Some media attention has recently focused on research into the concentration of carbon dioxide in the atmosphere, and its relation to the ability of plant life and the soil itself to hold on to CO2. The research (Knorr et al. 2007, Geophysical Research Letters, Vol. 34, pg. L09703) claims that because of changes in temperature and precipitation brought about by global warming, the rate at which plants and soil take in/give off CO2 has been changing. This could portend unpleasant changes which will cause atmospheric CO2 to increase faster in the near future. That could, in turn, increase the rate of global warming, which would further reduce the ability of plants and soil to hold on to their carbon, which would further increase atmospheric CO2, further exacerbating global warming, etc… a classic (and rather nasty) “feedback mechanism.”
Knorr and colleagues used a model of the global biosphere, including the effects of plant and soil respiration of CO2, to investigate how climate conditions affect the biosphere flux of CO2 into the atmosphere. They then attempt to find the possible relationship between climate conditions, specifically el Nino conditions and global land precipitation, and the fluctuations in the growth rate of atmospheric CO2.
That CO2 is increasing is beyond doubt, and the steady increase is unquestionably due to human activity: the burning of fossil fuels. Knorr et al. estimated global CO2 concentration by averaging the measurements made at Mauna Loa in Hawaii and those from the south pole (atmospheric concentration is expressed as “ppmv,” or “parts per million by volume”):
There are two strongest patterns in the change of atmospheric CO2. One is a steady rise, at a rate (averaged over the interval 1980 to 2005) of 1.56 ppmv/yr. The other is an annual cycle, with more CO2 in the atmosphere around May and less around September-October, due to the respiration of (mainly northern-hemisphere) land plants. During northern-hemisphere spring and summer, plants take CO2 out of the atmosphere to build their tissues, but during fall and winter they decay, returning that CO2 to the air. The northern hemisphere dominates because most of the land is in that hemisphere, so that’s where most of the land plants are.
We can easily model the steady rise as a linear increase, and the annual cycle as a seasonal pattern, mathematically:
Evidently, from about 1992 to 1998, measured CO2 was slightly less than the linear+seasonal pattern, while from 2002 to 2005 measured CO2 was higher than this pattern. In fact we can subtract the “linear+seasonal” pattern from the observed data, to find how atmospheric CO2 has changed in addition to its steady rise and seasonal pattern. This defines “CO2 anomaly”:
Now we can see much more clearly that in addition to the steady rise and seasonal pattern, there’s a bit of fluctuation in CO2 levels. This raises the question, what is the cause of these additional fluctuations? It’s probably not simply due to increased emissions from human activity; the CO2 anomalies don’t match similar anomalies in records of man-made emissions.
Since detailed measurements began in 1958, about half the CO2 emitted by human activity has been absorbed by natural systems; the other half is the source for the steady rise. One hypothesis for the fluctuations in CO2 anomaly is that the impact of earth’s natural systems — in particular, the biosphere — on the carbon cycle fluctuates in response to a number of factors, including climate.
Investigating this possibility is the purpose of Knorr et al.’s research. They studied the history of two climate factors, el Nino and total land precipitation, in an attempt to understand their influence on the biosphere contribution to the carbon cycle. They used a model of biospheric CO2 activity called “BETHY” to estimate the impact of climate on the biosphere carbon cycle, and combined it with an atmospheric mixing model called “TM2″ to estimate the effect on atmospheric CO2, then compared that to observations. What they actually calculated is the growth rate anomaly, which is simply the growth rate of the CO2 anomaly:
The green line is the estimated CO2 growth rate anomaly according to their model, the solid black line is the observed CO2 growth rate anomaly. The match is actually rather good, especially in explaining the fast growth rate of CO2 anomaly during the strong el Nino of 1997-1998, and during the unusually dry climate of 2002-2003. According to their estimation, the strong el Nino was the principal cause of the 1998 surge in CO2, but the unusually dry climate over global land areas was the principal cause of the 2003 surge. The model does a better job explaining the surge in 2002/2003 than that of 1998; the authors suggest that part of the reason for the higher-than-estimated CO2 surge in 1998 may be the large peatland fires observed in Indonesia, an effect which is not included in the BETHY model.
But the match is far from perfect. In particular, it doesn’t do very well at matching the low growth rate anomalies in 1991 & 1992 (which could be related to the Mt. Pinatubo explosion), or the fluctuations from 1999 to 2002. This in fact brings me to my strongest criticism of the paper, that no quantitative statistical test is reported to evaluate just how good is the match between prediction and observation.
Nonetheless, the results are suggestive that the biosphere carbon cycle is subject to greater fluctuations than have previously been appreciated. Furthermore, the work offers solid evidence that dry years over land, an expected consequence of global warming, are likely to trigger release of CO2 from plant life and soil, increasing the greenhouse effect and therefore global warming.
This could have very important repercussions for our planet’s climate future. But these results have not yet been subjected to sufficient review to be sure of their correctness. Therefore, while I strongly suggest that this is an area greatly in need of further study, it seems to me that the press reports are exaggerating the conclusiveness of the study, and the certainty of the implications. In fact, as the authors themselves say,
Further research will be required to investigate if this fluctuation carries features of projected future climate change and the CO2 growth rate anomaly has been a first indicator of a developing positive feedback between climate warming and the global carbon cycle.
So, I’m not yet ready to declare that the biosphere is rearing its ugly head, to bite us in the nether regions, creating a CO2 feedback disaster. But it is certainly an issue which deserves very careful watching.
21 responses so far ↓
fermiparadox // May 15th 2007 at 4:25 am
Does the model also account for the varying anthropogenic carbon emissions? I have plotted data from http://cdiac.esd.ornl.gov/trends/emis/tre_glob.htm . There are dips in 1992 and 1999, and it looks like they match at least partially your anomaly plot, but I did not make a statistical analysis.
[Response: No, the authors state, “We neglect ocean, land use and fossil fuel fluxes as they are believed to contribute less to interannual fluctuations.” I’d be very interested to know what the fluctuations in fossil fuel fluxes are. I have heard suggested, for instance, that the reduction in CO2 growth rate during 1992-1993 is partly due to the collapse of the economy of the former Soviet Union.]
fermiparadox // May 15th 2007 at 4:44 am
I have given you a link in my comment, you can use those data, not?
There is a slowdown in 1973 (oil crisis?), a decrease 1979-1983 (another oil crisis started in 1979) and those two I mentioned.
I have heard that the decrease 1992-1993 is due to the breakdown of the eastern European economies, too.
I don’t know what happened in 1998, though.
[Response: I looked at the emissions data, and compared the “emission rate anomaly” to the “CO2 growth rate anomaly.” The match is not very good, although the high emissions anomaly in 1995 corresponds well to the mid-high growth rate anomaly at the same time.
I would have expected that fluctuations in emissions would have a greater impact on atmospheric CO2. I’ll see whether I can find figures for the *amount* of CO2 involved in the biosphere fluctuations (rather than just normalized anomalies).]
Heiko Gerhauser // May 15th 2007 at 1:52 pm
It seems to me that their assumption is a very reasonable one, if you translate fossil fuel emissions into ppmv, then the fluctuations are small, it’s only the growth rate that fluctuates.
To have an impact on interannual fluctuations, we should be looking at 10% interannual falls/rises in emissions or something like that, not changes in the growth rate of 1 or 3%. Or am I overlooking something?
[Response: Pondering it, I get the impression you’re right. Emissions right now are about 7 gigatonne, and the annual CO2 growth rate is about 2 ppmv. That translates to *roughly* 1 ppmv/3.5 Gt. The interannual fluctuations in emissions are generally not more than 0.4 Gt, which would be a bit over 0.1 ppmv. But the observed CO2 growth rate anomalies are larger than that — around a full ppmv.]
Dano // May 15th 2007 at 7:09 pm
I haven’t looked at the paper (can’t get GRL), but in my mind I can visualize strong contributions of grassland/steppe to both the low and high C anomalies. The growth rates of biomass in these regions is rapid and fluctuates interannually as the biomass growth is highly dependent upon precipitation.
Absent a statistical test, this suggests to me a possible explanation for your fluctuations from ’99-02 is the precip variability across midlatitudes during those years (’99 was la Niña) – look also at the similar variability following the ‘91-92 El Niño (no la Niña). I’d like to see if anyone has soil C numbers during ’99-02 to see if there are any correlates there. Surely the FACE grassland folks should have useful data too…
Best,
D
the Grit // May 15th 2007 at 9:38 pm
Hi tamino,
First:
Knorr et al. estimated global CO2 concentration by averaging the measurements made at Mauna Loa in Hawaii and those from the south pole (atmospheric concentration is expressed as “ppmv,” or “parts per million by volume”)
Really? Two data points to represent the entire planet? Very sad.
On the other hand, I have half as many data points to go on, and I am about to harvest a record breaking hay crop. Heck, the hairy vetch is growing so fast this year that you risk being covered over if you stand still in the field for too long. The fruit crops are going to be spectacular as well, if the number of blossoms is any predictor. Even my wife’s roses had a spectacular first bloom, there’s a picture on our blog. Also, based on the number of mating pairs we’ve had buzzing around the farm, there’s going to be a bumper crop of humming birds. In a less pleasant statistical measurement, considering the number and kinds of critter corpses along the road side, that segment of the biosphere is doing quite well, other than still not knowing to look both ways before crossing the street. Oh, and we even have wild turkeys again for the first time in 30 years! That, by the way, referees to the birds, not the bourbon
Thus, at least here, higher CO2 levels and whatever warming we’ve experienced don’t seem to be having an adverse effect.
the Grit
[Response: How you doin’? If you look at CO2 measurements from around the world, you’ll find that they’re all *strikingly* similar, showing the same growth rates world-wide. The biggest difference in different locations is between the northern and southern hemispheres. The “mixing time” for atmospheric CO2 is on the order of 6 months, so the southern hemisphere shows less seasonal pattern than the northern. By averaging Mauna Loa and south pole measurements, Knorr et al. obtain a good representative estimation of worldwide fluctuations on short timescales.
I’m glad to hear you’re anticipating a bumper crop. I suspect it has more to do with the right amount of rainfall than with CO2 fertilization. Alas, on the other side of the planet, the farmers in Australia are suffering what probably is an effect of man-made climate change, the worst drought on record in Australia. Their agricultural production is in the toilet, and I hear that the suicide rate among farmers and ranchers is through the roof. Here’s hoping — sincerely — you never have to face a similar misfortune.
And there’s nothing wrong with the Bourbon!]
Dano // May 15th 2007 at 11:01 pm
That hay crop likely has less nitrogen in it in a higher CO2 world, making it less nutritious and requiring more hay to get the same amount of nutrition as before.
Best,
D
fermiparadox // May 15th 2007 at 11:24 pm
Grit, you will find globally averaged data at http://www.esrl.noaa.gov/gmd/ccgg/trends/
Just to prove tamino’s point that they are similar worldwide.
Dano // May 16th 2007 at 12:49 am
CO2 is a well-mixed gas.
Best,
D
Fielding Mellish // May 16th 2007 at 12:50 am
So, Grit could be describing his local weather, the weather in nearby geography, or climate. It could be climate or surrounding weather in the sense that declining nearby habitats are pushing nearby animal populations toward his current oasis. If warming climate is responsible for this Spring flush, then a heavier and more varied pest population mightn’t be very far behind. If he’s within the current range of Solenopsis invicta, then he may be seeing one or more additional swarms per year. If this Spring flush is due to AGW and if he’s near to, but not currently in, invicta’s range, he has a real treat in store. He’d wish he never heard of AGW. Meanwhile, the saying “make hay while the sun shines” applies as adroitly to the development of non-C energy alternatives.
george J // May 16th 2007 at 1:03 pm
Grit said: “at least here, higher CO2 levels and whatever warming we’ve experienced don’t seem to be having an adverse effect.”
Many of the worst effects of climate change (eg, flooding due to sea level rise) may not be felt for several generations — after you and I are dead and gone.
Does that mean it’s OK for us to ignore them now and let our grandchildren and their children deal with them?
Alexander Ac // May 17th 2007 at 9:41 am
Nice post.
If the CO2 anomaly goes up again after 2005 we are in more trouble (than the upward trend is evident!) - but this is what climate models predict - lower sink CO2 capacity with increasing temperatures. But in either case - with further warming not only biosphere will fail to absorb additional anthropogenic sources, but also the oceans will do the same…
further, the CO2 surge in not crisis *now*, but that is also the case for AGW!
And again, until everybody see that GW is a problem, it is far too late to do somthing
Best,
george J // May 18th 2007 at 5:39 pm
This may not be responsible for the recent surge (since saturation of the southern ocean was reached in the early 80’s, according to the study), but it certainly does not bode well.
Reported in The Independent
“Climate change itself is weakening one of the principal “sinks” absorbing carbon dioxide - the Southern Ocean around Antarctica - a new study has found.”
“the research team [from University of East Anglia, the British Antarctic Survey and the Max-Planck Institute for Biogeochemistry, published in the journal Science] has found the vast Southern Ocean, which is the earth’s biggest carbon sink, accounting for about 15 per cent of the total absorption potential, has become effectively CO2-saturated.”
the Grit // May 19th 2007 at 10:21 pm
Hi tamino,
First, thanks for the personal concern. We’re doing fine here, and I hope the same is true for you.
As to crops, we get this every few years. It’s a combination of a mild winter and just the right amount of rain. It also helps that we’ve had a mild spring. Around here, we can have some unusually hot days, this time of year, which most of the plants don’t like. Really though, the credit should go to my wife’s uncle, who left us the farm, and established the plant mix in the fields. We have other grasses that do well after a cold winter and a dry spring. While we don’t get as big a harvest, at least we get something
As to Australia, their drought, and Global Warming, what’s your proof? We’ve had the same problem here in the past, before GHG went up enough to make a difference. Also, I seem to recall being warned not to mistake the unusually cold winter in the North East as evidence that there was no Global Warming. The same advice would seem to apply here.
the Grit
Hank Roberts // May 24th 2007 at 12:52 am
http://environment.newscientist.com/data/images/ns/cms/dn11899/dn11899-1_600.jpg
http://environment.newscientist.com/article/dn11899-recent-cosub2sub-rises-exceed-worstcase-scenarios.html
George // May 24th 2007 at 5:06 pm
While it is certainly of concern that the yearly emissions (in Gt) are growing faster now than in the nineties (3% now vs 1% then), as indicated by the above study, it is clear that this change alone is not sufficient to account for the rather large recent surge in the yearly CO2 increment (in ppm) — ie, change in the yearly addition to the atmospheric CO2 concentration (in ppm) — which averaged about 1.5 ppm in the nineties, but has averaged about 2 ppm since 2000 — a 33% increase (as shown by Tamino above and in more detailon this NOAA site)
tamino // May 24th 2007 at 9:22 pm
At this point, many factors seem to be conspiring to accelerate CO2 increase in the atmosphere. Human emissions are rising faster, the biosphere (as addressed in the post) shows signs of reduced capacity to hold CO2, and new research suggests the southern oceans are becoming saturated, and will no longer be able to take up as much of human emissions as in past decades.
The future of atmospheric CO2 concentration isn’t certain, but at this point, a worrying number of factors give a very pessimistic outlook.
Ike Solem // May 28th 2007 at 5:17 pm
Just to add another main factor in future CO2 emissions, there’s the permafrost to consider:
“Permafrost soil blanketing northeastern Siberia contains about 75 times more carbon than is released by burning fossil fuels each year. That means it could become a potent, likely unstoppable contributor to global climate change if it continues to thaw. So conclude three scientists in a paper set to appear Friday in the journal Science” (June 2006)
http://www.physorg.com/news69692382.html
The trend towards a poleward expansion of the subtropical dry zones and a resulting transition from forests to grassland will also result in less carbon storage in the biosphere.
The indications that the oceans are becoming saturated with CO2 faster than expected is indeed ‘alarming’ since the the oceans have absorbed ~50% of all fossil fuel CO2 emitted over the past century - saturation means that a 50% reduction in fossil fuel use would have zero effect on the rate of CO2 accumulation in the atmosphere.
[Response: I would add one caveat. If the oceans stop absorbing CO2, *and* we cut emissions by 50%, the emissions cut would prevent the *doubling* which would otherwise ensue — hence it would have a potent (and likely, most necessary!) impact.]
George // May 29th 2007 at 4:53 pm
Land use effects may also be important in the recent surge.
The IPCC found the biosphere to be a net sink for CO2, but recent research indicates otherwise.
Biosphere is source, not sink, for carbon dioxide emissions, study shows
“Using a more advanced version of the Integrated Assessment Model, Atul Jain, a professor of atmospheric sciences at the University of Illinois at Urbana-Champaign, and graduate student Xiaojuan Yang report that the biosphere might now be acting as a source, not as a sink. Rather than storing carbon dioxide, the biosphere may have recently begun driving atmospheric levels higher. “
stevereenie // Jun 6th 2007 at 2:31 am
Just a short note in your response to the Grit,
You said: ” Their agricultural production is in the toilet, and I hear that the suicide rate among farmers and ranchers is through the roof. Here’s hoping — sincerely — you never have to face a similar misfortune.”
I say: Are you sure that the suicide rate is a result agricultural production or maybe the other way around? …… Next Stop Lauderdale
cilibrar // Jun 18th 2007 at 6:06 am
This CO2 thing seems like a “burp” compared to the reality of melting methane hydrate ice.
Wim van der Veen // Jun 18th 2007 at 11:17 am
As to your modelling of data where you started with, i would like to say that, looking at Mauna Loa data (which is available over longer periods) it makes more sense to use an exponential base line instead of a linear base line.
An exponential curve (starting at 275ppm in 1800) fits the data very well, at least the most recent 50 years. Using that, your anomaly analysis might be even more interesting.
[Response: It was the choice of Knorr et al. (in their research into the biosphere response to CO2) to choose the time frame.
The atmospheric CO2 concentration since 1958 (beginning of the Mauna Loa data) is certainly nonlinear. I often hear it referred to as exponential, but the data don’t support an exponential model over, say, a quadratic one. Whichever it turns out to be, the rate of increase is bigger now than it was when Mauna Loa observations began. My next post will be about the most recent trends in CO2 concentration.]
Leave a Comment