Guest Post by Willis Eschenbach
A recent study by Shindell et. al, entitled Improved Attribution of Climate Forcing to Emissions, (Science Magazine, 30 October 2009, Vol. 326) reports on interactions between aerosols and methane and other greenhouse gases. It has been discussed on Watts Up With That here <http://wattsupwiththat.com/2009/10/31/an-idea-i-can-get-behind-regulate-methane-first/>, as well as on other blogs. The Shindell study gives new values for the “radiative forcing” of various greenhouse gases. The “radiative forcing” is the increase in greenhouse radiation which is due to the increases in greenhouse gases since 1750.
UPDATE: The remainder of this article has been removed at the request of the guest author, Willis Eschenbach. During discussion, an error was discovered (see comments) and rather than leave this article with that error in place which may possibly mislead somebody in the future (if they didn’t read through comments) I’m honoring Willis’ request for removal. The comments remain intact. – Anthony
carrot eater (14:26:32), you say:
The issue is that the IPCC earnestly informed us that the effect of methane since 1750 was an instantaneous radiative forcing of 0.5 W/m2 ± 0.1 W/m2.
This means that if there were no methane increase since 1750, the most that it would change the forcing is 0.6 W/m2.
The current study, on the other hand, says that that if there were no methane increase, the least that it would change the forcing is 0.7 W/m2.
Clearly, both of those can’t be right …
“”” Willis Eschenbach (14:47:49) :
George E. Smith (13:40:40) :
I hate to even ask how many stations there are all over the globe that have the necessary equipment to measure the local radiative forcing W/m^2 contributed by each of the major GHGs. Something tells me that without proper mapping of the local values, it would not be possible to compute a (credible) mean global value.
Since you didn’t ask, I believe that the answer is “none”. As far as I know, given our current measuring devices, it can’t be measured from the ground, or from the air, or anywhere. It can only be estimated from a computer model “””
Just based on the extreme range of surface Temperatures on earth from about -90C (183Kat Vostok) to about +60 C (333K in tropical deserts) the bounds for Black Body limited Stefan-Boltzmann thermal radiation cover a range of about 11:1. Throwing in variations in spectral emissivity, may extend or contract that range.
So a given amount of any target GHG over such a site; would presumably yield something like an 11:1 range in forcing for the exact same amount of a given GHG species.
Actually it is worse than that; because as a result of that 183 to 333 K extremes of Temperature, the peak of the BB like spectrum also varies by a factor of 1.82; and the spectral radiant emittance at that peak varies as the fifth power of temperature rather than the fourth. Then you get a 19.95:1 range of peak thermal emittance; and that spectral peak moves even further away from at least the CO2 13.5 to 16.5 micron absorption band. Now that of course results in enhanced cooling past CO2 GHG trapping in the hottest regions, and enhanced trapping by CO2 in the coldest locations where the emission spectral peak sits right on the CO2 band at 15 microns.
Then you also need to factor in that the Doppler broadening of those absorption bands tends to go as the square root of the Temperature (Kelvins), and that will add further to the high temperature absorption, and further depress the low temperature absorption due to the narrower band of the absorption bands.
In any case; radiative forcing hardly seems suitable as a global constant for any species of GHG
Which is why I find the whole Arrhenius concept of “climate sensitivity” to be rather silly.
Algebraic-Hypnotics 101 Primer
Academic statisticians “succeed” by hypnotizing their victims with volumes of algebraic abstraction. This concentrates power in their clan.
Practical consequences:
1) bad management,
2) bad policies,
3) etc., etc., etc. …
… i.e. intolerably costly for individuals, society, & civilization.
Clarification:
Data analysis (& *descriptive stats) is indispensable.
It is misapplied statistical *inference*, based on untenable abstract assumptions, that runs individuals, society, & civilization off-the-rails.
For those who lack the background to identify statistical inference:
Look for “error bars”, “confidence intervals”, “p-values”, “statistically significant”, & related “uncertainty” terms/measures.
No matter how well “mathematically defined” these things are, they are all based on abstract assumptions, which are not necessarily tenable in reality [for example in biological & climate systems].
Conclusions cannot be divorced from assumptions. Due diligence in assessing assumptions is not a part of the mainstream “p-value” “error bar” “confidence interval” culture that has gripped the scientific community as a major multi-decadal fashion.
Here’s an example of how it goes:
Say a medical expert & a stats expert are working together on a timeline…
Assessing abstract assumptions used in inference requires intimate subject-area knowledge. The statistician may not have this and so may decide to “go with it anyway and let the medical expert decide”.
Abstract-assumption assessment also requires intimate understanding of stats models. The medical expert may not have this and so may decide to “go with it anyway and let the statistician decide”.
And so we see a “hot potato” scenario …and keep in mind timelines, reputation, careers, grant money, etc., all of which have the potential to impact researchers families – plenty of motivation for 2 bright sparks to bluff a united front.
Anyone who thinks the solution is as simple as educating interdisciplinary hybrids is failing to appreciate how complex the realm of mathematics & statistics is. The solution is the careful application of common sense & appropriate qualifiers – along with sober, insightfully-penetrating [as opposed to malicious] audit by independent parties lacking conflict-of-interest.
We don’t all have multiple lifetimes to become cutting-edge experts in multiple fields, but that need not imply that our common sense should obediently & uncritically succumb to the wizardry of algebraic-hypnosis.
Willis, You are looking at the wrong graph in the IPCC AR4 report. Figure 2.21 is the emissions-based forcings, not figure 2.20 (which is the abundance based calculation). Jorg is correct – you are not comparing like with like.
Willis: MASSIVE OVERSIGHT!!!! You have missed one key part of the IPCC chapter 2 in the Fourth Assessment Report: namely, pg. 205 Figure 2.21 and pg. 207 Table 2.13 both show the IPCC assessment of an “emission based value” as versus the 0.48 from the “abundance based value”. And the result showed in this figure and table in the IPCC 4th Assessment Report was…
0.856 W/m2!!!
That’s 0.57 for CH4 (some is destroyed by NOx which is why it is larger than 0.48), 0.2 for ozone, 0.07 for stratospheric H2O, and a dab for contribution to CO2 from fossil methane. No sulfate or nitrate interactions estimated, but well within the Shindell uncertainty range for the emission based value. Therefore, in order to make your above post accurate and honest, you need to correct the figure to include the IPCC estimate (or retract the post).
I’m also surprised that given the eagle-eyed scrutiny that posters on this board apply to any claim made by a “consensus” scientist, that not one has read the IPCC forcing chapter sufficiently carefully to notice that this number was in there in plain sight, in multiple places.
Controversial because its heresy
“New data show that the balance between the airborne and the absorbed fraction of carbon dioxide has stayed approximately constant since 1850, despite emissions of carbon dioxide having risen from about 2 billion tons a year in 1850 to 35 billion tons a year now.
This suggests that terrestrial ecosystems and the oceans have a much greater capacity to absorb CO2 than had been previously expected”.
http://heliogenic.blogspot.com/2009/11/controversial-because-its-heresy.html
Oh. I see that both carrot eater and Zimmerman both beat me to the conceptual answer (eg, that Willis is comparing Shindell’s orange to IPCC’s apple), they just missed that IPCC kindly provided an orange as well so that in fact, we can do both apple to apple (abundance) and orange to orange (emission) comparisons.
This isn’t to say that the Shindell study has not advanced the science, which it has, adding new interactions and showing how they might change the calculations of forcing going forward into the future (the IPCC GWP estimate for methane uses very rough approximations for ozone effects). It also doesn’t say that the IPCC error bars were perfect – in general, my baseline assumption is that error bars should probably always be a bit larger than they are. Of course, that baseline means that things might always be worse just as easily as they could be better.
However, in any case, the thrust of this post, which is that the old IPCC result falls outside of the bounds of the new Shindell result, is clearly wrong, as carrot eater and Zimmerman both tried to argue.
Oh, and Freddy beat me to the post too, and he included the actual Figure number. Oh, well. I also apologize for spelling Zimmermann’s name wrong. C’est la vie.
Willis, let me try again.
“Clearly, both of those can’t be right …”
They are both right, or at least, they are consistent with each other. They are coming out of the same model. They are just different ways of describing the same physics.
The lower RF value for methane is calculated from the direct impact of ~1800 ppb of methane itself. Forcing due to stratospheric water and tropospheric ozone are calculated separately, and listed separately. You can see that in both the IPCC report and “Abundance-based” forcings in this paper.
However, increases in stratospheric water and low-level ozone are due, in part, to methane emissions. Methane reacts to form these compounds. So, the authors decided to group the radiative forcings in a different way, to emphasize which emissions lead to which forcings. Hence, methane gets not only its own direct contribution, but parts of the forcings due to stratospheric water and tropospheric ozone.
The two different RF figures are expressing very different things, and thus they are different from each other. They are different by definition. Methane’s direct impact is x, +/- some error. Methane’s total impact is y, +/- some error. x and y will be different things. There is no inconsistency here.
WAG (11:38:57) :
So… we model the reaction of an economy we don’t understand to modelled actions we don’t understand in response to a modelled climate we don’t understand. Hey – what could go wrong?
I’m sure it will all fall somewhere within well-understood bounds (or not)
Marcus: many thanks for that; I totally forgot emissions-based figures are in the IPCC FAR. Yes, there’s a neat little table there on p207, showing the cross-interactions. In particular, the new paper adds interactions between sulphates and other species, beyond what was given at the time of IPCC FAR. But yes, this completes the loop. To really compare oranges to oranges, I’d remove the interactions considered in this paper but not the IPCC report.
Willis,
Of course, since we don’t understand the climate, to me it is somewhat circular to build a model of a climate system we don’t understand and then use the model to try to understand the climate system we don’t understand
Now you’re implying that there is no climate!
No.. really, I think that pretty much sums it all up.. particularly when both Lindzen and Spencer have demonstrated that all the models have water vapor feedback exactly wrong and all model predictions of catastrophic warming are due to water vapor feedback.
Re: Decision making under uncertainty
There’s a number of different methods that have been developed in different places for trying to make ‘optimal’ decisions when faced with uncertainties of different types, but pretty much all of them require some form of objective or utility function. That is, what is the ultimate goal that we’re trying to achieve. This should be set long before any discussion/disagreements about how the world works, and basically involves all parties to lay they cards on the table. This is where I believe everything will break down as I suspect it would be virtually impossible to get any sort of consensus about what humanity’s long-term goals should be amongst 200-odd world governments (of course it would be much easier if we only had 1 😉 ). e.g., status quo in terms of standard of living, getting 90% of the earths population above some magical poverty line, maintain healthy and functioning ecosystems across most of the earths surface (possibly to the ultimate detriment of humanity)?
@ur momisugly Paul Vaughan
I agree that misapplied statistical inference causes problems, but without some form of statistical inference how do you propose that scientific (not statistical) hypotheses of interest can objectively be compared? Where possible, assumptions should always be checked using the observed data so that you can have some faith that your reported conclusion are not too far off the mark. Unless you have a situation that is completely deterministic, with no potential for measurement uncertainty, I don’t know how you can compare hypothesis without some form of statistical procedure. Descriptive statistics can be useful of interesting, but have limited ability to advance hypothesis driven science.
As for your doctor and statistician example, yep, I’m sure that happens all the time and I’ve seen it happen, but that doesn’t mean people need to be an expert in both. It’s basically a communication problem. All fields have their own jargon which creates barriers. The parties need to either bring those barriers down themselves or find someone else to mediate (personally I’ve found beer to be very useful and breaking down such barriers, until I start babbling like a 6 month old). If people are too proud to admit they don’t understand something, then that’s an entirely different problem.
Sorry for the slightly tangential post.
Marcus (16:41:24), outstanding. This is why I love the web. You say:
Marcus, you are quite correct. This study did not say what I thought it said. My comments were 100% wrong.
The study does add something new to the question … it adds 0.08 to the methane emission-based forcing, through methane’s interactions with nitrate and sulfate. Be still, my beating heart.
I was misled by the fact that the study did not mention the IPCC emission-based forcing at all. Instead, they only referenced the abundance based forcings, and they only compared their results to the abundance-based forcings. I don’t know why, as it seems very relevant to the discussion. I was totally fooled by their claim that they had found something significant.
In their Abstract, the authors say:
“Substantially alter”? An addition of 0.08 W/m2 to the total methane forcing, from sulfate and nitrate interactions, doesn’t “substantially alter” anything. I am amazed that this trivial change, in the hundredths of a watt/metre squared, merited a scientific paper. How does this make any difference to anything? When I mistakenly thought that they were saying the methane forcing was doubled, “substantially alter” made sense … but a scientific paper for a 0.08 W/m2 difference? On what planet is this meaningful?
I am also amazed that anyone thinks that current climate models can measure forcings to the nearest hundredth of a W/m2 … when we can’t measure the forcings directly, so we have absolutely no evidence to confirm the model findings.
Finally, you say:
Sorry, Marcus, but that’s your job, to scrutinize things that seem wonky to you … and you have done it well. My congratulations. My job is to scrutinize things that seem wonky to me. I’m not surprised that you didn’t find the errors in the hockeystick, for example.
This is the beauty of the opening of science to a much wider audience on the web, that both of us can look for (and sometimes find) errors that have not been found in the peer review process. This is how science advances, because both sides investigate the claims that they doubt for whatever reason.
My thanks to you,
w.
To me, the most important quote from the IPCC FAR: “Best estimates are given where available.” Estimates? Doesn’t that mean POOMA* numbers?
* Preliminary Order of Magnitude Approximation
Willis, most science is incremental. The end numerical result of the paper might not seem impressive to you, but getting more aerosol chemistry into the models is one more thing they’ve got in there (how well they’ve gotten it there is hard to tell from this paper). Interactions like those between aerosols and clouds are another thing to improve, so aerosols really are of interest. Sadly, it’s just a short note in Science, so hopefully the details of the chemistry will be described elsewhere.
“In particular, methane emissions have a larger impact than that used in current carbon-trading schemes or in the Kyoto Protocol.”
That statement isn’t talking about the historical forcing, but rather the future forcings and therefore GWPs. _Those_ are changed more significantly by the paper (as I said in my first post, the methane chemistry was only included in a very rough fashion in the IPCC GWP estimation of “25”)
carrot eater (19:21:57), you are quite correct when you say:
That’s why I was surprised (and misled) by them saying:
Their result barely makes the “incremental” threshold, it definitely does not “substantially alter the relative importance”.
In particular, I’m not impressed by results of this type from a model. This is less than incremental, it is totally unverifiable. Yes, their model says that the methane forcing increases, but we have no data with which to compare that. How can we conceivably put error bounds on that tiny number? How do we even know that the sign of the change is correct? But I digress …
Thank you for your contributions to the thread. Our discussions here are another incremental step down the scientific path. My appreciation for your fighting my ignorance on this subject, and for your gracious and convivial tone.
w.
Marcus (19:49:41), you say:
After rising for a number of years, the atmospheric methane emissions have levelled off during the last decade. Nobody knows why this has happened. Heck, nobody knows why they were rising prior to levelling off. Nobody knows which way they will go from here.
As a result, any honest scientist would have to say that we don’t have a clue what methane levels will be a century from now.
Yes, if we assume (as the IPCC and Shindell et. al do) that methane will start to rise again at some assumed value, it could make a difference. If their model is correct. And if methane levels rise. And if GHG forcing is the secret global thermostat.
Since we have absolutely no independent confirmation of the accuracy of the model, and no idea what future methane concentrations will be, and no data or observations saying that temperatures are ruled by GHGs, that’s a big stack of “ifs” to be calling science.
Thank you again, Magnus, for your contribution. Rather than merely state your ideas, you backed them up with clear citations to the literature. My hat is off to you.
w.
If you’re looking for significant change in the numbers, it’s probably in the updated GWP of methane. I’m too tired to chase down exactly how those have changed.
“Yes, if we assume (as the IPCC and Shindell et. al do) that methane will start to rise again at some assumed value,”
They do? The IPCC FAR report says, “The reasons for the decrease in the atmospheric CH4 growth rate and the implications for future changes in its
atmospheric burden are not understood.” It goes on to give some discussion. Methane sources and sinks are not so well understood.
There is some danger of more methane release with melting permafrost.
carrot eater (21:13:49) :
The IPCC gives forty different scenarios for the future changes in GHGs. Of these, 36 of them give an estimate of future methane emissions.
In each and every one of these 36 scenarios, methane is shown as increasing starting now. In some of them the emissions continue to rise through 2100, in some of them the rise peaks somewhere between 2040 and 2080.
But in all of them it is shown as rising. So yes, the IPCC does assume that methane will start to rise again, as I stated.
w.
I am happy, the issue could be resolved. Thanks for carrot eater and marcus to further the argument, and especially for marcus, to provide the finally convincing citation.
By the way, “Marcus (16:49:53) : I also apologize for spelling Zimmermann’s name wrong.” No sweat. Having lived a year in the USA, I know how confusing the German spelling is. The really amazing thing is, what can happen to my first name. 🙂
“As a result, any honest scientist would have to say that we don’t have a clue what methane levels will be a century from now.
Yes, if we assume (as the IPCC and Shindell et. al do) that methane will start to rise again at some assumed value, it could make a difference. If their model is correct. And if methane levels rise. And if GHG forcing is the secret global thermostat.”
Shindell’s paper does not actually make any assumptions about the future path of methane growth. The way he calculates a GWP is to assume that we are at an equilibrium methane concentration, and then calculates the radiative forcing (not global temperature) perturbation resulting from emission of an additional ton of methane. This is also how the IPCC calculates GWPs. GWPs are the measure of radiative forcing of a substance relative to that of CO2. The key quote from the paper, as it relates to methane: “The 100-year GWP for methane is ~10% greater (~20 to 40%, including AIE) than earlier estimates (5)”.
I grant you that the IPCC scenarios almost entirely consist of rising methane concentrations. That is a separate discussion, and one that is likely to be revisited in AR5. I will note that the uncertainty there comes less from uncertainty about the chemistry and more from the uncertainty about trying to measure area sources of methane (rice paddies, swamps, etc.) and therefore get an accurate inventory of anthropogenic vs. natural emissions. I will also note that there has actually been a significant amount of methane emission reductions in the U.S., Europe, and Russia, through some combination of landfill capping, reduced gas infrastructure leakage, and better flaring of fugitive emissions. Therefore, to some extent the lower-than-expected methane growth may be, in part, a result of actual environmental policy. (to some extent it is understood natural variation – rainfall and wildfires and such interacting with methane sources, and to some extent it really is unexplained)
Willis, I follow you now. Well, the B1 scenarios result in little increase in the methane concentration, I think. The newish MFR scenario, which may not be in your SRES tables, has lower methane emissions than the others, though it has low emissions for other things as well. But I agree, we should all be watching to see if the plateau in methane concentrations continues.
By the way, to keep things clear, you should be careful with methane ‘concentrations’ and ’emissions’. I see when you point out the concentrations above, you said emissions. After all, the Shindell paper underscores the importance of keeping the two straight.
Willis, will you edit the original article to reflect the subsequent discussion? I sometimes refer back to older articles on WUWT, and I assume others do, too. It saves a lot of time if discussion-based corrections are folded back into the original story.