Open Mind

Yet More CO2

August 8, 2008 · 61 Comments

This has turned out to be “CO2 week,” with numerous posts on the topic already. I figured I was done with it for a while, but in comments to the last post, some questions were asked which deserved a detailed look at CO2 measurements. One reader asked why the increase is so linear in spite of the fact that human emissions have increased so much — shouldn’t CO2 levels be accelerating? I responded that the increase wasn’t nearly as linear as the graph might suggest, that in fact CO2 growth had accelerated. Another reader asked whether or not the amplitude of the annual cycle had changed, as one of the graphs suggested. I replied that yes, it had.


I’d like to show that these phenomena really are present in the CO2 data from Mauna Loa. This involves isolating the growth rate and the annual cycle from the fluctuations. Let’s begin by looking again at the Mauna monthly average CO2 concentration:

It’s evident that there is an increase (and hence a growth rate), and that an annual cycle is superimposed on it. We can see this even more clearly if we zoom in on the data since 2000:

By removing the annual cycle and smoothing the data, we can isolate the growth rate:

It’s obvious that CO2 concentration has risen quite consistently throughout the time span of these observations. But what’s the actual growth rate, and how steady has it been? Linear regression indicates that the average growth rate over the whole span of data is 1.42 ppmv/yr. Yet We can see from the above graph that the growth rate hasn’t been perfectly constant, but the deviations from constant growth (from a straight line) seem small in the above graph. We can get a better visual impression by fitting a straight line to the smoothed curve, then subtracting it to leave residuals; this will tell us just how the smoothed CO2 growth deviates from linearity:

Now we can see quite clearly that the growth hasn’t been steady. The downward slope of the residuals early in the time span tells us that CO2 was increasing at less than the average rate, while the upward slope late in the time span shows that now it’s increasing at more than the average rate, so it has indeed accelerated. Just how much has it done so? We could estimate the growth rate by subtracting each year’s average value from the previous year’s, but that would include a lot of the fluctuations as well, and we’re more interested in the persistent behavior than the fluctuations. So I determined the changing growth rate by two methods. First, I did a combined linear regression + Fourier analysis of each 4-year time span (Fourier analysis to remove the annual cycle, linear regression to estimate the growth rate). I also applied a “modified Savitzky-Golay filter,” which enables us not only to smooth the data but to determine its growth rate as well. They gave these results for the growth rate of CO2 at Mauna Loa:

The two methods are in excellent agreement. They also show that early in the Mauna Loa record, CO2 was growing at about 0.7 ppmv/yr, but now it’s growing at about 2.1 ppmv/yr. So indeed the growth has accelerated, so much so that today it’s about 3 times as fast as it was when measurements began at Mauna Loa.

We’re also interested in the annual cycle. The Fourier analysis + linear regression also records the amplitude of the best-fit sinusoid for each 4-year time span, which is an excellent approximation of the amplitude of the annual cycle. A wavelet analysis does the same thing. And here it is (the plot is of semi-amplitude, which is just half the amplitude):

So indeed it’s true that the annual cycle has gotten slightly bigger over time.

One other notable effect emerges from estimating (using a wavelet analysis) the phase of the annual cycle, i.e., the time of year at which the cycle peaks. Here’s the phase throughout time:

We see that the phase has decreased over time, so the peak in the annual cycle has tended to come earlier in the year as time has progressed. The change over the observed time span is about 0.026 years, which is just about 10 days.

We can confirm this by taking two 10-year blocks, one from early in the data (1960 to 1970), another from late (1998 to 2008), and using them to compute the average annual cycle. We’ll subtract the linear trend to isolated it from the growth rate, and by averaging the 10-year time span we’ll remove most of the fluctuations. This plot shows two complete cycles for each of these 10-year periods:

From this, it’s obvious that the annual cycle has changed, not only getting larger, but coming earlier in the year.

So it turns out that there’s quite a lot of information in the Mauna Loa record about the changes in atmospheric CO2. But finding those changes requires careful study of the data, and having powerful analysis methods sure helps too!

Categories: Global Warming
Tagged:

61 responses so far ↓

  • Joseph // August 8, 2008 at 2:38 pm

    I was looking at the seasonally-adjusted monthly data from 2000 to 20008. There might be a decelerating trend there (I tried to fit a second-order polynomial trend line). I doubt there’s statistical significance in it, though. But it won’t be surprising to see more of this in the future, when you consider facts such as drops in miles traveled.

    Also, if you consider a simple half-life model of excess CO2, as we put more excess CO2 into the atmosphere, more of it will be “reclaimed” by the planet every year. In theory, even if emissions were to remain at current levels, we should get to a point of “new equilibrium” if you will, where the atmospheric concentration doesn’t change. But that’s probably far off.

  • Tim McDermott // August 8, 2008 at 5:04 pm

    What a spectacular post! Not only a confirmation, and quantitization, of the acceleration of CO2 accumulation, but a confirmation, and quantitization, of spring coming earlier than it used to. And then you give us a conundrum. Why would the amplitude of the annual cycle increase? Is CO2 fertilization happening? Is something else going on?

  • nanny_govt_sucks // August 8, 2008 at 5:05 pm

    So indeed it’s true that the annual cycle has gotten slightly bigger over time.

    Could this not be a reflection of the greening planet? Wouldn’t more vegetation mean more CO2 taken up/released in the annual cycle?

  • George Tobin // August 8, 2008 at 6:18 pm

    Thanks for this outstanding piece of work.

    I can understand how the cycles might be larger–more vegetation in Siberia sucking up more CO2 and then releasing more, for example–as opposed to a smaller cycle if there were more ice cover or permafrost in the same spots.

    But what explains the earlier peaks? Why wouldn’t the down phase also start later? Just wondering if anybody has looked at that.

    Thanks again for this informative post.

  • Andrew W // August 8, 2008 at 6:28 pm

    Would the annual cycle getting larger, and coming earlier in the year, be a product of the change in the rate of increase?

  • Hank Roberts // August 8, 2008 at 6:46 pm

    > greening
    Eutrophication.
    Pond scum.
    http://www.google.com/search?q=Lake+Erie+algae

  • B Buckner // August 8, 2008 at 8:55 pm

    I will join the chorus, very well done and informative. Real value added to the raw data. Fascinating data on the annual cycle. Appears to indicate a longer growing season and some CO2 fertilization, by why only the spring expansion?

  • Dano // August 8, 2008 at 9:02 pm

    Wouldn’t more vegetation mean more CO2 taken up/released in the annual cycle?

    It likely affects the amplitude, off the top of my head. This would not be the explanation for the growth rate or absolute ppmv however. Off the top of my head.

    Best,

    D

  • B Buckner // August 8, 2008 at 9:36 pm

    Got it. Vegetation leafs out or sprouts in the spring based on temperatures. Warmer equals earlier. Plants stop evapotranspiring and the leaves fall off based on waning sunlight which is not temperature dependent, hense no change in CO2 levels in the fall.

  • Adam // August 8, 2008 at 10:13 pm

    “by why only the spring expansion?”

    From memory, for some/many plants, autumn is triggered by lower levels of insolation - eg when daylight hours reduce. Frosts can affect this, but tend to start later than the sunlight reduces.

    I would guess that this will set a lower limit on Spring onset too, but late frosts may be a the first limit (kill off early growth) and if the first change is to reduce these (or make them end earlier), then we will see the earlier start.

    Why the peak is earlier, I’m not sure and may count against what I’ve just written.

  • Hank Roberts // August 9, 2008 at 1:30 am

    > consider a simple half-life model of excess
    > CO2, as we put more excess CO2 into the
    > atmosphere, more of it will be “reclaimed?
    > by the planet every year.

    The tooth fairy model.

    Or you can look it up.

    http://www.sciencemag.org/cgi/content/abstract/1136188v1

  • Timothy Chase // August 9, 2008 at 2:21 am

    Adam wrote:

    Got it. Vegetation leafs out or sprouts in the spring based on temperatures. Warmer equals earlier. Plants stop evapotranspiring and the leaves fall off based on waning sunlight which is not temperature dependent, hense no change in CO2 levels in the fall.

    This sounds about right.

    Plants sense sunlight with their leaves, and depending upon the species will actually track the sun throughout the day. But you can’t sense something if you don’t have the thing you sense with. So they would rely upon temperature during the spring. And of course winter is warming more rapidly than summer.

    Additionally, heavier precipitation is coming during the winter but there is increasing drought during the summer — largely due to the increased rate of evaporation. We went from 20% of the world experiencing drought at any given time during the fifties to 30% more recently, and they are projecting 50% by the end of the century. (From Hadley brochure.)

  • Hank Roberts // August 9, 2008 at 2:30 am

    Sorry, I’m getting overly snarky. The simple linear description doesn’t match reality. Rates of change — many different ones — matter.

    Joseph, see Tamino’s earlier thread on carbon cycling, and, e.g.,

    http://www.enviroliteracy.org/article.php/478.html
    and David Archer’s online textbook on this.

    The problem is the rate of change. There’s likely a boost in biological productivity briefly with increasing CO2, but also a boost in ocean pH, followed by various dieoffs because biological systems don’t adapt to the rate of change we’re causing. See the history of previous asteroid impacts for analogous conditions, or the PETM excursion.

    Yes, eventually, natural processes do remove excess CO2 from the atmosphere. Not on a time scale that matters for human civilization though.

    http://www.pik-potsdam.de/~victor/archer.subm.clim.change.pdf

    http://geosci.uchicago.edu/~archer/reprints/

    http://www.agu.org/pubs/crossref/2005…/2004GC000854.shtml

    http://www.agu.org/pubs/crossref/2005…/2004GC000891.shtml

    http://www.isse.ucar.edu/kleypas/PUBS/caldeira_etal_grl_2007.pdf

  • B Buckner // August 9, 2008 at 12:32 pm

    Hank

    The fairy tooth model is described in Section 7.3.2.1 of AR4. The terrestrial biosphere and oceans have consistently removed 45% of fossil CO2 for the past 45 years. As emissions have increased, so has the take up by plants and the ocean.

  • S2 // August 9, 2008 at 1:16 pm

    Interesting and plausible discussion on vegetation above, but I’m not convinced about autumn/fall leaf loss being triggered solely by insolation changes.

    Looking at the daily CO2 figures from Mauna Loa (as well as Tamino’s Annual Cycle chart above), the entire cycle appears to have shifted - the “troughs” have moved forward in time by about the same amount as the “peaks”.

    The NH growing season is starting earlier, but I think that there is no evidence (from this data alone) that it is lasting any longer.

  • Joseph // August 9, 2008 at 1:37 pm

    Yes, eventually, natural processes do remove excess CO2 from the atmosphere. Not on a time scale that matters for human civilization though.

    From the emissions data, I would estimate that the equivalent of 1/3rd of the CO2 we pump into the atmosphere is removed every year. (A lot of uncertainty in that calculation though). In this case, to just get atmospheric concentrations to stop increasing - never mind drop - CO2 production will have to drop to 1/3rd of what it is. It seems to me that this kind of reduction (or whatever the right number is) needs to happen, and it needs to happen soon.

  • Hank Roberts // August 9, 2008 at 4:33 pm

    Joseph, you seem to be coming up with your numbers in a vacuum and thinking you can make everything very simple.

    Have you read about the many different known biogeochemical cycles?

    It’s easy to think you can describe an outcome as simply related to an individual input — as long as you look only at a very short time slice.

  • Hank Roberts // August 9, 2008 at 4:40 pm

    Joseph, here’s one place you’ll find summaries of the science over the past several decades. The carbon cycle papers will be worth reading:

    http://www.icsu-scope.org/downloadpubs/indexpub.html

  • Hank Roberts // August 9, 2008 at 5:07 pm

    PS, Buckner above is describing such a short term snapshot, referring to the AR4 chapter here:
    http://www.ipcc.ch/pdf/assessment-report/ar4/wg1/ar4-wg1-chapter7.pdf

    He’s conflating photosynthesis and solubility; these are very different, not equivalent ’sinks’ once you look past the atmosphere at the planet (ocean, atmosphere, biota, geology).

    Look again at the Science paper linked above and its references and links, and at the ocean pH references.

    CO2 dissolved in ocean water predictably reaches chemical limits within the current century that will greatly change ocean biology.
    Remember, the first half of the fossil thus far burned was burned before about 1970, at a relatively slow pace compared to the second half burned so far, all since 1970.

    See p. 512 (and remember this is not the published research — for that see the references. This is a very simplified summary):

    ——excerpt—–
    the observed increase in atmospheric CO2
    concentrations does not reveal the full extent of human emissions in that it accounts for only 55% of the CO2 released by human activity
    since 1959. The rest has been taken up by plants on land and by the oceans. In all cases, atmospheric concentrations of greenhouse
    gases, and their increases, are determined by the balance between sources (emissions of the gas from human activities and natural
    systems) and sinks (the removal of the gas from the atmosphere by conversion to a different chemical compound). … The concentration of CO2 is now 379 parts per million (ppm) and methane is greater than 1,774 parts per billion (ppb), both very likely much higher than any time in at least 650 kyr (during which CO2 remained between 180 and 300 ppm and methane between 320 and 790 ppb). The recent rate of change is dramatic and unprecedented; increases in CO2 never exceeded 30 ppm in 1 kyr – yet now CO2 has risen by 30 ppm in just the last 17 years.

    —–end excerpt——

    Key point — the AR4 Ch. 7 is _only_ about the atmospheric increase. It’s not addressing where the rest of the CO2 goes or what happens.

    Limiting your focus is the way one can imagine the “tooth fairy” takes away the problem.

    That’s snapshot, short term, fragmentary thinking.

    Biogeochemical cycling: nothing goes away.

  • Hank Roberts // August 9, 2008 at 5:34 pm

    Oh, and Buckner above refers to 7.3.2.1 and misinterprets it to say he believes the planet is taking up the increases in fossil carbon. Bogus. Read the actual document.

    The very last lines of 7.3.2 are:

    “… when the carbon cycle is included, the models
    consistently simulate climate feedbacks to land and ocean carbon cycles that tend to reduce uptake of CO2 by land and ocean from 1850 to 2100 (see Section 7.3.5).”

    The clue there? see Section 7.3.5.

  • Joseph // August 9, 2008 at 7:08 pm

    Joseph, you seem to be coming up with your numbers in a vacuum and thinking you can make everything very simple.

    No, in fact, I noted that I can’t be very certain of any such estimates, and I don’t see that the literature provides much more certainty. All we know is that it takes quite a while for CO2 to get removed from the atmosphere. I’ve seen excess half-life estimates that go from 19 years to 200 years. Of course, I realize that there’s not necessarily a fixed equilibrium point for CO2, and how easily CO2 gets removed depends on many things like vegetation and who knows what else, so the “half-life” concept is just a construct that helps us think about reality.

    But just to elaborate on my 1/3rd claim, that’s assuming a half-life of about 70 years. This is apparently the best half-life that explains the Mauna Loa data, using CDIAC emissions data as input. Basically, I estimate that of the amount of CO2 we have pumped into the atmosphere, an excess of 230,000 million metric tons remains. About 1% is removed every year (at the presumed half-life). So about 2300 million metric tons would be removed every year. As of 2004, we’re pumping about 7910 million metric tons into the atmosphere every year. So it’s more like 1/4rth.

    That changes considerably if we just assume a half-life of 20 years, though, which is not out of the realm of possibility. In this case, about 50% of what we pump every year gets removed.

    Again, I can’t be sure about these estimates. Take them for what they are. I guess we’ll have a better idea when and if the CO2 concentration levels off.

  • Hank Roberts // August 9, 2008 at 7:28 pm

    > All we know is that it takes quite a while

    But if you will read the references you will know much more than that. That’s what I’m suggesting you do. Study the science and you will know much more than you’re describing now.

  • Dano // August 9, 2008 at 7:53 pm

    BBuckner/Adam/S2:

    Plants green up in spring with day length and temperature; soil moisture availability is a control for perennials, esp grasses. Plants go dormant in the fall with day length and temperature; soil moisture availability is a control for woody perennials and some grasses (herbivory is a confounding factor for grasses).

    In the spring, temp is the stronger control, with day length a factor. In fall, day length is the stronger control with temperature a factor. You may have seen the term ‘degree days’ for crops, which may help visualize this - degree days being a function of both temp and day length.

    We saw this on the Colorado Front Range this year, as La Niña kept temps cool, delaying green-up, and the limited soil moisture adding to the delay and shorter green season for High Plains grasses and forbs. The gardeners here found beans and peas to be poor this year, with a long period of cool temps leading right into hot weather. OTOH our cool season brassicas and crucifers and leafies did well.

    HTH.

    Best,

    D

  • Hank Roberts // August 9, 2008 at 8:49 pm

    Joseph, did you just come up with the notion that CO2 has a “half-life” yourself? Or are you using a reference for this?

    It’s a poor term in this context; this isn’t radioactive decay, there’s no single reaction path chemically.

    Although the term has been used, you won’t find it in the scientific work I don’t think. I’d welcome a pointer to anywhere you are relying on.

    Regrettably you will find that term applied to CO2 in the denial/PR material, quite a bit. It’s used to confuse the issue, not understand it.

  • Aaron Lewis // August 9, 2008 at 8:52 pm

    Why is nobody looking at changes in human economic and social patterns that result in changes in the annual timing of emissions? The assumption is that the curve is going up because of human emmissions. Thus, the shape of the annual curve may also be affected by the temporal and geographic distribution of emmissions.

  • Adam // August 9, 2008 at 8:55 pm

    TC & Dano, thanks for the clarification. I was discussing from memory (should have looked it up first).

    More reading to do…

  • Hank Roberts // August 9, 2008 at 8:55 pm

    Try this:
    http://www.pik-potsdam.de/~victor/archer.subm.clim.change.pdf

    ——-excerpt——-
    Most of the CO2 drawdown will take place on time scales of centuries, as CO2 invades the ocean, but it is too simplistic to call the invasion timescale the atmospheric lifetime of the CO2, as is commonly done in popular and scientific discussion. We argue that a better shorthand for the lifetime of anthropogenic CO2 would be “hundreds of years plus a significant fraction that changes climate forever”.

    Introduction

    The ocean contains 50 times more dissolved oxidized carbon than the atmosphere does, and 70% of the surface of the earth is covered by ocean. For these reasons, the prevalent opinion among earth scientists in the early twentieth century was that the oceans would prevent industrial activity from increasing the pCO2 of the atmosphere. This view prevailed until precise measurements of free-atmosphere pCO2 values showed an increasing trend of (at that time) 0.8 ppm yr-1 (for current data see http://cdiac.esd.ornl.gov/trends/co2/contents.htm ).
    At about the same time as the first accurate pCO2 measurements, Revelle and Suess [1957] showed that the uptake of CO2 into seawater is enhanced by carbonate buffer chemistry, but only one tenth as strongly as might be naively inferred from the relative concentrations of carbon in the air and in the water….

    —–end_excerpt—–

  • B Buckner // August 9, 2008 at 9:48 pm

    Dano - Thanks for the more thorough explanation.

    Hank - Perhaps I misunderstood your original point. The reference you provided indicated the southern oceans are taking up less CO2. This post by Tamino is about measurements of CO2 in the atmosphere. The measurements clearly show the oceans and biosphere continue to take up more CO2 at the same rates they have been for at least 45 years (since we have been measuring). This is not a snapshot. If the southern oceans are taking up less CO2 it is not showing up in the data. If your point was that we should not rely on this mechanism into the future, I agree. The science indicates trouble ahead on many fronts.

  • Hank Roberts // August 9, 2008 at 10:41 pm

    45 years _was_ a snapshot.

    Taking up less refers to the proportion, not the absolute tonnage.

    Work forward from the link:

    “Based on observed atmospheric CO2 concentration and an inverse method, we estimate that the Southern Ocean sink of CO2 has weakened between 1981 and 2004 by 0.08 PgC/y per decade relative to the trend expected from the large increase in atmospheric CO2. This weakening is attributed to the observed increase in Southern Ocean winds resulting from human activities and projected to continue in the future. …

    http://www.sciencemag.org/cgi/content/abstract/1136188v1

    Published Online May 17, 2007
    Science DOI: 10.1126/science.1136188

    This is more recent than the IPCC’s latest numbers.

    Click the various links on that page for citing and related work since a year ago.

  • Brian Klappstein // August 10, 2008 at 4:45 am

    “…I was looking at the seasonally-adjusted monthly data from 2000 to 20008. There might be a decelerating trend there…”

    (Joseph)

    I don’t think there is any doubt the growth rate in CO2 has shrunk since 2002. Which would mean the correlation with anthropogenic CO2 growth is poor over that period since anthropogenic CO2 growth has been accelerating over that period.

    Regards, BRK

    [Response: You're wrong, and frankly your claim is absurd. There is NO evidence that the growth rate has shrunk since 2002.]

  • B Buckner // August 10, 2008 at 12:25 pm

    “This is more recent than the IPCC’s latest numbers.” The July 2008 CO2 measurements of the atmosphere from Mauna Loa show an unchanged proportion of CO2 being taken up by the oceans and biosphere. No evidence today of any reduction in the ocean sink.

    “45 years _was_ a snapshot.” You indicated previously that half of the fossil fuel ever burned has been since 1970. If a snapshot, it is a snapshot of the entire period during which the ocean have interacted with an atmosphere enriched with CO2 from burning fossil fuel.

    The link: http://www.sciencemag.org/cgi/content/abstract/1136188v1 You are hanging your hat on one paper. This paper seems far from settled science as evidenced by the comments:
    “Unlike Le Quéré et al. (Reports, 22 June 2007, p. 1735), we do not find a saturating Southern Ocean carbon sink due to recent climate change. In our ocean model, observed wind forcing causes reduced carbon uptake, but heat and freshwater flux forcing cause increased uptake. Our inversions of atmospheric carbon dioxide show that the Southern Ocean sink trend is dependent on network choice.”
    “We disagree with the conclusion of Le Quéré et al. (Reports, 22 June 2007, p. 1735) that poleward intensifying winds could continue to weaken the Southern Ocean sink in the future. We argue that altered winds, along with rising atmospheric carbon dioxide, will likely increase the efficiency of this sink in the 21st century.”

  • Brian Klappstein // August 10, 2008 at 2:02 pm

    “…[Response: You're wrong, and frankly your claim is absurd. There is NO evidence that the growth rate has shrunk since 2002.]…”

    (Tamino)

    I repeat my claim that there is clear evidence the growth rate has shrunk since 2002. You can see this in the recent trend of month minus same month last year data points. You’re using Fourier analysis to eliminate the annual cycle, but the month minus same month last year eliminates the annual cycle too.

    Using that metric the, highest CO2 growth ever was in August 1998 (3.92 ppmv increase). Since it would be “cheating” to start the trend in an El Nino and end in a La Nina, I’m just quoting the trend from January 2002:

    [Response: Trend? You don't know the meaning of the word. The slope of the regression line is -.08 ppmv/yr, but it's NOT statistically significant. NOT EVEN CLOSE.

    It's really typical of idiots that they'll take any number they can find to support their denial, ignore proper statistical evaluation, and make pronouncements about trends or changes in trends. It's no coincidence that such bonehead mistakes usually involve less than a decade of climate data and ignore autocorrelation.]

  • Brian Klappstein // August 10, 2008 at 2:14 pm

    Continued…..

    That trend is -.08 ppmv growth per year. As an interesting aside every growth in Co2 month over same month last year over 2.6 ppmv (since 1970) is preceded by a period where the SOI is less than -10. The lag period varies between 5 and 13 months. The converse is not true however, there are periods where the SOI drops below -10 where there is no following growth spurt over 2.6.

    Regards, BRK

    [Response: An interesting aside: I've wasted enough time on your incompetent "trend" analysis already.]

  • Hank Roberts // August 10, 2008 at 2:47 pm

    Buckner’s picked one comment out of the large material published in response the Le Quere paper. Le Quere et al long ago answered that point in the material there.

    You know know to find this.

    Read and see for yourselves — read the science, which continues to come in.

    It’s rhetoric to state a belief then go looking for and quote tiny bits out of the science, without more, in the blog as though it were a conclusive statement supporting the belief.

    The science is over there — pointing to the continued discussion is what we can do here.

  • Joseph // August 10, 2008 at 4:02 pm

    I also don’t think there’s statistically significance in what you see, Brian. If you think there is, why do you think that?

    Of course, it’s always possible that changes are occurring and we can’t detect them through statistical significance, yet. There’s not much that is surprising about a decelerating trend in CO2.

  • Brian Klappstein // August 10, 2008 at 7:32 pm

    “…I also don’t think there’s statistically significance in what you see, Brian. If you think there is, why do you think that?…”

    (Joseph)

    I get the sense my comments are no longer welcome here, however in the interest of science I’ll add a few more points. I don’t think there is statistical significance in the slowing growth of CO2 since 2002……yet. I never did say that, except perhaps by referring to this observation as a “trend”.

    However I do think there is statistical significance in the relationship between the SOI and the monthly CO2 growth rate since 1970. So far all I’ve gotten in response to this second hypothesis is a lot of angry bluster; no hard statistical analysis.

    On the topic of my second hypothesis, not only are the growth spurts in CO2 preceded by SOI drops, but for the most part the growth sags in CO2 are preceded 3 to 13 months by SOI values above 10, the major exception being the period following the Pinatubo eruption.

    Regards, BRK

    [Response: You originally said "I don’t think there is any doubt the growth rate in CO2 has shrunk since 2002." Now you say "I don't think there is statistical significance in the slowing growth of CO2 since 2002......yet. I never did say that."

    The relationship between SOI (and/or ENSO) and CO2 growth probably has a lot to do with the biosphere response to the changing weather patterns associated with SOI/ENSO. See this.]

  • David B. Benson // August 10, 2008 at 10:31 pm

    SOI?

    [Response: SOI = Southern Oscillation Index.]

  • Joseph // August 10, 2008 at 11:55 pm

    I get the sense my comments are no longer welcome here, however in the interest of science I’ll add a few more points. I don’t think there is statistical significance in the slowing growth of CO2 since 2002……yet. I never did say that, except perhaps by referring to this observation as a “trend?.

    You did say the following, Brian.

    I don’t think there is any doubt the growth rate in CO2 has shrunk since 2002.

    I’d advise more cautious language when making claims that are by no means certain.

  • Hank Roberts // August 11, 2008 at 4:46 am

    Brian did say “I don’t think.”

    Can’t argue with what people _believe_.
    Everyone’s entitled to their own opinion.

    It’s the facts that we’re trying to nail down, to actually describe the objective world as it is. That’s what takes time and tools.

  • Hank Roberts // August 11, 2008 at 4:48 am

    Brian, which of the many published relationships between ENSO and CO2 are you thinking of? Or do you have one not yet published?
    http://scholar.google.com/scholar?q=relation+between+ENSO+CO2

  • Brian Klappstein // August 11, 2008 at 6:06 am

    “…he relationship between SOI (and/or ENSO) and CO2 growth probably has a lot to do with the biosphere response…”

    (tamino)

    I agree with that but I’ll tell you why I think the SST control of the ocean CO2 sink has more influence:

    1. El Nino periods are warmer periods, globally speaking, and all other things being equal, that equates to greater CO2 fixing by the biosphere. In other words CO2 growth rate should drop after an El Nino, not accelerate.

    2. El Nino periods do coincide with dryer conditions in some areas, but these areas (Indonesia and Northern Australia) are to a degree balanced out by wetter conditions in the southern US, east central Africa and the west coast of South America.

    3. The winter month growth rates during an El Nino are dependent on biosphere response in the smaller land area southern hemisphere since the majority of the boreal forest (Sibera and Northern Canada) are dormant at this time.

    In summary, to believe the biosphere trumps the ocean sinks as modulators of the CO2 growth rate you have to also believe that for El Nino events that dryer trumps warmer, globally speaking. I don’t think that’s true.

    Regards, BRK

  • Brian Klappstein // August 11, 2008 at 6:39 am

    “…I’d advise more cautious language when making claims that are by no means certain….”

    It is certain that the average growth rate in CO2 over the last 2 years by the month over same month last year metric is below the rates seen in 1998, and also less than the 2003 - 2004 peak.

    You can argue that it’s not statistically significant, which as noted above I agree with. But….whether it represents the start of a new trend, which is entirely possible, or just noise, the fact remains that the current (last 2 years) growth rate is below average, average being defined as Taminos 2.1 ppmv “now its growing at” rate.

    I am absolutely certain of that.

    Regards, BRK

    [Response: You really are reading too much into the fluctuations that exist in the data. Fluctuations exist, and they go both up and down (by definition), but they're not harbingers of change in the system. The appearance of increase and decline, for no other reason than fluctuations, isn't just possible it's inevitable. That's why we insist on statistical significance testing. Read this.]

  • Brian Klappstein // August 11, 2008 at 6:54 am

    “….Brian, which of the many published relationships between ENSO and CO2 are you thinking of? Or do you have one not yet published?…

    (Hank Roberts)

    Tamino directed me to a paper (Knorr et al) which connects the dots between ENSO and CO2 growth, using the impact of ENSO on the biosphere as a lever on CO2 growth. I’m not aware of papers that more directly connect ENSO to CO2 growth through ENSOs effect on SSTs and thus the strength of the ocean CO2 sinks, but I would not doubt they are out there.

    If not, then the idea is ripe for further investigation, since preliminary investigation shows a potential relationship between the two parameters.

    Regards, BRK

  • Ray Ladbury // August 11, 2008 at 2:17 pm

    Brian, The biosphere is not, generally speaking, a long-term CO2 sink. What it absorbs it eventually gives back. True, carbon sequestered in old-growth forests may be out of circulation for hundreds of years, but old-growth forests will not be drastically affected by short-term fluctuations like ENSO. There is no reason to suspect any slowing in atmospheric CO2 growth.

  • dhogaza // August 11, 2008 at 3:49 pm

    Brian, The biosphere is not, generally speaking, a long-term CO2 sink. What it absorbs it eventually gives back.

    As it turns out, even when it’s sequestered in the form of coal and oil …

  • Hank Roberts // August 11, 2008 at 5:53 pm

    > Knorr et al
    Which?
    http://scholar.google.com/scholar?hl=en&lr=&safe=off&scoring=r&q=%2BKnorr+%2Bco2+%2BENSO&as_ylo=2003&btnG=Search

  • Brian Klappstein // August 11, 2008 at 7:18 pm

    “…Which?…”

    (Hank)

    (Knorr et al. 2007, Geophysical Research Letters, Vol. 34, pg. L09703)

    Regards, BRK

  • Brian Klappstein // August 11, 2008 at 7:50 pm

    “…Brian, The biosphere is not, generally speaking, a long-term CO2 sink. …”

    (Ray Ladbury)

    No issue with that. It was Tamino that brought up the hypothesis that the biosphere was a major modulator of CO2 growth, the biosphere itself being modulated by ENSO events. Makes sense, but I don’t believe it to be the major natural modulator of CO2 growth (which I think is SSTs, also modulated by ENSO events).

    I’ll reword my original point more cautiously as:

    POSSIBLY we are seeing the start of a new trend in CO2 growth rates.

    And add to it my reasoning. POSSIBLY this reduction in CO2 growth rate is a result of a POSSIBLE recent cooling trend in SSTs which would strengthen one of the major CO2 sinks.

    Regards, BRK

  • Hank Roberts // August 11, 2008 at 10:08 pm

    Paraphrasing

    If the current series were to be extended at the same rate without normal variation, for __ years, a trend would be significant at the ___ level. If there had been no correction to the Argo data, the uncorrected numbers could have correlated with such hypothetical trend, which, um, did we prove something yet?

  • Hank Roberts // August 11, 2008 at 10:09 pm

    Seriously, Brian, there are a lot of correlations in the literature; why not cite to them instead of to wishful thinking for your basis? It’s a tradition in the scientific literature, and a good habit to practice. You might come up with something that could actually be looked into?

  • ChuckG // August 11, 2008 at 11:04 pm

    http://www.gccoal.com/investor_info/archived_documents/pub_disclosure_01.html
    Grande Cache Coal Project
    No. 7 Underground Mine, No. 8 Surface Mine and Coal Processing Operation

    GCC’s contact person in Grande Cache is:

    Mr. Brian Klappstein, P. Geol.
    Project Manager
    Grande Cache Coal Company Inc.
    P.O. Box 8000
    Grande Cache, T0E 0Y0

    http://wattsupwiththat.wordpress.com/2008/06/07/jason-2-satellite-set-to-launch-june-15th-from-vandenburg-will-track-sea-level/
    Brian Klappstein (19:37:34) :
    “…3.2 mm/year…?
    Not lately. It’s been 2 years since a rolling 12 month average was above 3.2. The last 1 year rolling number available (centered late 2007) shows the one year average growth (fore and aft of that point) to be minus 7 mm/year.
    No real surprise, what with cooling oceans, the currently waning La Nina etc. Of course, lets get ready for the chorus from the AGW alarmists:……altogether now…….?ITS NOT STATISTICALLY SIGNIFICANT!?
    Not yet anyway…
    Regards, BRK

    Bold added.

    [Response: It seems that denialists always have to be told that it's not statistically significant. They can't, or won't, learn.]

  • ChuckG // August 12, 2008 at 2:25 am

    Should have bolded Brian Klappstein

    And from Deltoid
    Oreskes on Western Fuels Association’s war on science
    Category: Global Warming
    Posted on: August 11, 2008 2:41 PM, by Tim Lambert

    John Mashey emails me a link to a video a Naomi Oreskes talk about the Western Fuels Association’s PR campaign against the global warming science.

  • Hank Roberts // August 12, 2008 at 3:29 am

    It is difficult to convince a man of something when his paycheck depends on his not understanding it. (Upton Sinclair)

  • TokyoTom // August 12, 2008 at 4:50 am

    Hank, that quote tells us more than most people understand. It’s not merely that Brian Klappstein might be the same Brian Klappstein, geologist and project manager for Grande Cache Coal, which coal company is naturally interested in defending its business, and whose employees naturally may wish to think that their work activities do not generate climate risk. That’s all fairly obvious and understandable.

    What people are missing is that our state, provincial and federal governments, here in the US, in Canada and elsewhere, largely own the land from which coal is being extracted and collect very significant royalties as a result.

    Political progress is delayed because the federal government doesn’t want to cut off an important stream of revenues without a replacement, and because local governments are strongly opposed to losing such revenues and employment. If we want to see progress, the coal states out West and in Appalachia will need to be bought off.

  • TCO // August 12, 2008 at 2:44 pm

    I don’t think you should do a regression on smoothed data. Just do the regression on the noisy data itself. Fit the second order (or whatever to that). Removing the annual result is fine.

  • TCO // August 12, 2008 at 2:53 pm

    I would also be very catious of using some of these fancy filters and then discerning trends (with implicit statistical relevance). Smoothing and filtering are better for visual representation of data (and for some things like audio quality). Regressed filtered data is troublesome as you don’t know how much degrees of freedom have been compromised, treatment of outliers, etc. I would just be rather leary instead of rather FAST to pull out the big guns like that. (Just my sorta feeling. But I also felt the same way about how Steve McI “overmodeled” the “red noise” in his MBH comparison testing. I have these intutions where I just sorta smell a possible flaw despite not knowing any math. :)

  • Brian Klappstein // August 12, 2008 at 5:34 pm

    Actually the “project manager” is pretty ancient history. I’m a currently geological consultant to the mining industry, and needless to say my views are my own not my clients.

    This “outing” is pretty old too, since that happened on the Globe and Mail blogs about a year ago.

    I will say another thing which should be obvious but perhaps needs repeating. My motives can be whatever this audience wants them to be, but that is irrelevant to whether my hypotheses are correct or incorrect. No amount of casting aspersion on my character changes the R2 statistic on the relationship between SOI and the monthly CO2 growth rate.

    As a final point, if I was really into changing the political landscape on climate change, this wouldn’t be the blog to do it on. It’s too technical, low traffic, and populated by sincere AGW believers to make any progress on that front.

    Regards, BRK

    [Response: What is relevant to whether your hypotheses are correct or incorrect is your disdain for proper statistical evaluation, as indicated by your claims about a decrease in the growth rate of CO2, and the quote from your comment on Watts' blog.]

  • Petro // August 12, 2008 at 6:55 pm

    BRK, I truly hope your clients receive better statistical analysis from you than what you have presented here. Any economical decision which is based on such a sloppy skill is bound to be, if not incorrect, justified incorrecly.

  • Brian Klappstein // August 13, 2008 at 4:55 pm

    “…disdain for proper statistical evaluation,…”

    (tamino)

    Actually I was hoping to intrigue you enough to investigate in detail the statistical relationship between SOI and Co2 growth. A while back I posted my “cherry-picked” R2 of .64 and you noted due to autocorrelation it was unlikely as strong as I believed. But you never came back with any statistics counter to my preliminary investigation.

    Hank Roberts apparently was unable to dig up any papers on the relationship between SSTs and CO2 growth, so to my mind (as noted above) the topic is ripe for investigation. I think you also noted the Knorr et al 2007 paper was weak because it wasn’t supported by appropriate statistical analysis.

    All in all there are lots of reasons to pursue this hypothesis further. If I had the time and the expertise….but I don’t.

    Regards, BRK

  • Hank Roberts // August 13, 2008 at 6:46 pm

    > Hank Roberts apparently was
    > unable to dig up …

    Good grief, Brian. I’m not the miner, you are, if you’re _that_ guy. Are you?

    _You_ have to dig. I can help you figure out _where_ to dig (so can anyone else, it’s cut and paste work with the search engine).

    Let’s paste your text into Scholar one more time again yet:

    http://scholar.google.com/scholar?sourceid=Mozilla-search&q=relationship+between+SSTs+and+CO2+growth

    Consider LeQuere’s work for example.
    http://scholar.google.com/scholar?sourceid=Mozilla-search&q=Lequere+biological+productivity+sea+surface+temperature+climate+model

    Good grief, Brian.

  • TokyoTom // August 14, 2008 at 6:38 am

    Brian, let me ask you a question: do you understand how much warmer it was millions of years ago, when dinosaurs and tropical life flourished at the poles, and the CO2 that we have been releasing from “fossil fuels” wasn’t yet fossilized, but in the atmosphere. As the energy received from the sun was LESS than now (during the Carboniferous period (about 300 million years ago) the solar constant was about 2.5% less than today), where did the higher temps (6-8 C higher than today) come from, if not GHGs? http://www.gcrio.org/CONSEQUENCES/winter96/geoclimate.html

    Do you prefer that we turn the whole world into a Cretaceous Park, or is there some point at which you’d say that it’s fair for us to take control over the global thermometer from the fossil fuel firms?

Leave a Comment