While this article makes a strong case, looking at SST and CO2 can also be revealing:
A review of this WUWT post might also be instructive: A look at human CO2 emissions -vs- ocean absorption
From Columbia University: Oceans’ Uptake of Manmade Carbon May be Slowing
First Year-by-Year Study, 1765-2008, Shows Proportion Declining
(Click on image to view larger version)
Carbon released by fossil fuel burning (black) continues to accumulate in the air (red), oceans (blue), and land (green). The oceans take up roughly a quarter of manmade CO2, but evidence suggests they are now taking up a smaller proportion.
Credit: Samar Khatiwala, Lamont-Doherty Earth Observatory.
The oceans play a key role in regulating climate, absorbing more than a quarter of the carbon dioxide that humans put into the air. Now, the first year-by-year accounting of this mechanism during the industrial era suggests the oceans are struggling to keep up with rising emissions—a finding with potentially wide implications for future climate. The study appears in this week’s issue of the journal Nature, and is expanded upon in a separate website.
The researchers estimate that the oceans last year took up a record 2.3 billion tons of CO₂ produced from burning of fossil fuels. But with overall emissions growing rapidly, the proportion of fossil-fuel emissions absorbed by the oceans since 2000 may have declined by as much as 10%.
Some climate models have already predicted such a slowdown in the oceans’ ability to soak up excess carbon from the atmosphere, but this is the first time scientists have actually measured it. Models attribute the change to depletion of ozone in the stratosphere and global warming-induced shifts in winds and ocean circulation. But the new study suggests the slowdown is due to natural chemical and physical limits on the oceans’ ability to absorb carbon—an idea that is now the subject of widespread research by other scientists.
“The more carbon dioxide you put in, the more acidic the ocean becomes, reducing its ability to hold CO₂” said the study’s lead author, Samar Khatiwala, an oceanographer at Columbia University’s Lamont-Doherty Earth Observatory. “Because of this chemical effect, over time, the ocean is expected to become a less efficient sink of manmade carbon. The surprise is that we may already be seeing evidence for this, perhaps compounded by the ocean’s slow circulation in the face of accelerating emissions.”
The study reconstructs the accumulation of industrial carbon in the oceans year by year, from 1765 to 2008. Khatiwala and his colleagues found that uptake rose sharply in the 1950s, as the oceans tried to keep pace with the growth of carbon dioxide emissions worldwide. Emissions continued to grow, and by 2000, reached such a pitch that the oceans have since absorbed a declining overall percentage, even though they absorb more each year in absolute tonnage. Today, the oceans hold about 150 billion tons of industrial carbon, the researchers estimate–a third more than in the mid-1990s.
For decades, scientists have tried to estimate the amount of manmade carbon absorbed by the ocean by teasing out the small amount of industrial carbon—less than 1 percent—from the enormous background levels of natural carbon. Because of the difficulties of this approach, only one attempt has been made to come up with a global estimate of how much industrial carbon the oceans held—for a single year, 1994.
Khatiwala and his colleagues came up with another method. Using some of the same data as their predecessors— seawater temperatures, salinity, manmade chlorofluorocarbons and other measures—they developed a mathematical technique to work backward from the measurements to infer the concentration of industrial carbon in surface waters, and its transport to deep water through ocean circulation. This allowed them to reconstruct the uptake and distribution of industrial carbon in the oceans over time.
Their estimate of industrial carbon in the oceans in 1994—114 billion tons—nearly matched the earlier 118 billion-ton estimate, made by Chris Sabine, a marine chemist at the National Oceanic and Atmospheric Organization in a 2004 paper in the journal Science.
Sabine, who was not involved in the new study, said he saw some limitations. For one, he said, the study assumes circulation has remained steady, along with the amount of organic matter in the oceans. “That being said, I still think this is the best estimate of the time variance of anthropogenic CO₂ in the ocean available,” said Sabine. “Our previous attempts to quantify anthropogenic CO₂ using ocean data have only been able to provide single snapshots in time.”
About 40 percent of the carbon entered the oceans through the frigid waters of the Southern Ocean, around Antarctica, because carbon dioxide dissolves more readily in cold, dense seawater than in warmer waters. From there, currents transport the carbon north. “We’ve suspected for some time that the Southern Ocean plays a critical role in soaking up fossil fuel CO₂,” said Khatiwala. “But our study is the first to quantify the importance of this region with actual data.
The researchers also estimated carbon uptake on land, by taking the known amount of fossil-fuel emissions and subtracting the oceans’ uptake and the carbon left in the air. They were surprised to learn that the land may now be absorbing more than it is giving off.
They say that until the 1940s, the landscape produced excess carbon dioxide, possibly due to logging and the clearing and burning of forests for farming. Deforestation and other land-use changes continue at a rapid pace today—but now, each year the land appears to be absorbing 1.1 billion tons more carbon than it is giving off.
One possible reason for the reversal, say the researchers, is that now, some of the extra atmospheric carbon—raw material for photosynthesis–may be feeding back into living plants and making them grow faster. “The extra carbon dioxide in the atmosphere may be providing a fertilizing effect,” said study coauthor Timothy Hall, a senior scientist at NASA’s Goddard Institute for Space Studies. Many other scientists are now working to determine the possible effects of increased carbon dioxide on plant growth, and incorporate these into models of past and future climates.
Khatiwala says there are still large uncertainties, but in any case, natural mechanisms cannot be depended upon to mitigate increasing human-produced emissions. “What our ocean study and other recent land studies suggest is that we cannot count on these sinks operating in the future as they have in the past, and keep on subsidizing our ever-growing appetite for fossil fuels,” he said.
In a related paper in Nature, Khatiwala describes how the research was done.
Just a few comments from a physical chemist regarding Henry’s Law. This law says that the CO2 in the atmosphere will seek an equilibrium with the dissolved CO2 in the water (as carbonic acid – H2CO3). The exact proportions will depend on the temperature of the water, with warmer water less able to hold CO2 in solution. The mathematical statement is that at equilibrium the ratio of the partial pressure of CO2 to the concentration of H2CO3 is a temperature-dependent constant:
K(T) = P[CO2]/[H2CO3]
(see Wiki’s section on Henry’s_law#Temperature_dependence_of_the_Henry_constant)
This simple truth is complicated by the fact that there are two other equilibria involved, both of which depend on the pH of the water, rather than the temperature.
H2CO3 ionizes in water to produce a bicarbonate ion (HCO3-) and a proton (H+, acidity).
Bicarbonate ion ionizes further in water to procude a carbonate ion (CO3=) and a proton (acidity, again).
The pH of sea water ranges froma high of about 8.4 to a low of about 7.5.
At a pH of 8.4, 98% of the dissolved carbon dioxide is present as bicarbonate ion, and only 1 percent of the dissolved CO2 is present as carbonic acid, and another 1% as carbonate. That mean that 99% of the CO2 that the ocean absorbs becomes ions that cannot be released back into the air. Only carbonic acid can release CO2 back into the air.
The airborne CO2 is in equilibrium with only 1% of the CO2 in the sea. The sea is a terrific sink for CO2.
Even at a pH of 7.5, 90% of the CO2 in sea water is present as bicarbonate, and only about 10% as carbonic acid. [Carbonate holds only about 0.1% of the dissolved CO2 at this pH.] This relatively high concentration of carbonic acid means that the atmospheric CO2 reaches equilibrium with the dissolved CO2 faster in the more acidic portions of the sea.
Carbonic acid only predominates as the dissolved form of CO2 when the pH is less than 6.37 – a situation we have not yet seen. Above this pH the sea absorbs more CO2 that it releases.
CO2 can be driven out of solution by raising the temperature or by lowering the pH. Since the pH is not low enough to make an isothermal sea a net source of CO2, the way to take CO2 out of the sea is by raising the temperature.
Can anybody tell us if any of the IPCC models account for submarine heat sources? These heat sources are not trivial. I am talking about “black smokers”, mid-ocean rifting, and ocean floor vulcanism (e.g. Gakkel Ridge, Arcitc Ocean; Lo`ihi volcano, Hawaii:). I don’t believe that they do.
Steve Fitzpatrick (20:22:16) :
Thank your patience in doing the detailed explanation. I’m trying to get my head around the somewhat counterintuitive result that 2.6ppm by weight would equate to 1.7 ppmv even though CO2 is denser than air
Another thing that concerns me is the behaviour of dissolved gases in that emormous column of oceanic water where pressures get ridiculously high and the temperature drop must double the absorbtivity of CO2. It seems that even smaller changes in temperature would result in considerable outgassing, and the previous post talks about those hydrothermal vents and other recently discovered ocean-floor heat sources. But, how CO2 could migrate up from these depths without needing an upward water current?
And given the difficulty of measuring such tiny temperature differences in situ, and
also the problem of actually measuring the CO2 content of samples from these depths, there seems to be plenty more science to be done, and as a taxpayer I’d rather contribute to more research than be stuck with a huge international tax-hike.
(And I’d prefer the research done by chemists and physicists . Seems you can’t trust climatologists these days.)
Bart (19:02:57) :
The data say that for every unit of CO2 emitted by humans, about half accumulates in the atmosphere, and half accumulates in the oceans/biosphere/land. This distribution seems rather surprising to you, though I don’t follow why. You also seem to think that there are other ways of explaining the accumulation in the atmosphere. So, out with it. Put forth any sort of hypothesis. Where is the other source? Is it the ocean? Is it the soil? Volcanoes? Then, give a rough idea of what the net flow from that source would need to be, in order to account for the atmospheric accumulation, as well as drown out the effect of the human-related emissions. Describe the other flows around the system, as well. Then explain why this source would have become active just at the same time as human-emissions, and followed the same ramp. Also explain how this other source has the isotope signature of fossil fuel use.
You invoking an unknown unknown, and stopping there, seems hugely unproductive, as well as strangely incurious.
“2000 to 5000 Gtons, natural processes would add over 100 Tera-tons. ”
You continually make statements like this, without noting the simultaneous natural outflow. Those natural processes were active before human emissions, too. Are you somehow amazed then, that the atmospheric CO2 level was fairly stable during the thousands of years of the current interglacial, before human emission? That’s thousands of years of natural processes, adding gigantic amounts of carbon. And yet the atmospheric level remained within a band 20 ppm wide or so.
We are taking carbon that has been out of the cycle, and putting it back in. It has to accumulate somewhere. I don’t know why it is surprising to you that it accumulates in both air and sea. Where did you think it was going to go?
“That, however, may not be the case, depending on what he meant by “Gaussian trajectory”,”
A Gaussian distribution with a half-width of 150 years. That’s unambiguously defined; I don’t see how there is confusion.
I haven’t mentioned it yet, but I have no idea what I’m looking at, in the pink plot at the top. They’re plotting monthly data, fine, using a 12 month moving average for the CO2 level, and a 21 YEAR moving average for the SST?
I’ll allow that SST data is noisy (as is all temperature data), but a 21 year average?? Why plot a data point for each month if you’ll use that much averaging? Autocorrelation, anybody? I can understand smoothing the noisy side a bit, but this seems like an extreme mismatch. If you showed me this, and I knew nothing of the topic, I would ask to see CO2, SST anomaly and global surface anomaly, using the same smoothing, plotted against time, though I already know what that looks like.
I don’t see what you can hope to learn from the pink plot. SST is going up, CO2 is going up. If you use a 21 year average on the SST side to suppress the noise, the correlation looks quite nice. But we knew that. Though I thought some here were disputing the former, or that there was much of any correlation between the two.
carrot eater (15:40:43)
“You continually make statements like this, without noting the simultaneous natural outflow.”
Are you kidding?!!! It’s all I talk about. The natural outflow does not discriminate between the natural inflow and what we put in. It takes it all back out, and it does so elastically. It opposes the deviation from the mean. This is negative feedback in action. This is what negative feedback does. It is what makes feedback so marvelously useful, and feedback regulated systems so marvelously robust.
The whole point is that the system is removing all that naturally produced CO2 just fine, but then it supposedly chokes up on a measly few percent-at-most additional input. That would make it, most decidedly, a non-robust feedback loop.
“A Gaussian distribution…”
How do you read “trajectory” and get “distribution”? I could think of a dozen ways to generate a trajectory off the top of my head which would fit such a description. There are all kinds of possibilities, which is why I said this paper would never have gotten past peer review in my field. We take pains to outline each and every step, and rigidly define each and every quantity, in our papers so that they will be useful to others who wish to replicate our results. That does not appear to be the goal at all here, and that in itself is a travesty.
The system response can be extremely sensitive to the type of function used. As I stated before, frequency components beyond the bandwidth of the system can excite transients, and the plots in the paper do appear to show large transient spikes, though it is difficult to tell from the resolution of the time axis. If it is not a transient spike, then it makes no sense at all, for reasons I have outlined previously.
There is no reason at all to arbitrarily choose the profile of a Gaussian distribution. And, it would leave plenty of questions if that is what he did, such as, how far down the tail did he go for the start date? If, e.g., he used a distribution centered at 2100 with a “standard deviation” of 150 years, and he started the model in the year 2000, then he was already way up the curve when he started, and that would introduce horrendous high frequency content to feed a transient. Even starting 200 or 300 years previous would require a huge initial step. Where did he start?
“Where did he start?”
And, if the answer is, he started at the peak in 2100 and tailed off from there, then the whole thing is just gibberish, and not reflective of reality in the least. The transients, which reflect only the response to the sudden step input and not reality, would be enormous, and that may be just what we see in the plots.
Bart (16:43:05) :
I think I’ve finally figured out what’s driving your thinking.
Basically, you are surprised that CO2 is able to accumulate to any appreciable extent in the atmosphere, because you imagine the carbon cycle as a feedback control loop that keeps the atmospheric level in check. You are drawing analogies to whatever systems you study in your work, and you are thinking that the carbon cycle will work the same way, without stopping to consider the actual physical processes and time scales involved. I think this is a rather poor approach. You should put some time into studying the ocean-atmosphere exchange and the time scales involved, as well as the other processes involved, such as the weathering of rocks (which, as one might imagine, is quite slow). Only then would you have an idea of where the carbon should go, and how quickly it should go there.
On the level that you are thinking, you could flip the picture upside-down and reach the opposite conclusion. If man were pumping CO2 directly into the ocean, shouldn’t your feedback mechanism oppose the deviation from the alleged setpoint and promptly put all the CO2 into the atmosphere? How do you know that your feedback mechanism will work to keep stasis in the air, instead of the ocean?
These are the dangers that arise when you don’t actually look at how carbon is cycled throughout the system (and ultimately removed from the system), and just imagine it as some black box feedback control loop.
Bart (16:43:05) :
“How do you read “trajectory” and get “distribution”? ”
I think it’s pretty straightforward. The trajectory of the emissions look like a gaussian curve. The center is at time x; the half-height width is y. Doesn’t seem the least bit confusing to me. Start from zero, ramp up, hit the maximum at year x, then ramp down. Write to him, if you don’t believe me.
“There is no reason at all to arbitrarily choose the profile of a Gaussian distribution. ”
Why not? He has to choose something. The goal is to see the long term response; to see that clearly, the emissions have to ramp back down again.
Is the actual history Gaussian so far? No, nor need it be for the paper to be instructive. (you can see the actual history in Fig 1 here, http://wattsupwiththat.com/2009/11/10/bombshell-from-bristol-is-the-airborne-fraction-of-anthropogenic-co2-emissions-increasing-study-says-no/)
Bart, one more – Archer and Caldeira (freshly of Superfreakonomics fame), among others, recently published a intermodel comparison of how different models handle a pulse of CO2 emissions. I haven’t had time to read it carefully, but it has a better discussion, and being more recent, it gives more recent references as well. Mainly, I think it’d be useful to actually consider the different processes at hand, so you don’t imagine the thing as some sort of black-box PID loop.
The paper actually discusses the sorts of confusion seen among the public regarding issues such as CO2 lifetime; apt as this exact confusion was seen in the thread here about the Bristol work (Knorr).
http://dge.stanford.edu/labs/caldeiralab/Ongoing_changes.html , the 2009 publication with Archer, et al
carrot eater (19:04:52) :
“You should put some time into studying the ocean-atmosphere exchange and the time scales involved…”
No, that’s not it either. I have covered every possible variation for a continuous feedback system. These are very general principles I have been putting forward. More in a minute…
“If man were pumping CO2 directly into the ocean, shouldn’t your feedback mechanism oppose the deviation from the alleged setpoint and promptly put all the CO2 into the atmosphere?”
Define “promptly”. Feedback systems evolve in time based on their bandwidth. And, they regulate inputs according to how well that bandwidth covers the input bandwidth. This is why the form of any hypothetical input has to be very carefully constructed so as not to trigger false dynamics.
“How do you know that your feedback mechanism will work to keep stasis in the air, instead of the ocean?”
It is a different system response, and they are not symmetric. The ocean is the repository at the minimum energy state. Everything in nature wants to reach its minimum energy state. This is why things have reached their present state – it is as low down the energy scale as they can go in the presence of the external forcing of the Sun. I tried to explain to you earlier how this worked, how nature gets into a rut from which it will resist being dislodged, but I don’t think you considered it very carefully.
“Start from zero, ramp up, hit the maximum at year x, then ramp down. “
How? A Gaussian function has infinite tails. You can truncate at some point, but you then have a discontinuity, and you will elicit a false transient proportional to that discontinuity. If your Gaussian function has a “halfwidth” of 150 years, you have to go back (or forward) from the central point several centuries to get an insignificant discontinuity.
“Why not? He has to choose something”
How about something which is assured of having a frequency content which will not excite transients? Put in an FIR filter impulse response which is tailored to maintain frequencies within the bandwidth of the system. At least then you can lower bound the effect without causing potentially misleading transients.
“Mainly, I think it’d be useful to actually consider the different processes at hand, so you don’t imagine the thing as some sort of black-box PID loop.”
I have not done so. I have stated that if the system is overwhelmingly linear, there are fundamental limits on response no matter the actual form of the system. I have also stated that, if there are significant nonlinearities, then the sensitivity must be significantly amplified at the margins for this to all add up, and I have further stated that, that increased sensitivity should be recognizable in increasing variability as the CO2 level rises. These are very general principles, as I said. They must hold no matter what the actual form of the system.
Note: I am interpreting “half-width” as the width at which the function reaches half its peak, in reference to how power engineers measure the half power bandwidth of a network. But, maybe this refers to a width such that the tail of the distribution is “negligible”. If that is the case, then he is dumping up to 100% of the carbon reservoir of the Earth into the atmosphere in a very short time interval, which is not realistic.
Carrot eater —
“I’ll allow that SST data is noisy (as is all temperature data), but a 21 year average?? Why plot a data point for each month if you’ll use that much averaging?”
I’ve seen that pink graph before, but couldn’t find the source with a bit of browsing. But we can guess.
You normally choose a smoothing average based on the frequency of the noise you wish to filter out. Hence the one-year for the CO2, which has an annual period. With the SSTs, they are probably trying to smooth out the ENSO/PDO variability, which oscillates every 20 years or so.
Bart —
“I am not qualified to comment on the “stripping” action, but there is a hell of a lot of surface area there. Do you have any reliable references in which this is studied?”
Given a nominal pH of 5.7 for rainfall (the usual figure), and assuming all the acidity is due to carbonic acid, then each liter of rainwater should be removing 0.0000878 g of CO2.
Thanks, Contrarian.
To Carrot Eater – let me clarify something I have really only hinted at. There could well be a transient response to our forcing, which could cause some overshoot from the steady state. But, predicting that overshoot is like predicting the weather. It is fast dynamics which are extraordinarily sensitive to model parameters and to the form of the input.
The arguments put forward by some lay people to the effect of “how can we predict the climate in 50 years when we cannot predict the weather in 2 weeks,” as you and I both presumably know, are naive. Fast dynamics, sometimes to the point of being stochastic in nature, are devilishly tricky to estimate. Slower dynamics are much easier, and almost all of our technology, indeed the very function of our brains by which we form our observations, depends on this distinction.
Contrarian (03:24:04) :
According to this site, there is globally 914 trillion liters of rainfall on the planet each day, or 334 tera-liters per year. That, given your figure (I am assuming “g” is for grams), could remove 30,000 metric tons of CO2 per year. As our emissions appear to be estimated at about 30 Gtons per year, this does not appear to be very significant.
supercritical (07:56:04) ,
The volume fraction of CO2 in the atmosphere is proportional to the number of molecules of CO2 compared to the total number of molecules of everything (mostly N2 and O2, but also a little bit of argon, CO2, N2O, methane, halocarbons, and ozone). The weight fraction is proportional to the number of molecules of CO2 multiplied by its molecular weight divided by the sum of the numbers of each type of molecule multiplied by their respective molecular weights (or atomic weight, in the case of argon). This means (in essence) that the volume fraction and weight fraction of CO2 in the air are related to each other by the ratio of the molecular weight of CO2 compared to the “average” molecular weight of everything in the air:
Fw = Fv* (44/29)
or by rearranging the terms,
Fv = Fw * (29/44)
where Fw is the weight fraction and Fv is the volume fraction. The “29” is very close to the averge molecular weight of everything in the air. If you have any further doubt, a college level introductory general chemistry or introductory physical chemistry text will probably explain better than I can. Lots of sources are on the internet as well.
With regard to pressure: Pressure in the deep ocean is (of course) astonomical (400 bar at 4 Km depth), but this does not mean that the quantity of CO2 in the deep water is changed by that pressure. If you could increase the pressure of the atmosphere, more CO2 would, of course, dissolve in the surface of the ocean. But you can’t increase the pressure of the atmosphere, so the absorption/desorption always takes place at 1 atmosphere pressure at the surface, not at the high pressure present in the deep ocean. If you collect a sample of water from the deep and reduce its pressure to 1 atmosphere it is not going to out-gas CO2 like a freshly opened beer, because the concentration of CO2 in that water is not very high; it was set by the atmospheric concentration of CO2 when the water was last in contact with the air.
With regard to deep ocean vents: There could of course be some contribution of these vents to both the temperature of the deep ocean and to the concentration of CO2 in the deep ocean (assuming that CO2 is a significant fraction of what comes out of the vents). However, please keep in mind that the actual (measured) temperature of the deep ocean is quite close to the temperature of the sinking surface water at high latitudes, and that the measured concentration of CO2 in the deep ocean is really not high; it is about what would be expected for having been in contact with the atmosphere with a CO2 concentration of about 280 PPM by volume.
Anyway, ocean absorption of CO2, while complicated in the details, seems to me a pretty straight forward physio-chemical process when considered in light of the basic structure of the ocean and the very slow “turn-over” due to surface and deep currents. There is (of course) some uncertainty about the relative importance of absorption, biological ocean up-take, and increased CO2 uptake by land plants, as well as other factors (like effects from deep ocean vents, but the overall process is pretty well understood.
Contrarian (03:24:04) :
According to a site which the spam filter appears not to like (Wiki-answers) there is globally 914 trillion liters of rainfall on the planet each day, or 334 tera-liters per year. That, given your figure (I am assuming “g” is for grams), could remove 30,000 metric tons of CO2 per year. As our emissions appear to be estimated at about 30 Gtons per year, this does not appear to be very significant.
Maybe the spam filter distrusts the site for good reason. Maybe this is just rainfall over land, or is totally wrong altogether. Does anyone have a different estimate? I can’t seem to find one in a cursory search.
OK, the spam filter let up while I was away searching. Still, is this figure accurate?
Bart (08:15:42) :
Rainfall has to almost equal ocean evaporation. Nobody knows exactly what that is, but a reasonable (very rough) estimate is in the range of 0.25 cm per day on average.
If the ocean area is about 360 million sq Km (assuming I’ve done my math right), then the total daily volume should be 0.25 * 1000000 * 10000 * 360,000 liters per day, or 9 * 10^14 liters per day, or 900 trillion liters per day. So the 914 trillion number looks reasonable. Seems unlikely to be more than 50% off.
Bart (20:19:46) :
“No, that’s not it either. I have covered every possible variation for a continuous feedback system. ”
You are talking generic principles of feedback control loops. You have spent zero time considering the physical processes involved in moving carbon from one place to another. How you think you can understand the carbon cycle without actually looking at the carbon cycle is beyond me.
“Define “promptly”.”
That’s what I was hoping YOU would do. I’ve referred you to literature that discusses the time scales of the different processes involved. You seem to want the time scales to be shorter, but have not described the physics that would allow this. You can talk about bandwidth as long as you want, but it won’t have meaning until you look at the physics of the carbon cycle.
“The ocean is the repository at the minimum energy state.”
Actually, in the end it ends up as minerals and rocks on the sea floor and land.
Let me help out here – you were trying to write a overly simple model before; let’s resurrect it for a bit. Let me work with mass, instead of concentration: N = mass of carbon in the atmosphere. Then,
dN/dt = -k*(N-No) + human emissions(t)
where No is the equilibrium amount of carbon. For now, let’s call No to be the preindustrial amount, though as it turns out, that’s a bad assumption. We’re assuming a single simple linear term; we have no physical basis for doing so (and it’s incorrect), but it makes life easy for the moment.
So, what do we know? This is where you kept erring before. Let’s take an estimate for the current value of -k*(N-No): eh, let’s say it’s 2.5 Gton C/year. Looking at Fig 7.3 of the IPCC WG1, let’s say N is currently 760 Gton C, and No is 600 Gton C. Thus, k is equal to 0.016 1/yr.
Before, you kept confusing the issue by considering the total flows, not the net flows. I don’t know why; this model doesn’t have the physics to be able to describe the individual flows.
So, if emissions stopped today, what would happen? Taking the current conditions as the initial conditions Ni, we get
(N-No)/(Ni-No) = exp(-kt)
The time constant is 63 years. As awful as this exercise was, that’s not a terribly wrong result. However, as seen in the literature, there isn’t a single simple time scale to this process, but multiple physical processes, all operating at different time scales. Add the descriptions of those processes, and you get a ‘long tail’, not a simple exponential decay, and you don’t approach the pre-industrial value for a long, long time.
Happy? Or are you still hung up on why the term [-k*(N-No)] has been roughly equal to 1/2 of the human emissions, over time, as opposed to 1/10, or 1/5, or some changing fraction? For that, you have to consider the physics of how the carbon cycle reactions to the human perturbation. There is no alternative; you cannot analogise to other systems.
Contrarian (23:06:23) :
But with that much smoothing on the SST side, you shouldn’t be surprised the correlation looks so good. All the pink plot says is that both CO2 and SST are going up. But we knew that.
Bart —
I got 30 million metric tons/yr, based on the 914 trillion liters/day. Still not much compared to emissions.
Carrot eater —
“But with that much smoothing on the SST side, you shouldn’t be surprised the correlation looks so good.”
Why not? The smoothing does not effect the slopes of those curves, and thus the correlation.
Contrarian (15:33:09) :
Does the value of the slope have any particular significance? Does 143.6 mean anything to you? Anyway, the pink plot looks clean because of the huge amount of smoothing; if you just plotted monthly data or even yearly means, you’d see scatter. I suppose it’s fine if you don’t want to see the scatter, but remember that it would have been there. In any case, I don’t see why anybody would plot a different point for each month, if they were going to use 21 years of averaging. Anyway, I’m still puzzled as to what we’re supposed to draw from the plot. The earth has been warming for 30-some years. Great.
Carrot eater —
“Does the value of the slope have any particular significance? Does 143.6 mean anything to you?”
I assume the 143.6x is the conversion from monthly data to annual; the 334.1 the CO2 baseline. The slope agreement is significant because it is so close. It may have a bearing on the direction of causation. Ocean outgassing of CO2 with temperature occurs instantly, while ocean warming from increased atmospheric CO2 would be slower, and would lag. The tight correlation suggests (to me) the former causal arrow.
Maybe Anthony can give a link to the source of the graphic.