackouts. This is due to tens of millions of pounds/euros being put into intermittent renewable sources, while taxing round-the-clock fossil fuel stations. No new gas-fired power stations are being built, for instance.
To us, of course, this is a Basil Fawlty 'statement of the bleedin' obvious', but at least it seems to be couched in terms which even a politician might be able to understand.
I exclude Mr Potato Ed of course...
That'll be 'off-topic', then - please excuse fat fingers...
Leave the post up. It shows the honesty and scientific nouse of some of the readers here.
More determined fumbling in the dark by all concerned. Pursuing the mighty snipe, through dense jungle. SNAFU, in military-speak. The paper is meaningless, as CO2 level has nothing to do with the global mean surface temperature anyway. You're all fired, from the President on down.
You have a point, there, HDH; while CO2 levels continue to rise, temperatures do not. Surely, it has to be obvious that perhaps the two are not as inexorably linked as we have been told. However, for the sake of political expediency, the charade has to be maintained, as to admit it would be to consign so many careers to the scrap heap. Of course, there are also the many, many dumb folks who will never see that link in a month of Sundays… ho-hum.
Oh… and keep this post up – let it be shown that scepticism is applied to everybody!
It's gotten really frazzled since I last saw it.
In synopsis,
Evidently, the rather puny amount of CO2 which mankind adds to the atmosphere is neither here nor there, CO2 emissions are rising but so what? The link is made, between a natural background warming and CO2 emissions increase, one - CO2 follows the other as a result of warming.
Lets not put the chariot before the horse.
Funny enough athel when that was mentioned over at Paul Hudsons blog it was like something completely new to the religious faithful!
Mailman
Oh do keep up, Radical Rodent - "while CO2 levels continue to rise, temperatures do not. Surely, it has to be obvious that perhaps the two are not as inexorably linked as we have been told".
It's being masked by all those pesky El Nino events and normal apocalyptic service will be resumed in doublequick time.
I've been saying for ages they'll be claiming something like this sooner or later to explain the last 17 years.
So, Mickey C, it is now incontrovertible that a rise in CO2 WILL cause a rise in temperatures, and nothing can stop it – unless, of course CO2 levels rise, and temperatures don’t, which means something is stopping it, which only goes to prove that CO2 is such a dangerous gas… except when it isn’t. Errr… I think I understand.
"It's
gottenbecome really frazzled since I last saw it." There. Corrected it for you, TBYJ.(now mine needs correcting! Oh, for taking the short-cut and not reviewing!)
Eventually, as global mean temperatures decline, and atmospheric CO2 continues its stall in rate of rise in lockstep with temperatures, while human inputs continue accelerating, it will be realized that the former depends very little on the latter.
The claim that there was a linear correlation between atmospheric carbon dioxide concentration and global temperature was always a load of garbage.
Everyone knows that global temperature is static.
Even if the paper wasn't about gradients, it's already well known from isotope mixing ratios that most of the CO2 in the atmosphere is not from anthropogenic emissions. That does *not* mean, though, that the reason for the 40% increase is not anthropogenic.
It's like this. Imagine you have a pocket full of loose change amounting to about £10. Every day, you buy £2 worth of gobstoppers from the sweet shop, taking coins at random from your pocket, and every day you receive £2 in loose change by selling said gobstoppers to your friends. (While avoiding the teachers who would have your parents prosecuted for this breach of the health food regulations.) Any given coin only stays in your pocket for an average of 5 days. It has a one-in-five chance of being handed over at the sweet shop. But the amount of money in your pocket stays constant at £10.
Now suppose instead of charging £2 you charge £2.10. The teachers have been getting really cross about all the gobstoppers showing up in school and you have to charge danger money. And suppose the shopkeeper raises their prices too, having been threatened with closure by the local council, to £2.05. Now every day you pay out £2.05 in randomly selected coins and receive £2.10. After 80 days of this you find you have £14 in your pocket. You have £4 extra, from all those 10p's you've been getting. Does that mean that of the £14 in your pocket, £4 of it is made up of those 10p's?
No, because the £10p you got on the first day has a one-in-five chance (roughly) of being paid out to the shopkeeper the next day, and another one-in-five the next day, and so on. So each coin only hangs around for an average of five days, and the chances of it lasting 80 days are pretty much nil. And similarly with most of the 10p's that came after. Almost all of them will have entered your pocket and then swiftly left it. The only ones still in your pocket at the end are the ones you received in just the last 5-10 days, perhaps 50p's worth, which are a very small fraction of the £14. If we had marked all the 10p's somehow, so we could track where they ended up, most of them would be with the shopkeeper.
Does that mean that the extra £4 isn't there because of the extra 10p's accumulating each day? Again, no. It's the net imbalance that cause the rise or fall, and the change in price has contributed a large part of that.
If we had traced the extra coins and found the expected number of 10p's remaining in your pocket, could we deduce from that that the 10p's were solely responsible for the rise? No. Finding the 10p's there would show that 10p's were being contributed, but the net imbalance depends on more than just the danger money. Maybe the shopkeeper knew the deal and charged an extra 10p danger money themselves, but their supplier had coincidentally dropped the wholesale price by 5p. Had the wholesaler not dropped their price, the net would have been £2.10 minus £2.10 or zero, and no change would have resulted. You would still have £10 in your pocket. Even though you're charging 10p extra, and even though the expected number of 10p's are showing up in your pocket, it would have made no difference. Had you not charged them, the pocket would have still filled up. The rise is entirely because of the unrelated drop in wholesale price.
So you can't just look at the amount extra going in, the amount accumulating, and the number of marked coins in your pocket, and figure out why the rise is happening. You have to understand the reason for the price changes - you have to understand the dynamics.
Similarly, you can't tell what's causing the rise in CO2 just by doing an isotopic analysis, measuring how much of the CO2 in the air at any one time was once fossil fuel. Most of it isn't, and what there is would be there anyway. The only way you can know is to understand what controls all the other flows, and how they are affected by overall level.
The long-term rise is very likely anthropogenic in my opinion, on the basis that nobody has offered any solid evidence for another mechanism (and Occam's Razor), but proving it is more difficult than the standard story would suggest.
"Similarly, you can't tell what's causing the rise in CO2 just by doing an
isotopic analysis, measuring how much of the CO2 in the air at any one time
was once fossil fuel."
I would assume burning fossil fuels would assume probably increase the atmospheric CO2. I would also assume the ocean/(limestone eventually) would suck it all up eventually. We must know the answer by now though as the thousands of climate scientists must have been taking measurements the last few decades to show for certain where it all goes..
Rob,
We don't know where it all goes. But my point was that even if we did know where it was all going, that wouldn't actually help unless we also knew why it did so.
Well, it seems clear that the biome is an enlarging sink, not to say its enlargening is accelerating.
===============
Speculating from inadequate data, not to say poor, such as Mauna Loa, something may be acting against the accelerating anthro release of CO2 else the Mauna Loa rise would not be so linear. At the least, some evidence that sinks are not being exhausted.
===============
Took some time to get and read the paper. JH appears right, and the thrust of this thread wrong. HockeyShtick goofed up this time, as we all do eventually. The fate of humanity.
Lesson for all on all sides of this debate: Nullius in Verbim.
Hear, here.
========
The information you have picked up is unreliable. Natural emission and uptake of CO2 are about 20-fold larger than emission of CO2 from burning fossil fuels. No one in their right mind expected the atmosphere to contain more than about 5% C14 depleted CO2 from the burning of fossil fuel.
Ice cores show that natural emission and uptake of CO2 were in near equilibrium for 15,000 years.
"Ice cores show that natural emission and uptake of CO2 were in near equilibrium for 15,000 years."
Which tells you either:
A) the ice core record is unreliable for measuring historical CO2, or
B) atmospheric CO2 is well regulated, and humankind's puny inputs have little effect
Natural systems in equilibrium generally remain that way because they are tightly regulated, and require powerful external disturbances to budge them off their set positions. When you have light regulation, your system tends to wander far and wide via the accumulation of random fluctuations, like the movement of particles suspended in a fluid undergoing Brownian Motion.
The very notion that atmospheric CO2 was well regulated for thousands of years, and has been upset by our activities, is glaringly inconsistent from a scientific perspective.
I don't agree that you should take the post down. Maybe an update acknowledging the misinterpretation of the headline would save some credibility when the warmist hoards arrive to point and laugh?
Even without burrowing into the details of this paper, the reader can get from it a sense of the difficulties of observing and tracking CO2 emissions and estimating CO2 levels and distributions on local and regional scales within the first few hundred metres above the surface. Most routine emissions of CO2 occur at the surface, be they ours or Mother Nature's, and the direct infra-red interactions of that surface with CO2 molecules take place very close to it. But oh what complexity is to be found there with heterogeneous 'everything'. The GCM operators have to make some gross simplifications to reduce this, and other processes higher up, into a single number, the so-called 'external forcing' attributed to CO2 at the top of their model atmospheres. The early attempts to do that led to such nonsense that 'flux adjustments' had to be added ad hoc to preserve enough credibility for raising more funds. This fund-raising role may have been the most important contribution these models have made. This paper points to some of the barriers to them being of much use in any other way for climate prediction.
The trouble with most of these reports on the carbon cycle is that the error margins are so much larger than the searched-for human signal. Innumeracy in action!
Bart: There is a significant difference between a system being in natural equilibrium and a system that is actively regulated. The fact that CO2 has risen to 400 ppm is proof that the system is not actively and tightly regulated - whether the rise was caused by burning fossil fuel or by natural variability in emission and/or uptake. The fact that Ice cores show that CO2 was unchanged at 265-285 ppm for 8,000 years indicates that natural variability was small (+/-10 ppm) when the climate was stable during the Holocene. The rate of uptake from the atmosphere depends on the concentration of CO2 in the atmosphere, but the rate of emission does not. 275 ppm of CO2 happens to be the concentration of CO2 that makes the uptake and emission of CO2 equal under average Holocene conditions.
"The fact that CO2 has risen to 400 ppm is proof that the system is not actively and tightly regulated..."
Not necessarily. It can just mean that the equilibrium level has shifted. You acknowledge that the equilibrium level is not necessarily constant when you cite "the concentration of CO2 that makes the uptake and emission of CO2 equal under average Holocene conditions." Obviously, during different eras, the equilibrium level was different.
Or, the ice core proxy measurements are wonky. As there is no way to validate them, assuming they are true is a leap of faith.
"The fact that Ice cores show that CO2 was unchanged at 265-285 ppm for 8,000 years indicates that natural variability was small (+/-10 ppm) when the climate was stable during the Holocene."
A lightly regulated system is essentially an integrator. And, an integrator accumulates random inputs into a random walk, with dispersion increasing as the square root of time. No natural system is immune to random fluctuations. If CO2 remained in a tight band for such a stretch, then it must be tightly regulated.
" The rate of uptake from the atmosphere depends on the concentration of CO2 in the atmosphere, but the rate of emission does not."
But, the rate of uptake thereby depends on the rate of emission. It is a dynamic system with feedback. Simple accounting does not work to analyze such a system because the offsets in the ledger are always changing.
I would refer you to my post above at Jul 21, 2014 at 6:41 PM. It is very clear that the rate of change of concentration is an affine function of temperatures. This is all you need to project atmospheric CO2 concentration for at least the last 56 years, since accurate measurements became available. You do not need to know the rate of emissions - they are essentially superfluous. If you try to add the rate of emissions in with the temperature relationship, the slope of the overall trend line will be too large. If you scale and offset the temperature relationship to make room for the emissions, the temperature relationship no longer matches the local variability.
That tells us that the emissions are rapidly sequestered, and the atmospheric concentration is essentially governed by other processes which have shifted, or are shifting, perhaps as a time lagged response to a localized shifting event, the equilibrium condition in the modern age. It is becoming clear that the apparent rough affine match between emissions and concentration which appears over the past half century was merely superficial, and the two are diverging as temperatures stall. With the rate of change of atmospheric concentration stalling, while emissions accelerate, it will not be terribly long before the discrepancy becomes too large to continue denying.
Bart wrote: "But, the rate of uptake thereby depends on the rate of emission. It is a dynamic system with feedback. Simple accounting does not work to analyze such a system because the offsets in the ledger are always changing.
The uptake systems have no idea of how much emission is currently occurring - the uptake systems only know how much CO2 is in the atmosphere at the time uptake is occurring. It doesn't "know" how to balance emission.
Bart wrote: "A lightly regulated system is essentially an integrator. And, an integrator accumulates random inputs into a random walk, with dispersion increasing as the square root of time. No natural system is immune to random fluctuations. If CO2 remained in a tight band for such a stretch, then it must be tightly regulated."
But Frank had already written: "The fact that CO2 has risen to 400 ppm is proof that the system is not actively and tightly regulated..."
and Bart replied: Not necessarily. It can just mean that the equilibrium level has shifted."
Yes. The equilibrium level has shifted upward. Equilibrium (as opposed to regulation) means that the rate of emission happens to be equal to the rate of uptake. A shift in equilibrium means either emission rose or uptake declined. However, after 8,000 years when natural emission didn't rise or natural uptake didn't decline at a significant rate (compared with the 20th century rate of change), it isn't likely that natural emission or uptake suddenly changed (during normal Holocene climate present). There is no doubt that anthropogenic CO2 emissions rose this century and that they totaled enough to raise CO2 above 500 ppm.
Bart: "Or, the ice core proxy measurements are wonky. As there is no way to validate them, assuming they are true is a leap of faith."
Oh please! Air surrounds fallen snow crystals and diffuses in and out while more snow is falling. Eventually the snow is compressed into ice, sealing the air inside the ice. We know sealing happens because we don't find current CO2 levels in ice taken collected hundreds of meters below the surface. So air is permanently trapped in the firn at some point in this process. Does air diffuse freely through the deepening snow for 25 years before getting trapped? Or 50 years? Or 200 years? Given an 8000 year record of stability, it doesn't make any difference. The time needed to seal the firn does make a difference for many other things. It certainly makes a difference when we try to align the Mauna Loa record of CO2 with the ice core record. However, if natural uptake and emission varied enough to raise or lower CO2 by 50 ppm any time during the Holocene, the ice core record would have told us.
The link you provided showed CO2 data from the last 50 years (from Law Dome?). Firn doesn't form ice that fast, so recent data comes from the atmosphere and has been merged with the ice core record. Read the last sentence of:
http://cdiac.ornl.gov/trends/co2/law_dome_methods.html
"The uptake systems have no idea of how much emission is currently occurring - the uptake systems only know how much CO2 is in the atmosphere at the time uptake is occurring."
And, the amount in the atmosphere depends in part on how much is emitted. If the regulation is tight, there is a high feedback gain, so that even a small amount of emissions triggers a rapid response to soaking it up.
"...it isn't likely that natural emission or uptake suddenly changed..."
Says who?
"There is no doubt that anthropogenic CO2 emissions rose this century and that they totaled enough to raise CO2 above 500 ppm."
The virtual accumulation is meaningless for an active feedback regulated system. The evidence I have provided indicates that it was rapidly sequestered, and had little effect on overall atmospheric concentration.
" Air surrounds fallen snow..."
This is just a narrative. It means little without independent confirmation. This is the heart of the scientific method: form an hypothesis, and then test and validate. If you miss the test and validation part, then you are not performing science.
"The link you provided showed CO2 data from the last 50 years (from Law Dome?)."
No, from the direct measurements carried out at the Mauna Loa Observatory. These are the most reliable, verifiable records we have. Look at the plot again. It's all there. An astoundingly good match, given the quality of the data and the bulk averages involved. It explains essentially everything. No anthropogenic emissions required.
Bart asks "Says who [it isn't likely that natural emission or uptake suddenly changed]".
Says the stability in CO2 over the last 8000 years. When a system is at equilibrium, either: a) neither uptake nor emission vary substantially (my assumption) or b) uptake and emission always shift in parallel, in which case they are tightly linked by some mechanism. You agreed above that the recent rise ruled out a tight linkage.
Sorry for confusion about the source of the CO2 data. There is no doubt that temperature has some influence on fluctuations in the rate at which CO2 is climbing. The 1998 El Nino did result in a slight larger increase in CO2 that year. The question is temperature or man was responsible for the bulk of the CO2 increase in 1998. The question is whether man or temperature is responsible for the bulk of the increasing rate of rise - the overall slope from lower left to upper right. You could to add to this plot the anthropogenic annual release of CO2 every year which also slopes from lower left to upper right and then do a multivariable regression analysis showing how much of the change in CO2 can be attributed to man and how much to temperature. Unfortunately, temperature and CO2 have risen in parallel, and multivariable regressions don't work very well when two potential explanatory variables move in parallel. Either man or temperature to could explain gradually increasing rate at which CO2 is accumulating. Correlation is not causality. (See Al Gore.)
If you look at the ice core record, you will see that four centuries of the LIA drove down CO2 by only 10 ppm. This is strong evidence that even relatively large changes in temperature (greater than we have experience since 1960) don't produce large SUSTAINED changes in CO2.
So, the ice core record is correct because there can't have been a change in natural emission or uptake because the ice core record shows that they are stable.
Do you see the logical flaw here?
What you say about multivariable regression is correct. But, the variability allows us to separate the the effects. Look at the plot again. It is scaled to match the variability. Compare the emissions here. They have no comparable short term variability.
When we scale the temperature to match the short term variability in the rate of change of CO2, we also match the trend in CO2. But, the emissions also have a trend. There is little room for them to add in to any significance.
Look again at the comparison between atmospheric concentration and emissions. For the past decade plus, emissions have continued ramping up, while concentration has flattened out, in lockstep with the temperatures.
There really is no doubt about it. Temperatures are in the driving seat. Emissions are, at most, a small contributor. If you continue to disagree, well, you are among the many who do so. Keep watching, though. Global temperatures should be set for a 30-ish year gradual decline, in step with their behavior over the past century, and the rate of change of atmospheric CO2 will simultaneously decline. Assuming emissions keep accelerating, the discrepancy will rapidly become more and more visible.
PS: if you are interested in my projection of what temperatures are going to do for the near future, and concomitantly the rate of change of atmospheric CO2, it is here.
Bart wrote: "There really is no doubt about it. Temperatures are in the driving seat."
There is no doubt that temperature influences the rate of CO2 accumulation. During the 6/97-6/98 El Nino, CO2 rose an usually large seasonally-adjusted 3.34 ppm. However, during the 6/98-6/99 La Nina - one of the largest temperature drops on record, CO2 STILL rose 1.41 ppm. Temperature modulates the CO2 rise - it doesn't control it. Look at your plot at WoodForTrees: How many times has CO2 fallen? (I presume this is monthly data.) Not once since 1980! How can temperature control CO2 when temperature has fallen many times during this period and CO2 has never fallen? (Answer: CO2 is being driven upwards by something in addition to temperature - that dominates temperature - ensuring that CO2 never falls.) What has happened during the pause? Temperature hasn't gone up in more than a decade, but CO2's rise has continued unabated. Look carefully, the flat areas that align after 2000 say that CO2 rising at a constant rate and temperatures is remaining constant! When you align the derivative of a time series against an ordinary time series, flat areas near zero on the y-axis mean the same thing (no change), but flat areas well above zero mean that one is changing at a constant rate while the other is not.
Consider the following thought experiment: Let CO2 increase with t^2 to roughly mimic the growth of human emissions. Let temperature rise with the log(CO2) which is linearly with time. (log(t^2) is 2*log(t)). Now let's add noise to the temperature data. Now make a small change in CO2 proportional to the latest temperature CHANGE. Now take the derivative of the CO2 data. The derivative of the sum of these two terms contributing to CO2 rise is the sum of the derivations. The derivative of the t^2 is 2*t - a constant slope from the lower left to the upper right. The derivative of the noise is going to put a lot of scatter in that line of constant slope, and all the fluctuations will align perfectly with the noise in the temperature data. What is driving CO2 rise in this case? The t^2 term obviously will dominate eventually, with scatter in the data provided by and correlated with the noise in the temperature. You are being fooled by the correlations in the plot at WoodForTrees. Remember Al Gore's mistake with ice core data: Correlation is not causation!!!!
Thanks for the interesting discussion
"How can temperature control CO2 when temperature has fallen many times during this period and CO2 has never fallen?"
It is an integral relationship. Since at least 1958, it has unfolded essentially according to the differential equation
dCO2/dt = k*(T - To)
where k is a scale factor, and To an equilibrium level. Atmospheric CO2 doesn't start decreasing until T goes below To.
To preempt your next objection, yes, there is no upper bound on the result if temperature stays above To. But, this is merely an approximate relationship which is seen empirically to hold reasonably well for the past 56 years. It is a local, in time, approximation, valid for the era 1958-present. What the dynamics would be outside that era would be speculative at this time - we do not have enough information. But, it has held for 56 years, and the expectation would be that it will hold reasonably well into at least the near future.
"The derivative of the noise is going to put a lot of scatter in that line of constant slope, and all the fluctuations will align perfectly with the noise in the temperature data."
It isn't just noise. It is a trend plus short term variation in the rate of change of CO2 which matches the trend and short term variation in temperature remarkably well. Correlation is, indeed, not causation. But, this correlation is strong between two well separated (in frequency) time series components. The likelihood of that happening by chance is vanishingly small.
If you are interested in a discussion of how I think the differential equation relationship above probably comes about, and how it can be consistent with rapid sequestration of human inputs, the proprietor of the blog here picked up my surmise in a comment at WUWT and posted it.
Also, you may find Dr. Murry Salby's related lecture here of interest.
Bart: Both your mechanism and mine have a d(CO2)/dt = f(T) to explain the correlation, but mine has additional terms. Your dCO2/dt = k*(T - To) mechanism and mine may be able to explain the CO2 data since 1960, but yours can't explain many other phenomena. Ice cores show only a 10 ppm drop in CO2 during the LIA - four centuries when the T was presumably below T0. Furthermore, there has been a slow long-term cooling ever since the Holocene Climate Optimum that would have taken us away from any T0. During the last ice age, T was massively below T0 for 100,000 years, but CO2 dropped and then stabilized. (If you believe that CO2 functions as a greenhouse gas, your formula has the potential to produce a runaway greenhouse or icehouse because it is inherently unstable.) If there are two mechanisms that can explain the correlation between between d(CO2) and T since 1960 and one mechanism can't explain what happened earlier, which one is likely to be right?
Worst of all, you can't ignore the fact that there is an anthropogenic d(CO2) term from burning fossil fuels every year which is about twice the observed rise in CO2. Today that emission is equivalent to a 4 ppm/yr rise. To explain the observed average annual rise of only a 2 ppm/yr, nature must be taking up half of this anthropogenic emission by an excess of uptake over emission. Once a CO2 molecule from fossil fuel is in the atmosphere, natural uptake won't treat it any differently than a CO2 molecule emitted from some natural source. Any sensible explanation for changes in atmospheric CO2 MUST have at least two terms: d(CO2)/dt = f(T) + aCO2, where aCO2 is anthropogenic emission and f(T) provides the variation that is seen with temperature that has been observed over 50 years at Mauna Loa AND in ice cores for millennia. (A thousand millennia if you believe scientists have correctly handled the problem of layers flowing and thinning under the weight of all the ice above, but my points rely on the much simpler information for the Holocene and some of the LIA). The ice cores tell you that the f(T) term can't be k(T-T0) because of the long periods of stability when T-T0 must have been as large as it is today.
You raise many points which I would like to address, but I find long-winded replies can be counterproductive, as the reader loses interest. So, I will focus on just a couple.
"The ice cores tell you that the f(T) term can't be k(T-T0) because of the long periods of stability when T-T0 must have been as large as it is today."
In the modern era, the f(T) term is undeniably k(T-T0). You can see it right here.
You raise an issue with the ice cores, but again, the ice core estimates cannot be validated.
In addition, the system is assuredly nonlinear overall, and the k and T0 which hold now provide a mere linearized approximation to what is truly going on in the large. But, they hold in the modern era 1958-present, and this is the era in which the lion's share of the observable rise in the past century has been observed. Extrapolating this model well beyond the period of observation is no more likely to produce a reliable estimate of behavior than extrapolating the gain of a transistor beyond its nominal operating range.
"If there are two mechanisms that can explain the correlation between between d(CO2) and T since 1960 and one mechanism can't explain what happened earlier, which one is likely to be right? "
But, the anthropogenic forcing mechanism cannot explain the correlation. The short term variations in the rate of change of atmospheric CO2 do not match the anthropogenic forcing. They only match the temperature variations. But, the temperature also has a long term trend. You cannot take in the short term variation, and leave out the long term trend.
Nature has no way of performing that high pass filtering operation in general, and certainly not with no observable phase distortion. You have to take all the components of the temperature input to CO2. When you do that, you have enough trend to explain the trend in the rate of change of CO2. You cannot add in an additional significant amount of anthropogenic forcing - the trend becomes too large.
Frank said: "If there are two mechanisms that can explain the correlation between between d(CO2) and T since 1960 and one mechanism can't explain what happened earlier, which one is likely to be right? "
Bart replied: "But, the anthropogenic forcing mechanism cannot explain the correlation. The short term variations in the rate of change of atmospheric CO2 do not match the anthropogenic forcing. They only match the temperature variations."
Frank continues: Have you honestly tried to see if the anthropogenic mechanism is incapable of producing a d(CO2) vs T plot that looks like observation??? I proposed a thought experiment that should have alerted you to the possibility that an d(CO2) = aCO2 + f(T) model could work about well as your k(T-T0). Remember, scientific theories are never proven to be true try by experiments (especially one experiment); we use observations and experiments to reject hypotheses that are inconsistent with data. Theories are what survive. I suspect most people with a scientific viewpoint reject your k(T-T0) mechanism because it doesn't take into account what happens to aCO2 (which certainly exists) and the ice core record. However, your plot clearly shows that the real mechanism must include some sort of f(T) term. Everyone should have already realized that an f(T) term has to exist, because an f(T) term is needed to explain the ice core data (low CO2 during glacials, high CO2 during interglacials). So you haven't discovered anything unprecedented - just that the alarmists have chosen to ignore the role of f(T) in the recent record. So your challenge is to demonstrate that a d(CO2) = aCO2 + f(T) mechanism(s) can not produce a d(CO2) vs T plot like the one we observe and thereby discredit such mechanisms. If you try to do so, I think you will recognize why having a modest amount of emission linked to temperature can explain observations even though the bulk of the change is being driven by aCO2. We've got aCO2 data (but not monthly) and T need to try.
I can futz around, and come up with a small band of allowable impact from emissions. However, A) the allowable band still does not allow humans to dominate the process, and any changes we could conceivably make to our inputs have negligible impact, and B) amoung all the possible system configurations, ones in which emissions are effectively nulled in their impact occupy an overwhelmingly larger probability space than those in which there happens to be a fine balance between emissions and natural processes. Occam's Razor comes down firmly on the side of assuming that emissions have utterly negligible impact.
The fact that the rate of change of atmospheric CO2, for the past decade plus, is stagnant, whilst emissions are raging ever upward, gives additional support.
I hope I have at least opened your, and possibly other, minds to the possibility that it is not inconceivable that humans have negligible impact on atmospheric CO2 levels globally. All I think we can do at this point is wait and see what happens. As I have mentioned, keep track of the accumulated emissions and the atmospheric concentration. If what I say is true, you should see increasingly rapid divergence between the affinely mapped comparison of the two as temperatures decline.
I expect then we will be treated to the climate establishment scratching their heads, and wondering why sink activity seems to have picked up, as more and more of the human emissions appear to be getting sequestered. When they do, you can tell them it is just continuing as it has since at least 1958, and there is nothing new or remarkable going on.
Oh yeah, keep this thread. Bart and Frank, thank you very much.
===================