Back in the day, the IPCC published its 3rd assessment report and declared a high confidence in the science of doom – humans have caused the unprecedented modern warming and the future is very bleak.
In their view the science was well enough understood that climate models could be relied upon – especially as they predicted the past so nicely.
As well as their Summary for Policymakers they also published a Complete Report. It’s a very informative document and runs to just under 800 pages in 14 chapters, not including appendices. Worth a read for the serious student.
Chapter 5 examines the role of aerosols, which they introduced in Chapter 1 (p.93):
The effect of the increasing amount of aerosols on the radiative forcing is complex and not yet well known. The direct effect is the scattering of part of the incoming solar radiation back into space. This causes a negative radiative forcing which may partly, and locally even completely, offset the enhanced greenhouse effect. However, due to their short atmospheric lifetime, the radiative forcing is very inhomogeneous in space and in time.
This complicates their effect on the highly non-linear climate system. Some aerosols, such as soot, absorb solar radiation directly, leading to local heating of the atmosphere, or absorb and emit infrared radiation, adding to the enhanced greenhouse effect.
Aerosols may also affect the number, density and size of cloud droplets. This may change the amount and optical properties of clouds, and hence their reflection and absorption. It may also have an impact on the formation of precipitation. As discussed in Chapter 5, these are potentially important indirect effects of aerosols, resulting probably in a negative radiative forcing of as yet very uncertain magnitude.
(My emphasis added). A few snippets of interest from chapter 5, which runs to 60 pages:
p. 304, Under Summary of Main Uncertainties Associated with Aerosol Sources and Properties
Perhaps the most important uncertainty in aerosol properties is the production of cloud condensation nuclei (Section 5.3.3).
Earlier they lead up to cloud condensation nuclei being the tricky part – with the other bits being easier? p291-2:
.. An analysis of the contributions of the uncertainties in the different factors needed to estimate direct forcing to the overall uncertainty in the direct forcing estimates can be made. This analysis leads to an overall uncertainty estimate for fossil fuel aerosols of 89% (or a range from –0.1 to –1.0 Wm–2) while that for biomass aerosols is 85% (or a range from –0.1 to –0.5 Wm–2 ).
.. An analysis of the contributions of the uncertainties in the different factors needed to estimate indirect forcing of the first kind can be made. This analysis leads to an overall uncertainty estimate for indirect forcing over Northern Hemisphere marine regions by fossil fuel aerosols of 100% (or a range from 0 to −2.8 Wm–2).
.. The indirect radiative effect of aerosols also includes effects on ice and mixed phase clouds, but the magnitude of any indirect effect associated with the ice phase is not known. It is not possible to estimate the number of anthropogenic ice nuclei at the present time. Except at very low temperatures (<−40°C), the mechanism of ice formation in clouds is not understood. Anthropogenic ice nuclei may have a large (probably positive) impact on forcing.
So clearly the science was settled back in 2001.
(By the way, I think the scientists who put chapter 5 together did a great job and have clearly been hard at work trying to understand an extremely complex subject).
Of course, science moves forward so let’s take a great Six Year Leap to 2007 and the 4th Assessment Report.
What do we find? First let’s note that the radiative equivalence of the anthropogenic increase in CO2 is around 1.7 Wm–2 (the radiative equivalence can view the CO2 increase as if it was an equivalent amount of radiation from the sun).
p29:
Direct aerosol radiative forcing is now considerably better quantified than previously and represents a major advance in understanding since the time of the TAR, when several components had a very low level of scientific understanding. A total direct aerosol radiative forcing combined across all aerosol types can now be given for the first time as –0.5 ± 0.4 Wm–2, with a medium-low level of scientific understanding.
(My italics). Note that one aspect of aerosols is somewhere between no effect and cancelling out half the anthropogenic CO2 warming.
p30:
Anthropogenic aerosols effects on water clouds cause an indirect cloud albedo effect (referred to as the first indirect effect in the TAR), which has a best estimate for the first time of –0.7 [–0.3 to –1.8] Wm–2
And the other key climate impact of aerosols is somewhere between not much and cancelling out all of the anthropogenic CO2 warming.
Thank goodness we now have a much better understanding of aerosols from 2001 when we knew for sure that humans had caused – and would continue to cause – significant warming.
Now we can quantify the aerosol effect, we know that they either do nothing or completely cancel out anthropogenic global warming. Back in 2001, we didn’t have to let that dent our confidence!
Just as a little footnote, those IPCC skeptics had to try and calculate error bars, but at such a low level of scientific understanding the error bars themselves might be in question.
And another footnote, I used the “Report accepted by Working Group I of the Intergovernmental Panel on Climate Change but not approved in detail” because I downloaded this when it was first released in 2007.
The Science of Doom 
So taking the best estimates, the radiative equivalence of aerosols is -1.2, compared with +1.7 for anthropogenic CO2.
Would subsequent (positive) feedbacks be applied to the +1.7 or the net +0.5? Or a bit of both?
Mr Potarto
Not quite.
If you download Chapter 2 of the IPCC report (from ipcc.ch), and take a look at p136 (really the 8th page of ch 2) you see all of the various forcings with their various error bars.
The total works out to 1.6W/m^2 (-1.0, +0.8), i.e. between 0.6W/m^2 and 2.4W/m^2 – it’s common to see error bars that aren’t symmetrical.
Other positive forcings are other “greenhouse” gases like methane, other negative forcings are land use.
From a conceptual point of view, once all of the first order forcings are considered then the feedback effects are calculated from the resulting temperature change for this first order forcing.
But GCMs are very sophisticated, even with their many limitations. So if, for example, a feedback (positive or negative) happens quickly then it will act on immediate changes in temperature. Other feedbacks might be slower and so will only act on longer term changes.
Jup, the aerosol forcings (and their link to clouds) are very uncertain. Should be a point of focus for research. However, the estimations for climate sensitivity for 2xCO2 also have a very large range. It is just that it is not very likely that the aerosol forcing is near its most extreme negative value, so it is not likely that it can counteract AGW (note globally!) completely, also for other reasons (short lifetimes, inhomogeneities, etc) explained in the AR4 (you should read the WG1 document completely for once if you haven’t done that already (or wait for AR5..)). It’s likely that AGW will warm the planet, however how much exactly is somewhat uncertain. And, please, don’t ask about regional changes… or about the possible secondary effects, such as on ecosystems, humans, etc. There is still a lot more to do and to explore! My personal opinion is that AGW could be “inconvenient” ;), especially if the sensitivity turns out to be large.
This blog entry goes back a way but I thought I’d post a note on a few things that aren’t clear in the text since there are bound to be some new people working through this excellent site.
CO2 is the only anthropogenic “positive” forcing mentioned here and it is heavily implied that the potential magnitude of “negative” forcings from aerosols can wipe this out in one blow, thereby bringing into question the IPCC declaration of detectable human-induced warming.
As alluded to in scienceofdoom’s reply to Mr Potarto there are other anthropogenic greenhouse gases. CO2 represents about 50% of the total magnitude of human-induced “positive” forcings. CO2 is often used in discussions as short-hand for all anthropogenic climate effects for three reasons:
1. It is the largest single forcing.
2. The best estimates for all other negative and positive forcings cancel each other out almost perfectly so the final forcing total (1.6) is very similar to CO2 on its own (1.66).
3. CO2 has a very long lifetime in the atmosphere so changes to this climate variable will have long-lasting effects.
(4. People don’t know any better. There is a reasonable case to be made that CO2 is often over-emphasised in relation to other factors).
Interestingly if you assume all forcings, both warming and cooling, are actually at the most negative point of the error bars then you do get a small net cooling. This is very unlikely of course, but one reason why conclusions are given in terms of very high confidence rather than virtual certainty.
Another thing about anthropogenic aerosols, going slightly off-topic, is that they have a very short life in the atmosphere. Their quantities will effectively only increase in line with economic growth. CO2, on the other hand, is effectively cumulative. We could introduce a global steady-state economy right now and CO2 levels would continue to rise, while aerosols might become a constant.
Hi SOD, long time reader and fan. Ran across this slightly older article, not sure if such comments get flagged for you, but a question about what you post here. From other discussions I’ve seen, folks like Hansen seem to argue along the following lines: if aerosols are on the extreme low impact end, then obviously they are not stopping AGW and so we have to worry about it. If they are on the high end of impact, then it’s actually worse because of the residence time problem – it means greenhouse-driven warming is stronger and has been dampened by the stronger aerosol effect, but as greenhouse eventually muscles out it will be roaring. Have you authored any posts that look specifically at this question? I realize it’s a difficult topic given the obvious measurement problems.
Geoff,
I haven’t looked specifically at your question. And I haven’t looked at aerosols for quite a long time.
In a slightly related area, Jeffrey Kiehl posed a question about why models are able to reproduce the observed warming given that they have a large range of sensitivity to GHGs – Twentieth century climate model response and climate sensitivity, Jeffrey T. Kiehl, GRL (2010):
The paper is free, easy to read and short. I recommend it.
Going back to what Hansen says, it makes sense in principle. It’s just simple maths in words.
There are a lot of uncertainties in understanding climate and modeling climate that are discussed in papers but they don’t make it into executive summaries of IPCC reports.
The uncertainties don’t mean the range expressed in climate models is wrong but I suspect the real uncertainty is much higher. I hope to develop this idea further.
Geoff Price,
Hansen is posing a false dichotomy in my opinion. He appears to ignore the possibility that the modeled climate sensitivity to ghg forcing is too high, not that aerosols or some other unknown factor that will go away soon or be overwhelmed by ghg forcing is keeping the temperature from rising rapidly. Models have clouds as a strong positive feedback that increases the effect of increasing absolute humidity at higher temperature. I haven’t seen any measurements that would confirm this.
Sure, Hansen tends to say he is pretty convinced that sensitivity will be somewhere near that 3C midpoint based on his understanding of the paleoclimate-based lines of evidence, and so doesn’t think it likely that climate sensitivity is on the very low end.
As I understand it models vary as to how cloud feedback currently nets out; with some it is closer to zero because of the dueling effects of albedo increase and greenhouse enhancement (high level vs. low level clouds etc.) Clearly uncertainty about clouds and how to estimate/validate feedback effects is high, which is why we’re stuck with “1.5 to 4.5″…
But what if cloud feedback is negative? Sure, cloud cover reduces the rate of heat loss from the surface at night, but not by all that much. The cloud tops still radiate constantly and that energy has to come from the surface by sensible and latent heat convection. Meanwhile, the surface receives nowhere near as much energy during the day. It is counterintuitive to me that if there were 100% cloud cover, the surface would be warmer than it is now. That’s the obvious implication of cloud positive feedback.
Citing Hansen, Geoff wrote: “if aerosols are on the extreme low impact end, then obviously they are not stopping AGW and so we have to worry about it”.
This argument is totally bogus. Given the observed amount of 20th century warming, Hansen’s assumptions about ECS place limitations on the amount of cooling aerosols could have caused. See Figure 1 in the Kiehl paper mentioned by SOD. If Hansen believes climate sensitivity is around 3 K, then current total forcing must be around 1.4 W/m2. Since the current forcing from well-mixed GHGs is about 3.0 W/m2, Hansen’s assumption that ECS is 3 degK implies strong cooling from aerosols (-1.6 W/m2). Given observed warming, one must believe in either: a combination of high sensitivity to both GHGs and aerosols or low sensitivity to both GHGs and aerosols. The combination of high sensitivity to both GHGs and aerosols leads to catastrophe because someday we will be stuck with large amount of CO2 in our atmosphere and little aerosols to offset its warming (without geo-engineering). The combination of low sensitivity to both GHGs and aerosols means future warming will be less of a problem that the IPCC’s estimates: less warming due to GHGs and less warming when we emit fewer aerosols.
The situation with respect to aerosols has evolved since SOD wrote this post. The slight cooling/hiatus in warming from 1950-1970 prompted climate scientists to believe in high sensitivity to aerosols, and climate models today still use a high sensitivity to aerosols. A variety of factors are now causing estimates of aerosol cooling to come down: 1) The current hiatus in warming is not caused by rising aerosols (except a minor contribution from volcanos). 2) Pinatubo produced less cooling than anticipated. 3) Despite the fact that aerosols are localized in the NH, the NH has warmed more rapidly than the SH. (This can be partly explained because the SH has more ocean.) These observations prompted AR5 to raise its best estimate of aerosol forcing to -0.9 W/m2 (reducing sensitivity to aerosol cooling. Using Kiehl’s Figure 1, that implies that ECS is about 1.9 K! Instead of lowering their best estimate for ECS, the IPCC lowered their range (to 1.5-4.5 K from 2.0-4.5 K) and stopped making a best estimate (apparently because of the large disagreement between climate models and these considerations (called energy balance models)). Just recently, one expert produced a best estimate of aerosol forcing of -0.5 W/m2. Nic Lewis has been publishing on and publicizing the consequences of changing estimates of aerosol forcing, but you don’t need to take a skeptics word for it. Just look at Figure 1 from Kiehl (a non-skeptic, co-author of the K-T energy balance diagram).
There was a recent workshop of experts on aerosols at Ringberg, and the presentations are posted online. Modelers claim that the response to aerosol and GHG forcing develops on different time scales, making simple models unreliable. In other words, all W/m2 (forcing) are not created equal in their ability to produce non-equilibrium warming. .
SOD’s link for Kiehl didn’t work for me. This one did:
Click to access Kiehl_2007GL031383.pdf
http://climateaudit.org/2015/03/19/the-implications-for-climate-sensitivity-of-bjorn-stevens-new-aerosol-forcing-paper/
http://www.mpimet.mpg.de/en/science/the-atmosphere-in-the-earth-system/ringberg-workshop/ringberg-2014.html
According to the framework, cloud feedback would have to be very significantly negative to achieve lower end of the sensitivity spread, given the magnitude of the water vapor feedback in particular which it has to offset (and the other larger one also being positive in albedo.)
As far as effect on heat loss, not sure how small the cloud effect really is, though you may understand more of the relevant physics. Studies assessing regional DWLR strength over time certainly seem strongly affected by cloudiness. And in terms of intuition, Venus has 100% cloud cover, and much more reflective than earth, but still quite balmy no…?
The surface pressure on Venus is 90+bar with the vast majority of it CO2. Under those circumstances the average of 20W/m² of sunlight that makes it to the Venusian surface is enough to maintain the ~750K surface temperature. Comparing that to a thought problem of the Earth with it’s current atmosphere but 100% cloud cover is not particularly enlightening.
To put it another way: Venus would be much, much hotter if it were not covered with clouds.
Using MODTRAN with the US Standard atmosphere, the upward radiation to space for clear sky is 260.2W/m². For a 2km thick layer of cumulus clouds starting at 0.7km, the upward radiation is 223.0W/m². For 60% cloud cover, the average upward radiation is 0.6*223+0.4*260.2=237.8W/m², or a reasonable approximation of the calculated TOA upward radiation of 239W/m². If the cloud cover is increased to 100%, then the upward radiation is reduced by 237.8-223.0=14.8W/m². That means that for the no feedback case, albedo could only increase by 14.8/342=0.0433 or from 0.30 to 0.3433. For strong positive feedback it would have to increase even less than that.
Does it seem reasonable to you that an increase in cloud cover of 67% would increase albedo that little? It certainly doesn’t seem reasonable to me. But perhaps I’m being overly simplistic.
You can also pose the opposite thought problem: What would happen if there were no cloud cover? Again, if clouds are a positive feedback, the GMST would drop. But in this case we have a better idea of the change in albedo. If there were no cloud cover, the albedo would drop to about 0.1, leading to an increase in energy absorption of 68W/m². But the emission would only increase by 21W/m². Clouds keep the planet cooler than it would be otherwise and therefore clouds must be a negative feedback, not neutral or positive, much less strongly positive.
Clouds do cool the net climate, but that does not mean they are a net negative feedback. For example, imagine that low-level clouds decrease as the climate warms — that would produce a negative feedback despite the fact that clouds are a net cooler.