Added: I’ve added a bit to the post, in italics, after some reflection and looking for satellite measurements of SO2. Quick take: we’ve only had satellite data sensitive enough to monitor fossil fuel SO2 contributions for 5 years or so. I haven’t found trend data yet, but I’ll keep looking.
A recent paper from a team of statisticians at Boston University contains a claim that the lack of global warming during the first decade fo the 21st century is due to the cooling effect of sulfur emissions from Chinese coal plants. It is an interesting hypothesis, physically plausible, but I’m still digesting the paper. Here are a few musings after a read through.
Volcanoes can cause measurable global cooling due to SO2 emissions. However, a large volcanic eruption injects sulfur compounds high into the stratosphere where they are distributed widely across the global and linger for several months. As I understand it, coal-fired power plant stack emissions do not reach the stratosphere in any large amounts and are mostly confined to the troposphere. There are transport mechanisms for bosoting fossil fuel sulfer emissions into the stratosphere (e.g., thunderstorms) so the steady increase will add to the stratospheric SO2 load. Tropospheric sulfur is quickly washed out in the form of acid rain. Coal stack sulfur emissions are mostly a regional problem. Evidence for this is the fact that the US EPA shows a consistent drop in US sulfur dioxide measurements since around 1980. The Chinese coal emissions aren’t pushing up SO2 levels in America’s skies. The EPA measurements are made near the ground, so they can miss stratospheric concentrations. Still, the stratospheric component will contribute to near ground level measurements as it drifts down. It would be interesting to attempt to infer the global loading by looking at near ground measurements and perhaps acid rain data. Unless I’m missing something, if we don’t see elevated sulfur levels here, the cooling effect must be largely a regional, not global, effect just from physical considerations. Further, as the IPCC in earlier reports classed our understanding of the effect of sulfur compounds on climate as ‘poor.’ Simply because volcanic sulfur emissions in the stratosphere are effective at cooling the climate, it doesn’t follow that we understand tropospheric emissions enough to be sure the paper’s model is in fact correct.
Speaking of models, as is so common in the climate papers, we have models used to predict both the SO2 cooling contribution, and more importantly, models that predict the warming from CO2 forcings, solar cycle forcings, and ENSO forcings. The authors conclude that their model shows the declining solar insolation of the Sun’s solar cycle, plus La Niña cooling, along with SO2 aerosols just balance out the CO2 warming effect from about 1998 to 2002 and have a small net cooling effect thereafter till 2008. Add it up and all the manmade warming and cooling effects just about cancel out, leading to no warming, and neatly explaining the lack of increasing global temps during the late Nineties and the Oughties (1998-2008). But there are other effects from burning coal besides higher sulfur emissions. As Chinese sulfur emissions rise, so do soot (black carbon) emissions, which tend to warm the air, and which the paper sort of glosses over by lumping the black carbon effects with stratospheric water vapor (which are deemed to be warming neutral, despite other research claims to the contrary). I don’t understand how the authors lump together stratospheric water vapor and tropospheric black carbon, nor how they simply dismiss their effects.
This is all a nice just-so story, and it may even be true. But I am skeptical of any climate result that comes strictly from models and not actual physical measurements that have real error bars. My main concern is that what the authors do is compare one set of models with another set of models. That is fine as far as it goes, but the map is not the territory. Models are simplifications. Models omit details, sometimes intentionally and sometimes unintentionally. And those details can be very important. In fact, if the authors are correct, the detail of SO2 emissions that was omitted from earlier climate model behaviors was just such an important detail! There is a 2005 paper that points out this very behavior, so perhaps we do have a ‘prediction’ which was not considered significant at that time. What other important details are missing from the models? Inquiring minds want to know!
I’m withholding judgement on this till others look to see how all this should play out. In particular, since I like to see physical predictions used to validate model conclusions, I’m looking for a regional climate signal due to the regional nature of the SO2 increase. In fact, if coal-fired power plant stack emissions are so effective at cooling, we should expect to see a warming effect in the US as our sulfur levels have dropped. But if anything, the US regional climate has been cooling faster than the global results. This seems counterintuitive, if the mechanism in the paper is indeed the correct explanation for global temps.
We should know more once real specialists have a chance to study the paper.
 I’ve used the WUWT link as the Proceedings of the National Academy of Sciences full text version is behind a paywall.