Let’s put together a consensus report: things about the Earth’s sensitivity to various forcings that AGW-skeptics, AGW-deniers, and AGW-affirmers will all sign off on. We remember, of course, that just because something isn’t controversial doesn’t mean it’s true.
The Earth’s average temperature has substantial yearly variations, and no theoretical model is going to be able to capture those–because of extreme sensitivity to initial conditions, if nothing else. However, the equilibrium mean temperature for a given average solar irradiation, CO2 and water vapor concentration, albedo, etc can be determined theoretically, irrespective of initial condition issues, because it’s just energy conservation. The net energy absorbed by the Earth from the sun has to be equal to the net energy radiated away by the Earth. Therefore, the fact that we can’t predict the weather more than a few days in advance doesn’t mean that we can’t predict the climate decades in advance (up to unmodelable changes in forcings, either regarding solar, volcanic, or human input).
Climatologists describe the Earth’s thermal sensitivity in ways that may be not be intuitive to some, so it’s worth reviewing them. They start out from some reference state, usually the Earth at pre-Industrial times. This reference state is presumed to be in equilibrium. Now suppose the system is perturbed. The perturbation is described by a change in the radiative forcing of the Earth, measured in energy per time per area (W m^-2). In equilibrium, the net forcing is zero (radiation in = radiation out). Let’s call the forcing “F“, so F=0 in the original equilibrium. F can become nonzero if the solar input increases or if the radiation out of the atmosphere is reduced. The latter is what the greenhouse effect does. Radiation is absorbed in the atmosphere and reradiated up and down. If the temperature profile doesn’t change, the net radiation leaving the earth will decrease. The net forcing is now F=dF=/=0.
For carbon dioxide greenhouse forcing, there is a simple formula (actually an analytic fit to 1D radiative equilibrium models):
F = 5.5 ln(C/C_0)
where C is the CO2 concentration, and C_0 is its preindustrial level. So, doubling the CO2 concentration (which humanity will probably do by mid-century) means a forcing of 3.7W m^-2. AGW-skeptics sometimes talk as if that natural logarithm is a big secret, but really it’s discussed on quite a few AGN-affirming websites e.g. (here and here). If the effect of C on T_s were linear, we’d be talking about a much larger temperature change than 3 degrees.
For the terrestrial system to come back into equilibrium, the surface temperature T_s must increase. Heating produces a forcing, F_T, to counteract the greenhouse forcing. This forcing is proportional to dT_s, the change in temperature. We can write this as F_T = Φ_T dT_s, introducing a new constant, Φ_T, the (temperature) climate feedback parameter. Assuming no other forcings (which will turn out to be a bad assumption), equilibrium will be reestablished when F_T=F and dT_s=F/Φ_T. So, to find the change in surface temperature, we just need to know Φ_T. The climate feedback parameter for temperature change with everything else held constant (water vapor, clouds, ice cover, etc) is one of the less controversial calculations–skeptics like Happer cite it. One gets Φ_T = 3.7-4.4 W m^-2 K^-1. Then the equilibrium temperature change for a doubling of CO2 is about one degree K.
As you all know, the IPCC favored value for the actual expected climate change for CO2 doubling is 2.1-4.4 K. The first thing to notice is that even the simple, uncontroversial calculation gives something of this order of magnitude. Changes of order a degree are what should be expected. However, the actual number is a factor of three different, the reason being that everything else does not stay constant when added CO2 heats the atmosphere. There are feedbacks. Three are particularly important: 1) water vapor changes (H2O being an important greenhouse gas in its own right), 2) ice/snow cover changes (melting reduces the Earth’s albedo, so it absorbs more solar radiation–a positive feedback), 3) cloud cover (increased clouds increases albedo, meaning more of the sun’s radiation is reflected–a negative feedback, while clouds can also increase greenhouse trapping of infrared waves–a positive feedback, so the net effect of clouds is complicated). So we can define a total climate feedback parameter Φ, which sums over all the forcings (forcings being additive for small changes):
F = Φ dT_s
Φ = Φ_T + Φ_H2O + Φ_ice + Φ_clouds
In the standard picture, Φ_H2O, Φ_ice, and Φ_clouds all work out to be positive feedbacks, meaning that they are all negative, so that the total Φ<Φ_T. Note that, for our purposes, low Φ is bad, because it means more heating to offset a given forcing, while high Φ is good. The current best guess for Φ is around 0.9-1.6 W m^-2 K^-1. This is based on an ensemble of computer models, meaning there are almost certainly systematic uncertainties that are not captured, and the real uncertainty is larger than this range. We can put limits on Φ by looking at past climate variability if we know both the external forcing and temperature change. For example, if Φ=∞, the temperature would never change, and we know that’s not right. If Φ<=0, the atmosphere would be thermally unstable to a greenhouse runaway, something that may actually have happened on Venus in the distant past. Observations of temperature change within recorded times from the solar cycle and from volcanic eruptions give Φ around a range of 1-4 W m^-2 K^-1, meaning a CO2 doubling dT_s of around 1-4 degrees. Of course, these numbers assume that researchers have at least correctly estimated their uncertainty in the forcings that were active during these times/events.
There are good reasons to believe that the last half-century’s global warming has been due to greenhouse forcing of some kind. There is the fact that the stratosphere has cooled while the troposphere has warmed, something that would be expected from an increase in the infrared opacity of the atmosphere but not, for example, from an increase in solar radiation. There is the fact that average nighttime temperatures have raised more than average daytime temperatures. Again, this is expected from stronger radiation trapping in the atmosphere. Finally, satellites cannot yet accurately measure the radiation imbalance that’s heating the troposphere (some AGW-affirming websites seem a bit misleading on this point), but they have registered the effect of increased greenhouse gas concentrations on the IR spectrum of the atmosphere, and these spectra can be used to validate 1D radiation transfer models that contain the expected greenhouse forcing.
So, is the skeptic’s case sunk? Not necessarily; it depends on what his alternate explanation or prediction is. Some alternative explanations of global warming (increased direct solar heating, volcanic CO2) are definitely excluded by observations. Among the more respectable objections to the AGW position:
- Freeman Dyson thinks the feedback from stimulated plant growth and resulting carbon sequestration has been underestimated. He points out that human policy may actually have a significant effect on this feedback and suggests an appropriate change in focus.
- William Happer thinks that the net effect of feedbacks will turn out to be small, or even slightly negative. That is, he thinks Φ is about equal to, or slightly greater than, Φ_T. He cites the many unknowns of cloud physics as the main source of uncertainty. The climate establishment admits there’s a lot of uncertainty here, but thinks Happer is being recklessly optimistic.
- Some have speculated that other forces, e.g. a dearth of cosmic rays, have caused a reduction in cloud cover, and that this has contributed somewhat to recent warming. Mostly they argue that we don’t understand cloud formation well enough to say anything. IPCC thinks we understand it well enough to say it’s not the main driving force of GW. Also, if you just change the albedo, you don’t recover all the greenhouse markers, but clouds do lots of things and other feedbacks can come into play, so I don’t know that we can rule this out. I’ll have to look into it more.
Since this is a consensus report, I won’t agree or disagree with any of these claims here.
Coming up next, I’d like to go deeper into a few issues:
- ocean heat storage–it’s role in delaying equilibrium
- the early 20th century warming, to which anthropogenic emissions only made a modest contribution
- more on alternate (non-AGW) theories
- whether traditionalist conservatives have a dog in this fight
Filed under: Climate