Absorption of Carbon Dioxide
from the Atmosphere

by
Dr. Jarl Ahlbeck

First: the Abstract

Analysis of data published by the CDIAC (1999) for anthropogenic emissions of carbon dioxide from fossil fuels and cement industry for the period 1960-1997 together with measured values of the atmospheric carbon dioxide concentration, indicates that the net sink flow rate, or the uptake rate of anthropogenic carbon by the biosphere and the oceans, is primarily controlled by the concentration of carbon dioxide in the atmosphere. This means that the "consensus" carbon cycle models (IPCC 1996) seem to overpredict the future atmospheric carbon dioxide concentration to a great extent.

And now: Review Comments on Dr Ahlbeck's Paper

Subject: A new carbon sequestration model
Date: Wed, 28 Jul 1999 10:27:41 +0100 (BST)
From: richard@courtney01.cix.co.uk (COURTNEY)
To: jarl.ahlbeck@abo.fi
CC: daly@vision.net.au

Dear Dr Ahlbeck:

I write to congratulate you on your new carbon dioxide sequestration model recently published on John Daly's web site. Your model incorporates effects of biology and, therefore, is a significant advance from the model of Dietze that has successfully demonstrated an ability to emulate empirical reality. I note the important fact that your model is validated by its ability to emulate the Mauna Lua data.

However, I would welcome a clarification of your conversion of your equations 6 and 7 to become your equation 8. You say, "Now the change of the carbon content of the surface water can be eliminated from equations (6) and (7) and we obtain the following expression for the net sink flow rate:". You then state your equation 8.

I did not obtain your equation 8 when I divided your equation 6 by your equation 7. Instead, by conducting the division, I obtained an expression for the change of the carbon content of the surface water as a function of the net sink rate.

Clearly, I have confused the calculation and so I would be grateful for an explanation of how you converted your equations 6 and 7 to obtain your equation 8.

Again, congratulations.

All the best

Richard Courtney

Subject: Re: A new carbon sequestration model
From: "Jarl Ahlbeck" <jarl.ahlbeck@abo.fi>
To: richard@courtney01.cix.co.uk (COURTNEY)
CC: daly@vision.net.au

Yes, I am glad that you read my reports to that extent that they check the equations ! Thank you for that ! I like this discussion on the net ! I think much junk science could be avoided if a large crowd of people have this kind of scienticical chat.

I checked the deriving. It is no use to divide the equations. Instead dmsurf/dt is solved out from eq. (6) giving:

dmsurf/dt = Fem/(1+y/zkH)-Fbio/(1+y/zkH)-Fbulk/(1+y/zkH)

When this expression is inserted for the first term into eq (7) :

Fsi = dmsurf/dt + Fbio + Fbulk

you can write:

Fsi=[1/(1+y/zkH)]*[Fem-Fbio-Fbulk+(1+y/zkH)Fbio+(1+y/zkH)Fbulk]

because we have to multiply the eq (7) Fbio and Fbulk with (1+y/zkH) if the whole expression would be written as divided by (1+y/zkH) that gives eq. (8) after cleaning the [ ** ]

have a good time, Jarl

Subject: Re: A new carbon sequestration model
Date: Wed, 28 Jul 1999 23:59:54 +0100 (BST)
From: richard@courtney01.cix.co.uk (COURTNEY)
To: "Jarl Ahlbeck" <jarl.ahlbeck@abo.fi> CC: daly@vision.net.au

Dear Dr Ahlbeck:

Thankyou. This completely explains my mistake.

All the best

Richard

Subject: Re daly/co2-conc/ahl-co2.htm
Date: Thu, 12 Aug 1999 13:14:04 +0200
From: "P. Dietze" <091335371@t-online.de>
To: jarl.ahlbeck@abo.fi
CC: HVolz@t-online.de, Onar Åm <onar@netpower.no>, Jack Barrett <100436.3604@compuserve.com>, Heinz.Hug@t-online.de, albrecht@biochem.mpg.de, Bjarnason <agust@rt.is>

Dear Jarl,

my congratulation to your new paper! I have now distributed the news to 600 people.

For the eddy mixing into the deep sea you verbally use the IPCC eddy diffusion approach that yields a completey flawed small uptake in future, but on the other hand you quite exactly come to my carbon model solution (apart from the correction for the mixed layer uptake in partitioning mode). I cannot solve the contradiction about diffusion from the mixed layer into the deeper ocean. May be you can explain to me why you need no calculation of the back-pressure that would reduce the diffusive uptake of the mixed layer.

In my German carbon model paper Dietze, P.: "Der Klima-Flop des IPCC" Klima 2000 Vol 2, 24-35 (1998 5-6) and http://www.wuerzburg.de/mm-physik/klima/cmodel.htm I say that there is eddy mixing +take away, but the used CO2 diffusion coefficient of 2.4 cm²/s is indeed double of the 'observed' figure and 3*10^5 of the physical reality. So a diffusive flux (driven by CO2 gradient) does indeed not exist. And I say that because of upbuilding back-pressure IPPC's future CO2 uptake (of the mixed layer) reduces considerably (by up to a factor of 6).

You say

The "Bern" model has been tested by comparing the calculated carbon dioxide profiles with the observed oceanic profiles of C-14 following nuclear bomb testing. Unfortunately this test is not relevant for testing the calculation of the net sink flow of anthropogenic carbon by the model because the influence of recently increased partial pressure of atmospheric carbon dioxide cannot be checked in that way.

The mixer influences the bulk efficiently by strong primary mixing to only about 3 % of the total depth (the backmixed layer). Due to secondary currents the concentration gradients cause diffusional mass transfer from the saturated (equilibrated) surface water to the huge liquid bulk of absorbent salt solution.

The latter should not be stated because at last you end up with a sink flux proportional to the concentration increment quite identical to mine (apart from your partitioning Fem):

Fsi = 0.10*Fem + 0.05* (Cair - 280 ppm)

This does not conform with the deep sea diffusion approach used by Oeschger, Siegenthaler, Joos and Sarmiento. It has to be explained why the diffusional CO2 flux is restricted to the boundary layer. You should make clear that by vertical eddy mixing a similar CO2 distribution can be observed (but only in some parts of the oceans) as if there was diffusion, but there exists indeed neither a diffusive mass transfer nor is a CO2 gradient necessary for eddy transport which would result in some arbitrary CO2 gradient. If cause and effect is mixed up, only a present situation can be explained, but such an approach (like IPCC's, compare the PhD thesis at http://web.mit.edu/globalchange/www/rpt39.html) can principally not model ocean uptake and predict future CO2 concentration.

Best regards, Peter

Subject: Re: Re daly/co2-conc/ahl-co2.htm
Date: Sun, 15 Aug 1999 11:18:24 +0300
From: "Jarl Ahlbeck" <jarl.ahlbeck@abo.fi>
To: <091335371@t-online.de>

Dear Peter,

think about the IPCC model like a thin (abt 100 m) backmixed surface + a complete stationary bulk under it, thermodynamically behaving like a breakfast egg. To describe the heat- and mass diffusion in the non-mixed egg, the second order differential equation for transient heat flow or the corresponding Fick's second law of diffusion is used by the IPCC (and MIT report) because the transient flow into the deep is assumed to create an increase of the oppositing concentration at every level.

The Westerlund calculation shows, however, that an equivalent (calculated from the chemistry and the sink flow rate) backmixed layer of 350 is needed to explain the ocean sink chemically, or a quite thick layer. This is not a real backmixed layer, but only an adapted number. Because we know that the real backmixed layer is not that thick, the question is therefore what is the mechanism for transporting the carbon from the thin backmixed surface to the depth ? As the IPCC model needs an assumption of no secondary currents in the region 100-350 m (about) and ideal transient diffusion, I prefer the Fick's first law of diffusion which should be reliable if the surface water "sees" the bulk quite directly in the contact between the secondary currents and the bulk. Westerlund has used the information of oceanic secondary currents and he came up to the opinion that they are large enough to create a non-complete backmixing of more than three times the complete backmixed layer. If it is so, the Fick's second law of diffusion is not valid because no mechanical mixing is allowed in this law. It should be the first law that I (and you) have used.

The most important thing is that the real measured data treated by regression analysis do not support the use of the Fick's second law of diffusion. See also Hartwig's kinetic equation where he shows that we do not need any information about the surface layer and the gas-liquid mass transfer rate if we state that the absorption by the bulk should be the rate determining step. The difference between the atmosphere and the bulk should be the driving difference if we, as Hartwig, use the Fick's first law of diffusion. I think we understand the physics in the same way but we only use a different approach.

regards, Jarl

Subject: Re: daly/co2-conc/reviews.htm
Date: Wed, 18 Aug 1999 08:13:04 +0200
From: 091335371-0001@t-online.de (P. Dietze)
To: Courtney <richard@courtney01.cix.co.uk>
CC: daly@vision.net.au, jarl.ahlbeck@abo.fi, Onar Åm <onar@netpower.no>, HVolz@t-online.de

Dear Richard,

you wrote:

Your model incorporates effects of biology and, therefore, is a significant advance from the model of Dietze that has successfully demonstrated an ability to emulate empirical reality. I note the important fact that your model is validated by its ability to emulate the Mauna Loa data.

It is not true that I have not incorporated the biosphere (see my Fig.3 of http://www.wuerzburg.de/mm-physik/klima/cmodel.htm where my numerical test in Tab.1 emulates Mauna Loa - Jarl did not directly show such a test). You should be more careful and not use confusing statements.

In linear first order approximation I modelled (in the same way as Jarl) the biosphere to react similar to the ocean and this is why the uptakes are simply put together. Not sheer 'empirical reality' but the physics/chemistry behind my model, i.e. the uptake proportional to the atmospheric concentration excess, is just what Jarl has worked out more in detail. That is why (apart from his correct small buffer term with Fem for the mixed layer) Jarl came to a similar result:

Fsi = - 14.0 + 0.10*Fem + 0.05* Cair or in my case

Fsi = 0.0514* (Cair - 280 ppm) i.e. 3.6 GtC/a for 70 ppm

Regards, Peter

Subject: Re: daly/co2-conc/reviews.htm
Date: Thu, 19 Aug 1999 09:02:47 +0100 (BST)
From: richard@courtney01.cix.co.uk (COURTNEY)
To: 091335371@t-online.de

Dear Peter:

I am sorry if you are offended by my comment, but I stand by it.

Firstly, my comment contains a statement of positive support for your model; viz., "the model of Dietze that has successfully demonstrated an ability to emulate empirical reality".

Secondly, you say that you modelled "the biosphere to react similar to the ocean and this is why the uptakes are simply put together". With respect, this means that your model assumes a single CO2 sink mechanism, and you calculate that mechanism to act in accordance with Hooke's law. Ahlbeck's model considers two sink mechanisms (i.e. a physical mechanism and a biological mechanism) and this permits his model of each of these mechanisms to be amended as understanding of these mechanisms improves. Hence, I think I am correct and accurate when I observe that Ahlbeck's "model incorporates effects of biology and, therefore, is a significant advance from the model of Dietze".

Thirdly, I do not think a comment is a "confusing statement" when it is correct and accurate.

Finally, and very importantly, I do not agree that the models of Dietze and Ahlbeck are emulations of the mechanisms that govern CO2 accumulation in the atmosphere, but I do consider these models to be very valuable tools to aid our understanding of the past and present (perhaps future) CO2 accumulation. I have repeatedly observed that these models fail to replicate seasonal fluctuation of atmospheric CO2 concentration. The CO2 accumulation is the integral of the seasonal fluctuations, and therefore a true model of the input and sink mechanisms would predict the seasonal variations: the models of Dietze and Ahlbeck don't predict the seasonal variations (and nor do any other models).

Please note that I have great respect for the models of Dietze and Ahlbeck. They are the best tools we have for considering atmospheric CO2 accumulation. But respect does not require blind acceptance. It requires questioning and doubt to determine possibilities of improvement.

All the best

Richard

Subject: Re: Carbon Dioxide Date:
Thu, 19 Aug 1999 09:19:11 +0100 (BST)
From: richard@courtney01.cix.co.uk (COURTNEY)
CC: "Chris deFreitas" <c.defreitas@auckland.ac.nz>, "Fred Singer" <ssinger1@gmu.edu>, "John Daly" <daly@vision.net.au>, "Martin Manning" <m.manning@niwa.cri.nz>, "Pat Michaels" <pjm8x@wreck.evsc.virginia.edu>, "Peter Dietze" <091335371-0001@t-online.de>, "Peter Toynbee" <petertoynbee@paradise.net.nz>, "Wayne Hennessy" <wwh@wgtnserv.crl.co.nz>

Dear Vincent:

Thankyou for your comments on 'IPCC CO2 models' that I wholeheartedly agree, especially your conclusion that "The models have a spurious air of accuracy, as they combine an apparently scientific treatment of ocean chemistry with completely wild guesses of carbon absorption by the various regions of the biosphere." I consider this problem you cite to be true of all 'CO2 models'.

I have very recently had the following correspondence with Peter Dietze, and my comment in the paragraph begining "Finally, ..." is pertinent.

All the best

Richard

Subject: SV: daly/co2-conc/reviews.htm
Date: Thu, 19 Aug 1999 10:26:46 +0200
From: "Onar Åm" <onar@netpower.no>
To: <091335371@t-online.de>, "COURTNEY" <richard@courtney01.cix.co.uk>
CC: <daly@vision.net.au>, <jarl.ahlbeck@abo.fi>, <HVolz@t-online.de>

I have repeatedly observed that these models fail to replicate seasonal fluctuation of atmospheric CO2 concentration. The CO2 accumulation is the integral of the seasonal fluctuations, and therefore a true model of the input and sink mechanisms would predict the seasonal variations: the models of Dietze and Ahlbeck don't predict the seasonal variations (and nor do any other models).

This is an odd statement. The seasonal variation is just the biological growth in response to the winter-summer oscillation. Why is that important to model when considering long term change in the CO2 concentration? Jarl's and Peter's models both have a lowest resolution of 12 months, i.e. their model is predicting the CO2 signal smoothed with a 12 month running mean. I don't see the problem with that. It could equally well be argued that they don't have a "true" model since they don't model day to day or minute to minute or millenium to millenium variations, or even regional differences in CO2. But to me that's like saying that we don't have a "true" model of thermodynamics because we don't model the motion of individual atoms. Sometimes resolution is very important (such as e.g. predictions of the aerosol cooling effect), but in this case I cannot see why seasonal variations would matter.

Onar.

Subject: Re: Carbon Dioxide
Date: Thu, 19 Aug 1999 13:01:34 +0200
From: 091335371-0001@t-online.de (P. Dietze)

Dear Vincent,

I have the impression, you consider only the emissions from combustion of fossil fuels, cement etc. for IS92a, whereas Jarl's numbers seem rather to be the total amount (compare his Fig.1 at http://www.john-daly.com/ipcc-co2/ipcc-co2.htm and my Tab.1 at http://www.wuerzburg.de/mm-physik/klima/cmodel.htm). Of course, the community's airborne fraction is related to the fossil emissions, but it makes more sense to consider the total amount.

There is no basis for a figure of 0.5, and not much of an argument for its continuing decline, or for the extrapolations in your paper, as it depends on what happens in the biosphere in the future, which, I contend, is at present unpredictable.

You ignore that Jarl has properly shown in that Fig.1 how well IPCC assumes about 50% remaining in the air over the next century. Looks as if you have not understood the essentials of the carbon model and you still reject it, being mixed up with observations/extrapolations. Because of natural variations (e.g. El Niño, solar), you cannot consider annual values, neither can you extrapolate over several decades. You forget that Jarl's model is significantly backed by statistic analysis of smoothed values (see http://www.john-daly.com/ahlbeck.htm). I admit, the ocean-like biomass representation is poor and in reality an overreaction on CO2 increment and growth can be expected for some decades and later more decay. I wrote to you about the biomass paper of Goudriaan and Kettler in Climate Change 6, 167-192 (1984) and the carbon modelling of Goudriaan in Vegetatio 104/105, 329-337 (1993). But Jarl's (and mine) chosen simple linear biomass uptake approach seems not that wrong, at least it is on the safe side for the future (i.e. not too high).

There should be an effort to set up measurement stations above the various land areas to clarify what is happening. Until we know more about land absorption (and emission) I regard the various ocean models as rather futile as they give us no guide to future atmospheric carbon dioxide concentrations.

Here I do absolutely disagree. Like IPCC you are not aware that even if we knew exactly all present source and sink areas (and fluxes, how they vary with seasons, rain, wind and sunshine), this will hardly help us to model the future concentration - if we lack a dynamic global model based on good physics and chemistry. You indeed support the way IPCC is wiggling about to collect vast amounts of hardly relevant data and wasting money without coming to realistic projections. Compare the identification efforts and suggestions of Heimann in his German paper http://www.mpimet.mpg.de/deutsch/forsch1999.html.

Regards, Peter

Subject: Re: SV: daly/co2-conc/reviews.htm
Date: Thu, 19 Aug 1999 16:09:55 +0100 (BST)
From: richard@courtney01.cix.co.uk (COURTNEY)
To: "Onar Åm" <onar@netpower.no>

Dear Onar:

My point is simple. The model method of Ahlbeck and Dietze assumes inputs and outputs of CO2 to the atmosphere. If their models are correct then the models should ascribe true values to the total inputs and the total outputs. The seasonal variation of CO2 is several times larger than the annual accumulation. This indicates that during part of each year the CO2 sinks are sufficient to collect all the annual input. But the annual accumulation shows an increase for most years. These facts are not compatible with the model method.

All the best

Richard

Subject: Carbon Dioxide
Date: Thu, 19 Aug 1999 10:16:39 +1200
To: Jarl Ahlbeck <jarl.ahlbeck@abo.fl>

Dear Jarl

I have just been catching up with your various papers on Daly's site. for which many congratulations. I have the following comments.

I was surprised at your discovery that the IPCC "Bern" model simply applies a 50% airborne fraction. for its CO2 projections. Note that the airborne fraction is the ratio of the carbon dioxide increase in the atmosphere to the emissions from combustion of fossil fuels, cement etc over the same period. I have no truck with "deforestation" emissions (see below)

The IPCC have been extremely coy about their CO2 projections. I know of no publication which lists the actual figures, so that if you want them, you are obliged to try and get them off the extremely small diagrams (Figure 5b, page 23, and Figure 2.3 page 83) in "Climate Change 95".

Better detail up to 2030 is to be found in an IPCC Technical Paper No 4 "Implications of Proposed CO2 Emissions Limitations" by Tom Wigley; which gives a graph up to the year 2030 with fairly clear points on it from which it is possible to calculate fairly accurately the figures for 1990, 2000 and 2025. The later figures are the best I can do from Figure 5b, page 23 of Climate Change 95. Note that the IPCC assumed an atmospheric CO2 concentration increment for 1990 of 1.8ppmv, as against an actual measured value of 1.29ppmv, and an average over a period of around 1.4. The following table applies for IPCC scenario IS92a

 Year Assumed Emission by IS92a GtC CO2 if all in atmosphere, ppmv IPCC Assumed CO2 ppmv Airborne Fraction 1990 6.0 2.82 1.80 0.63 2000 7.0 3.30 1.64 0.50 2025 10.7 5.04 2.54 0.50 2050 13.2 6.22 3.25 0.52 2100 19.8 9.33 4.1 0.44

So, I do not think it is true that the IPCC model gives a uniform airborne fraction of 0.5. Certainly it is not deliberate.The models have a spurious air of accuracy, as they combine an apparently scientific treatment of ocean chemistry with completely wild guesses of carbon absorption by the various regions of the biosphere. The uncertainties that should be attached to the forecasts are simply horrendous.

In my paper The Missing Sink on http://www.john-daly.com/bulletin/bull-111/bul111.htm I showed that the airborne fraction, since 1960, has fluctuated between 0.25 and 1.0 on an annual basis, and, when fitted with a second order polynomial has increased from 0.56 in 1960 to 0.6 in 1976, then declining to 0.52 in 1994. It has fallen further in 1995 and 1996. There is no basis for a figure of 0.5, and not much of an argument for its continuing decline, or for the extrapolations in your paper, as it depends on what happens in the biosphere in the future, which, I contend, is at present unpredictable.

I would like to upbraid you for continuing to propagate the myth of "deforestation". Although it was obvious from early ocean absorption models that there must be net absorption of carbon dioxide by the biosphere, the 1990 IPCC report, intent on claiming that everything "anthropogenic" must be bad, tried to pretend that there was a net emission from the biosphere in the form of "deforestation", thus creating the existence of a "missing sink". It is possible that there are some land areas with net emissions, but this is also true of some parts of the ocean. We do not try to isolate "emitting" ocean regions.

The biosphere clearly is a net absorber of carbon dioxide. There is no point trying to argue whether this is "anthropogenic" or not. Most of the land surface is subject to human influence, as it is also subject to influence by every other organism. The extent of its interaction with atmospheric carbon dioxide is complex and highly regional, and, surely, virtually unknown. The fact that the IPCC could make a major error in assuming a net emission ("deforestation") when the land is obviously a net sink, displays the extent of our ignorance. Atmospheric carbon dioxide measurements are almost invariably made in places remote from the land, so we have almost no information of concentrations above the different land localities; industrial cities, forests, pasture, wetlands, etc. There should be an effort to set up measurement stations above the various land areas to clarify what is happening. Until we know more about land absorption (and emission) I regard the various ocean models as rather futile as they give us no guide to future atmospheric carbon dioxide concentrations

Vincent Gray
Crofton Downs
Wellington 6004
New Zealand
Phone/Fax 064 4 4795939

Subject: Re: Carbon Dioxide: Clarification
Date: Sat, 21 Aug 1999 10:45:09 +1200
To: Jarl Ahlbeck <jarl.ahlbeck@abo.fl>

Dear Jarl & Co

Further comments on the airborne fraction and IPCC. I still insist that the only useful definition of the airborne fraction is the ratio of the atmospheric increase in carbon to the amount emitted by combustion of fossil fuels and cement, both of which are quantities fairly accurately measurable, so that the ratio has some meaning. The addition of "deforestation" is a response to strictly political pressure from environmental activists, plus the delusion that there is a net "anthropogenic" emission from the earth. "Deforestation" is a vague term whose boundaries are undefined, and may or may not include "regrowth. Its estimates are highly inaccurate, so it gives the "airborne fraction" a similar inaccuracy.

In the Table of my last Email I gave figures from IPCC IS92a which did not include "cement" emissions. I have also checked the IS92a proposed CO2 levels and amended some of them.. Below I give figures for the airborne fraction when it is calculated as the ratio of the total IPCC (fossil fuel and cement) assumed emissions; and also a second set which includes "deforestation". Note that their "deforestation" figures have a deduction for "regrowth"

IPCC IS92a Scenario

 Year Emissions FF/Cem GtC/yr Emissions FF/Cem/Defor GtC/yr Carbon Increase in atmosphere GtC/yr Airborne Fraction FF/Cem Airborne Fraction FF/Cem/Defor 1990 6.2 7.4 3.31 0.53 0.45 2000 7.2 8.4 4.22 0.59 0.50 2025 11.1 12.2 6.01 0.54 0.43 2500 13.7 14.5 7.02 0.51 0.48 2100 20.4 20.3 10.73 0.53 0.53

As you can see, it is not true that the IPCC IS92a model amounts to an airborne fraction of 0.5, whichever way it is measured.. Their assumptions are also not a plausible extrapolation of existing trends in the airborne fraction; again, however measured.

Best Regards

Vincent Gray

Subject: SV: SV: daly/co2-conc/reviews.htm
Date: Sat, 21 Aug 1999 07:26:19 +0200
From: "Onar Åm" <onar@netpower.no>
To: "COURTNEY" <richard@courtney01.cix.co.uk>

> Dear Onar:

> My point is simple. The model method of Ahlbeck and Dietze assumes inputs and outputs of CO2 to the atmosphere. If their models are correct then the models should ascribe true values to the total inputs and the total outputs.

Yes, and they do -- within the range of frequencies specified by their models. Again, why on earth would they care about a strongly cyclical input/output signal that is well-understood and has no bearing on the accumulation of CO2 in the atmosphere?

The seasonal variation of CO2 is several times larger than the annual accumulation. This indicates that during part of each year the CO2 sinks >are sufficient to collect all the annual input.

Yes, your point being? That plants grow a lot in the growing season?

But the annual accumulation shows an increase for most years. These facts are not compatible with the model method.

I still don't understand. Why are these not compatible?

Onar.

Subject: Re Carbon Dioxide
Date: Sat, 21 Aug 1999 10:53:47 +0200
From: 091335371-0001@t-online.de (P. Dietze)

Dear Vincent,

Here you state correctly

Bulletin No 106.. pointed out that for the 25 years from 1971 to 1996 a 54% increase in fossil fuel emissions had had no apparent influence on the approximately constant annual carbon dioxide increment in the atmosphere

Fig.1.. shows a decline in the airborne fraction from 1971 to 1995 from 0.58 to 0.50, a reduction of 13%

>[in Fig.3] a second order polynomial trend line has been drawn. It shows that the emissions absorbed by the earth have increased from 1.2GtCarbon in 1960 to 3.1GtC in 1995. The rate is accelerating.

The trend lines that you show here, distinctly back Jarl's and my carbon models. I do not understand why you now deny all that with your rather destructive comments (see on 19 Aug and my response at http://www.john-daly.com/co2-conc/reviews.htm) like

I do not think it is true that the IPCC model gives a uniform airborne fraction of 0.5

There is no basis for a figure of 0.5, and not much of an argument for its continuing decline, or for the extrapolations in your paper

I am disappointed that there is so much negative response from our friends re the carbon models, giving the impression (especially to IPCC people that we try to convince that we are right) that our models are incorrect - mostly from misunderstanding the essential concept and purpose - and not from finding real bugs. Supposed, Jarl and/or I would prepare a publication in 'Climate Research' and Chris de Freitas would ask you and Richard for review, our model will not even be printed. Though it quite well emulates the (smoothed) Mauna Loa curve, and being based on sound physics and chemistry, it should be useful for future projections.

And Richard should rather build up a good hypothesis about the origin of the seasonal CO2 swing (and its very unevenly distribution over the globe) instead of giving the impression, our models are incorrect. As already shown by Onar and been broadly discussed at Daly's with Richard and Steve Hemphill round about my lake model, the models have nothing to do with seasonal swings as we model the CO2 concentration trend for the next century. Richard said

If their models are correct then the models should ascribe true values to the total inputs and the total outputs

If smoothing and trending is no more allowed, we can forget all about the global warming debate!! Our critics seem not to realize what means "true". Take the temperatures for example. Are this the instantaneous local values in all locations around the globe or the daily, weekly, monthly or yearly or decadal averages or the running means (with arbitrarily chosen time span) or the linear or polynomial trend (within a chosen interval - see Santer), regionally, global, on ground or from satellites??? Though all may be 'true', all look different and on all you can draw different conclusions from applying different (more or less meaningful or erroneous) analyses and hypotheses.

Re CO2, for example once a (nonserious) contrarian said, global warming is rubbish because CO2 shows a reverse correlation: In the summer it is warm & low CO2, in the winter it is cold & high CO2. This is true, but still nonsense.

Dear Jarl & Co.. I still insist that the only useful definition of the airborne fraction is the ratio of the atmospheric increase in carbon to the amount emitted by combustion of fossil fuels and cement, both of which are quantities fairly accurately measurable, so that the ratio has some meaning

I do not agree to this. What we (as carbon modellers) are interested in is the question: Supposed, we consider the averaged equilibrium natural conditions as our base - and then emit varying amounts of additional CO2 - how much of this will the system sequester, how will the atmospheric CO2 increase and how should the dynamic model physics look like?

So all additional CO2 (even that from forest clearing!) has to be coped with (though I agree, it is not easy to cleanly define what all is caused "anthropogenic"). It was just the problem of IPCC modellers, that they first only coped with fossil emissions in the models and thus got into errors, mismatches and missing sinks.

To make my position more clear, let me assume, there is another folk of highly intelligent climate concerned monkeys living in another corner of the world, who just consider us humans as part of the natural equilibrium. Let me further assume, the present 'true' total yearly airborne fraction may be 60%. Now they emit CO2 and find that some 110% (of theirs) "accumulates" in the atmosphere, and this value wildly swinging up and down. Trying to develop a carbon model based on that fractional approach, they would get crazy.

Just as a statistical logging figure you can certainly use IPCC's fossil airborne fraction, but for carbon modelling and prediction purposes it is a far too shaky and improper parameter. The small variations from the 0.5 airborne fraction that you listed, rather confirm than contradict Jarl's important findings re IPCC's modelled airborne fraction for the next century (see his Fig.1 at www.john-daly.com/ipcc/ipcc-co2.htm). Note, that he used a linearized IS92a scenario.

Regards, Peter

Subject: Re: Re Carbon Dioxide
Date: Sun, 22 Aug 1999 09:04:27 +1200
To: <091335371@t-online.de> "Peter Dietze"

Dear Peter

Sorry to be so negative, but I do not see how you can have a viable carbon cycle model unless a lot more is known about the interaction of carbon dioxide with the biosphere and a better idea of how it is likely to change in the future

I have shown that Dr Ahlbeck's contention that IPCC CO2 concentrations are related to their emission curves by a constant airborne fraction of 0.5 is not true. It turns ouit to be absurd if you work out the results for IS92c, for which the airborne fraction for the year 2100 (excluding "deforestation") is 0.09, and with "deforestation" (for which they give a negative value of 0.2GtC) it is 0.1.

I still say that "deforestation" confuses the issue, and should be considered as part of an overall appraisal of biosphere/CO2 interaction.Without "deforestation", the airborne fraction is a fairly well-defined quantity, and since 1960 it fluctuates (up to 1.02 in 1973, and down to 0.25 in 1993) around a constant value of 0.57. My second order polynomial was from Excel, but two more points for 1995 and 1996 make it unnecessary. By all means extrapolate it for a few more years, but can you be confident that the extrapolation will be valid intil 2100?

I am heavily critical of the IS92 scenarios, ( see my paper in "Climate Research" 1998, Vol 10 pages 155-162), but they do show how the airborne fraction can be greatly influenced by almost all the factors they consider; population increase, amount and type of energy usage, economic development, how we treat our forests and agriculture, how crazy we are about global warming, etc. IS92a can be completely discredited as a possible futures scenario since it is so wildly incompatible with what has happened since 1990. I am surprised that any of you can take it seriously.

Sensible economists are unwilling to forecast more than 20 years ahead, and they are usually wrong, even then. Only climate scientists have the nerve to tell us what will happen 100 years from now, and Tom Wigley is prepared to forecast as far ahead as 2300.

Vincent Gray

Subject: Re: Re Carbon Dioxide
Date: Mon, 23 Aug 1999 08:41:08 +1200
To: Onar Åm <onar@netpower.no>, <091335371@t-online.de>

Dear Folks

"Fertilization" is certainly preferable to "deforestation", since it can be looked upon as "beneficial". If "fertilization" is "relatively predictable", then why do you, or the IPCC, not go ahead and predict it, and we might have plausible futures scenarios, instead of the unbelievable set of IS92.

Have you noticed; "emissions" are falling. According to "Nature" 5 August 1999 Vol 400 page 494, the figure for 1998 was 6.32GtC, compared with 6.29 in 1997 and 6.52 GtC in 1996.

Regards

Vincent Gray

Subject: SV: Re Carbon Dioxide
Date: Mon, 23 Aug 1999 05:20:05 +0200
From: "Onar Åm" <onar@netpower.no>

"Fertilization" is certainly preferable to "deforestation", since it can be looked upon as "beneficial".

Beneficial is of course a relative term, but many people would regard larger crops and less sensitivity to drought and heat as benificial, especially people in poor countries who cannot afford expensive watering systems and as a result starve to death.

If "fertilization" is "relatively >predictable", then why do you, or the IPCC, not go ahead and predict it, and we might have plausible futures scenarios, instead of the unbelievable set of IS92.

Actually I think that Jarl and Peter are doing a pretty good job at that. The assumption that the biomass sink is proportional to the CO2 concentration, is a good one as it fits very well with the massive amount of empirical data we have on the CO2 fertilization effect. It is likely that we will never experience that the biosphere becomes carbon saturated.

Have you noticed; "emissions" are falling. According to "Nature" 5 August 1999 Vol 400 page 494, the figure for 1998 was 6.32GtC, compared with 6.29 in 1997 and 6.52 GtC in 1996.

Yes, I also know that 1998 was an exceptionally warm year with low winter heating costs worldwide and that there was a recession in Asia. Both have now ended so expect the emissions to increase.

Onar.

Subject: Re: Re Carbon Dioxide
Date: Mon, 23 Aug 1999 10:20:44 +0300
From: "Jarl Ahlbeck" <jarl.ahlbeck@abo.fi>

Dear Vincent,

Please mail me a copy of your calculation about the IPCC "airborne fraction" for the other emission scenarios than IS92a !

Are we talking about the same parameter and the same way of calculating it ? You cannot see the implications of IPCC-modeling by picking one year at a time from their graphs, you must smooth the emission curve and the concentration curve by regression analysis (linear or parabolic) and then see what value of the "airborne fraction" can explain the connection by using the integrated mass balance. For nonlinear emission curves, this is not a trivial calculation.

For a linear emission scenario curve, a constant airborne fraction (backmixing) gives a parabolic concentration curve, just as in the IPCC curve for IS92a.

I am absolutely sure that the IPCC model implies a close to constant value of "my" airborne fraction that is Concentration Time Mean Derivative * 2.123 (as a function of time) divided by smoothed total emission rate (from the scenario) (as a function of time).

Something can also be seen from the structure of the equations that they use: Their ocean.surface + biosphere is kind of backmixed (that means theoretically constant airborne fraction). The contribution of the deep ocean sink is probably quite close to "constant airborne fraction" in their model too as they use the Fick's second law of diffusion for transient mass or heat transfer (the well-known second-order differential equation that can be used for calculation of the temperature profile in a boiling breakfast egg). If the sink is modeled in this way, the influence of the atmospheric partial pressure is much smaller than when using the first-order approach (Fick's first law of diffusion) and assuming that the backpressure of the bulk is created by the pre-industrial bulk concentration that is assumed to be constant (Dietze and me). Our simplification of the back-pressure is rough, but not more rough than the IPCC diffusion model, which is extremely rough too.

Even if the IPCC modeling of the sink to the ocean bulk as a second-order diffusion process would be correct, their modeling of the biosphere as "backmixed" is so over-simplified that it is stupid. I have never said anything about the proportion between the biosphere sink and the ocean bulk sink, only stated that a stringent data analysis show that the net sink flow rate, in contradiction to the IPCC model, is strongly dependent on the atmospheric partial pressure. Do you agree or not ? If not, what does your data analysis show ?

The "airborne fraction" is not anyway a reliable measure of the carbon cycle that should be described in terms of flow rates and accumulation rates only and how these flows are dependent on different parameters.

regards, Jarl

Subject: Re: Re Carbon Dioxide & reviews.htm
Date: Mon, 23 Aug 1999 10:08:18 +0200
From: 091335371-0001@t-online.de (P. Dietze) Reply-To: 091335371@t-online.de

Dear Vincent,

you still seem having problems with accounting for fertilization and deforestation. The question is not whether CO2 terms are preferable, beneficial or accurately known or predictable. From the carbon modelling point of view all additional CO2 - in comparison to a supposed undisturbed natural state - has to be allocated to emissions (called 'total' emissions), whereas all CO2 uptaken by the natural system has to be coped for within the model's buffers and sinks (a simple fertilization approach being implemented).

Then applying the model for the past and for example for the total IS92a emissions, we can emulate Mauna Loa and project the resulting future concentration. None of us ever said, IPCC's BAU and reference scenario IS92a is realistic - indeed we emphasized it is exaggerated (burning 1500 GtC whereas ~1300 only are to be expected). The reason that we use IS92a, is to get results demonstrating the deviations against IPCC and to show that even if we would do nothing about reduction and burn even more carbon than realistic, no 'dangerous', but rather a beneficial CO2 concentration and temperature change would occur. Mine is ~570 ppm in 2100 and ~0.55°C against preindustrial, so we would not even reach the medieval optimum.

Best regards, Peter

Subject: CO2 models
Date: Mon, 23 Aug 1999 17:09:20 +0100 (BST)
From: halley@courtney01.cix.co.uk (COURTNEY)
To: "Onar Åm" <onar@netpower.no> CC: <jarl.ahlbeck@abo.fi>, <daly@vision.net.au>, <HVolz@t-online.de>, <OWildgruber@csi.com>, <ssinger1@gmu.edu>, <h.heuseler@businessnet.de>, <Heinz.Hug@t-online.de>, <c.defreitas@auckland.ac.nz>, <petertoynbee@paradise.net.nz>, "Wayne Hennessy" <wwh@wgtnserv.crl.co.nz>, <m.manning@niwa.cri.nz>, "Pat Michaels" <pjm8x@wreck.evsc.virginia.edu>

Dear Friends:

Please forgive me not replying until now but I have had other priorities. I hope you don't mind that I intend to respond to both Onar and Vincent in this email and not individually. Also, I don't think Peter's comments to me call for a response; is that OK, Peter ?

Onar, you query my comment that the seasonal variation in atmospheric CO2 accumulation is not compatible with Peter's model and you say that Peter's model works at the temporal frequencies he considers.

Firstly, I agree that Peter's model does work as an emulation of the annual CO2 accumulation. In fact, I proclaim that his model works very well indeed for this (or at least, it does throughout the existing historic record). My point is different and has induced much debate between Peter and myself.

Peter claims that his model emulates the real world mechanisms that determine the CO2 concentration. But his model does not emulate the seasonal effects. The important point is that he needs to explain why there are mechanisms operating at seasonal rates that have no longer term effect when the magnitude of effects observed seasonally is much greater than the magnitude of the modelled effects operating annually. What are these mysterious seasonal mechanisms that do not overcome nor interact with the less powerful modelled effects ?

Please note that I believe the most valuable property of any model is its ability to provide greater understanding. The fact that Peter's model works for annual but not seasonal mechanisms demonstrates that we do not understand the mechanisms governing atmospheric CO2 concentration. In my opinion, this demonstration is a very valuable atribute of Peter's model and should not be ignored. It tells us to consider why the annual accumulation is not swamped by the powerful mechanisms that determine seasonal variation. I have repeatedly said that I would welcome an explanation of how this can be. If Peter's model eventually does induce an answer to this question then his model will have become a major milestone in our understanding of the atmospheric carbon cycle.

There is one caveat to my above argument. As Peter himself points out, atmospheric CO2 concentration varies around the globe. This spatial variability provides uncertainty to the temporal variability used to validate Peter's model.

Vincent is clearly concerned at the inaccuracy inherent in the temporal variability data. And he is concerned that the biological mechanisms affecting the variabilty have effects with very uncertain magnitude. In my opinion, this biological uncertainty is good reason for separating physical and biological effects in any 'CO2 model'. The seperated effects can be independently adjusted to agree with any improved empirical data that becones available.

Vincent, you say, "Have you noticed; "emissions" are falling. According to "Nature" 5 August 1999 Vol 400 page 494, the figure for 1998 was 6.32GtC, compared with 6.29 in 1997 and 6.52 GtC in 1996." Yes, I had noticed and expected this. It seems that the CO2 emissions are decoupling from reported energy use according to an analysis by an oil company. Well, given the Kyoto Protocol, this is not surprising. There is uncertainty in data of energy use and of CO2 emissions. If these two parameters move together they will appear to seperate when most countries want to report the lower limits of their emission estimates. This reporting problem also increases the uncertainty of annual CO2 emission data to be modelled.

All the best

Richard

Subject: Re: Re Carbon Dioxide
Date: Tue, 24 Aug 1999 11:03:34 +1200
To: "Jarl Ahlbeck" <jarl.ahlbeck@abo.fi>, <091335371@t-online.de>

-Dear Jarl

I attach my calculations for the airborne fraction for the IS92 scenarios. I have estimated the atmospheric increment of CO2 for each year for which they assume emissions from the slope of the curves of atmospheric CO2 concentrations for the varios scenarios. Scenarios. IS92a, IS92c and IS92e are fairly accurate, as there are good curves published in IPCC Technical Paper No 4 "Implications of Proposed CO2 Emissions Limitations" by Tom Wigley. For IS92b IS92d and IS92f I have been reduced to the extremely small graph on page 23 of "Climate Change 95".

There are no two ways about it. The "airborne fraction' is a completely arbitrary quantity whose future value can be varied as much as you like by your assumptions on economic development, energy usage and land usage.

The comparatively steady value over the past 40 years is probably an artifact, and there is no reason why it should continue. It is true that IS92a gives projections which add up to a fairly constant value for the airborne fraction, but that is a sheer accident. The IS92a scenario is nonsense, anyway and I cannot understand why you still continue to take it seriously. If you want a better scenario why not choose IS92c which does at least sljghtly resemble what is going on in the climate. But you wont like it because the airborne fraction varies.

The basic problem is the interaction of CO2 with the biosphere. You grossly exaggerate the ease with which this may be guessed. May I draw your attention to the paper by R.A. Houghton, J.L. Hackler and K.T. Lawrence in "Science" 23 July 1999, Vol 285 pages 574-578 which does a fairly honest job of estimating the net carbon emission/ sink value of the USA from the year 1700, but gives hardly any guidance on how it is likely to change in the future. Similar exercises need to be done for all land areas. but even then, how do we know what will happen in 20 years' time? I still say that arguing over the validity of the various ocean carbon models will not solve the basic problem of where we are going.

Vincent Gray

Subject: SV: CO2 models
Date: Tue, 24 Aug 1999 03:01:25 +0200
From: "Onar Åm" <onar@netpower.no>
To: "COURTNEY" <halley@courtney01.cix.co.uk>

Peter claims that his model emulates the real world mechanisms that determine the CO2 concentration. But his model does not emulate the seasonal effects.

Well, Newtonian physics claims to be a pretty good model of the dynamics of physical bodies, but it does not model the behavior of individual atoms. This is in fact true for ALL models of reality, i.e. no scientific theory exhausts reality. This does not, however, make them inferior models. For the limited area in which they fit reality they are indeed good models of reality.

The important point is that he needs to explain why there are mechanisms operating at seasonal rates that have no longer term effect when the magnitude of effects observed seasonally is much greater than the magnitude of the modelled effects operating annually.

But this is not a mystery that needs to be solved. It's a well-understood effect that is even taught in primary school.

What are these mysterious seasonal mechanisms that do not overcome nor interact with the less powerful modelled effects ?

Now you are confusing me. You obviously know why the CO2 concentration fluctuates seasonally (you must since this is common knowledge) and yet you present this as a mystery. Why?

Please note that I believe the most valuable property of any model is its ability to provide greater understanding. The fact that Peter's model works for annual but not seasonal mechanisms demonstrates that we do not understand the mechanisms governing atmospheric CO2 concentration.

???? Again you present this as a mystery that no-one understands. Peter's (and Jarl's) model doesn't "work" for seasonal fluctuations because it has been deliberately designed not to "work" for it because it is an unintersting signal. It is analogous to the fact that winter and summer temperatures are filtered out of the global temperature signal and instead we are given deviations from normal.

In my opinion, this demonstration is a very valuable atribute of Peter's model and should not be ignored. It tells us to consider why the annual accumulation is not swamped by the powerful mechanisms that determine seasonal variation.

Please explain more clearly what you are getting at. I don't understand what the problem is. This is like asking why the annual accumulation of heat in the atmosphere (due to global warming) is not swamped by the powerful mechanisms that determine seasonal variation in insolation. It doesn't make sense.

Onar.

Subject: Re: Re Carbon Dioxide
Date: Tue, 24 Aug 1999 09:19:30 +0300
From: "Jarl Ahlbeck" <jarl.ahlbeck@abo.fi>

Thank you Vincent for your message and the attached tables !

First of all, you should not misunderstand me, I really don't prefer any scenario. The reason why I used IS92a as a demonstration of 1) The IPCC model 2) The alternative models by me (and Dietze), is only the fact that this scenario is used in the local debate here in Scandinavia as the "most probable scenario".

When I look at your tables, I see that the mean value of the airborne fraction for every scenario in the range between 10 and 36 GtC/a for the year 2100 and for any time interval, as calculated by you from the IPCC report, is 0.5133 with a standard deviation of only 0.1 or just as I have stated.

The only clear exception is one value, or the value for the year 2100 and scenario IS92c that is 0.08 when the emissions have beed radically reduced to 4.9 GtC/a. For such a radical reduction of the emissions, even the IPCC ocean second-order diffusion model seems respond to the increased partial pressure to some extent because this low value is possible if the sink flow continues high due to the increased partial pressure and the emissions are reduced at the same time.

My model gives a negative value of the airborne fraction for a situation like that, the sink flow rate amplified by the increased partial pressure from recent emissions is much larger than the rapidly decreased emissions.

I never said that the IPCC model implies an exactly constant value of the airborne fraction, I said a CLOSE TO constant value.

I am glad to see that your calculations confirm my findings. However, the airborne fraction is not a good measure of the carbon cycle.

regards, Jarl

Subject: Re: SV: CO2 models
Date: Tue, 24 Aug 1999 09:32:19 +0100 (BST)
From: richard@courtney01.cix.co.uk (COURTNEY)
To: "Onar Åm" <onar@netpower.no>

Dear Onar:

School textbooks may mention some mechanisms of seasonal variation to atmospheric CO2 concentration, but so what ? Please see the continuing discussion between Vincent Gray and Jarl Ahlbeck to gain an understanding of the extreme doubts concerning these mechanisms.

My point is not whether Peter's model works; I have repeatedly stressed that I agree it does. I am questioning Peter's claim that his model emulates the behaviours of the actual mechanisms effecting atmospheric CO2 concentration. Peter's model is not expected to "exhaust reality", but Peter claims his model represents the actual physical mechanisms of reality. If it does then it must explain the seasonal mechanisms. As I have already spelled out (on Daly's web site) the 'level of a lake' analogy (that Peter proposed) suggests that the seasonal mechanisms should swamp the annual mechanisms.

Please note that an annual accumulation of CO2 in the air is the net effect of seasonal variation of CO2 in the air during that year. Hence, it is not reasonable to assert that the seasonal variations are "short term" and therefore irrelevant to the longer trends.

The seasonal mechanisms have much greater effect than the mechanisms of the annual accumulation. The annual accumulation is the net effect of the seasonal mechanisms. Therefore, emulation of the actual mechanisms creating annual accumulation must describe the seasonal variation that provides the annual accumulation.

All the best

Richard

Subject: Re co2-conc/reviews.htm
Date: Wed, 25 Aug 1999 10:22:15 +0200
From: 091335371-0001@t-online.de (P. Dietze)
To: Courtney <richard@courtney01.cix.co.uk>

Dear Richard,

I proclaim that his model works very well indeed for this (or at least, it does throughout the existing historic record).. The important point is that he needs to explain why there are mechanisms operating at seasonal rates that have no longer term effect when the magnitude of effects observed seasonally is much greater than the magnitude of the modelled effects operating annually. What are these mysterious seasonal mechanisms that do not overcome nor interact with the less powerful modelled effects?

After all discussion, I was expecting this intelligent objection of yours. I fully agree to your problem: Why are there (unmodelled) processes that fastly absorb even more than the emitted CO2 during the NH growing season and then release most of it again – and still the weaker long-term natural absorption processes (that being only modelled and yielding a 55 yr CO2 lifetime), control the long term sequestration?

These seasonal mechanisms (mostly related to growth and decay of light biomass) are by no way mysterious but well known. You can imagine the function – in terms of the lake model – as a huge piston, sucking water out of the lake into a cylinder and later inject it again. This process is simply superposed to the normal sink functions (for CO2: sequestration in oceans and trees) and does not interact much. I suppose, the cause may be not only the NH growing season, but considerable (phase shifted) temperature and CO2 solubility swings in surface waters as well.

Regards, Peter