You can imagine the start of a climate geoengineering programme in a number of ways. The way that most appeals to me is as part of a policy portfolio aimed at reducing the future risks of climate change. This would entail careful consideration of a variety of proposals for reducing incoming sunlight, research into the weaknesses of all of them and the choice of a preferred option. Then, if as the result of a deliberative process that has been going on in parallel to this, with each informing the other, you — for a suitably inclusive, legitimate value of “you” — decide that such risk management is worth trying you start implementing on such a programme, with the aim of slowly but steadily ramping up to the level of offset you have decide is wise, while continuing with other mitigation and adaptation measures.
On the other hand, a programme might be triggered by a specific event — for example, something sudden and dire happening in the Arctic. Some such events (lots of methane coming out of permafrost) might indeed be checked by prompt cooling, though you might need rather a lot of it. Other catastrophes (radical destabilisation of Greenland ice) probably wouldn’t be helped at all. But such an emergency might trigger demands for prompt climate action that politicians found hard to ignore, and climate geoengineering might be the prompt action they turned to whether or not it met the needs of the specific emergency.
I’ve always seen this as a rather worrying scenario. Much better to think carefully about climate geoengineering’s merits and dangers and build it into a portfolio of climate action than to be bounced into it as some sort of new alternative. Among other drawbacks, a programme put together in the context of a climate emergency might have to be sized so as to deliver a dramatic effect — one with a cooling that might be measured in watts per square metre, rather than something a tenth that size — right away. This seems likely to be imprudent.
A new paper by Jim Haywood and colleagues at the Met Office and the University of Exeter in Nature Climate Change brings up a new version of this question, though, one which I find intriguing. What about the use of geoengineering to counteract a natural, rather than man-made, climatic event? (Disclosure: after hearing Jim talk about this work at a meeting last year I discussed it with him quite a lot, to the extent of eventually being acknowledged in the paper. That’s why I’m writing about it in a personal capacity here, rather than in print somewhere. Obviously the opinions in this post are purely my own.)
The research started out looking at what might happen if a climate geoengineering programme put sulphate aerosols into the stratosphere of either just the northern hemisphere, or just the southern hemisphere. The first order answer is that the hemisphere with the hazy stratosphere ends up cooled and the other doesn’t, which is kind of obvious. What is not so obvious is what happens to the Sahel, the band of countries arrayed along and below the southern edge of the Sahara. If you cool just the northern hemisphere, the Sahel dries out very badly. If you cool just the south, it gets a lot wetter, and plant life there (as measured by net primary productivity in vegetation models) does peculiarly nicely. This is apparently due to movements of the intertropical convergence zone, the line where the northern and southern hemisphere trade winds meet. If you cool one hemisphere, the ITCZ gets pushed towards the other. If the ITCZ moves south after a northern cooling, less moisture gets from the Gulf of Guinea to the Sahel, and dry times ensue. (This is, as I understand it, on top of a generic tendency of this sort of geoengineering to dampen monsoon circulations, at least in the short run, by cooling the land more than the sea.)
This fits with the twentieth century record. Three of the four worst drought years in the Sahel followed volcanic eruptions which put lots of sulphate into the stratosphere in the northern hemisphere but not the southern hemisphere (if the gunk starts off far enough north, very little gets over the equator) and thus cooled the north but not the south. Volcanic eruptions in the south — such as Agung, in 1963, which cooled only the southern hemisphere — and near the equator — such as Pinatubo, in 1991, which cooled bothe hemispheres — did not have the same effect.
It all sounds like a pretty coherent story. A number of different model runs simulating both single-hemisphere geoengineering and various volcanic eruptions show the drying of the Sahel following northern cooling. A fairly straightforward mechanism accounts for why this might be. The historical record seems consistent with it. Is it a sure thing? no. Is it the way to bet on current knowledge? Yes. If a large volcano significantly north of the equator erupts, it seems that theres a good chance of a subsequent drought in the Sahel.*
So what does this mean for geoengineering? One implication is that climate geoengineering deployed in just the northern hemisphere looks like a very bad idea. Programmes in just the north have been considered and studied, in part because of the worries people have about something suddenly going wrong in the Arctic, something that needs “fixing” quickly. This research makes such approaches look dangerous.
More interesting, and more novel, is the implication that geoengineering might be used to avert a Sahelian drought caused by a volcano. If the stratospheric sulphates released in a major northern eruption were promptly countered by a deliberate release of sulphates into the southern hemisphere, both hemispheres would cool. The ITCZ would stay put, and a drought might well be averted. For a major drought, that would be a big win. The drought in the 1980s, which followed on the 1982 eruption of El Chichon in Mexico, killed about a quarter of a million people and turned millions more into refugees.
One way to think about such an intervention is that it would turn the hemispherically asymmetric effects of an eruption like that of El Chichon into the global effects of an eruption like that of Pinatubo in 1991. Pinatubo was a bigger eruption in terms of the overall amount of emissions, and it had various measurable effects, some of them disturbing (drop in stratospheric ozone, slowing of the global hydrological cycle). But if El Chichon did indeed cause, or help cause, the Sahelian drought of the 1980s then its effects were a lot worse.**
And if the Earth is left to its own devices, such droughts will happen again. Last century there were two eruptions that cooled the north and were followed by drought in the Sahel. The north is better endowed with volcanoes than the south, since the Pacific “ring of fire” is more a horseshoe of fire, with a gap in the south but a continuous arc in the north. The odds of at least one eruption in the Pinatubo-to-Krakatoa range somewhere of the Earth in this century are better than even. The chances of one happening in the north are obviously lower; but the odds are hardly long.
If humans had had the technological wherewithal to stop the 1980s Sahel drought in its tracks, would people have wanted to use it? It seems likely that there would have been a constituency for it, not least in the Sahel. And many of the reasons people have for objecting to geoengineering as an inappropriate “technical fix” to man-made climate change might apply with rather less force if the technology was being used to forestall a natural disaster on a continental scale.
That said, to balance out a large eruption is not really something to be attempted from a standing start. The hardware needed to lift millions of tonnes of sulphur into the sky could not be whistled up on the fly. And some of the arguments against embarking on geoengineering in a rush that I made at the top of this post would clearly apply.
But what if one were already in the climate geoengineering business, with a fleet of aircraft (or balloons, or whatever) dedicated to maintaining a permanent stratospheric haze at all latitudes? Then things might look rather different. Stopping operations in the northern hemisphere and upping them in the south might be quite feasible. This might well also be the case for other geoengineering techniques, such as the brightening of marine clouds. Providing more brightening, and thus cooling, in the Atlantic Ocean off Namibia might also be a way of keeping the ITCZ put in the aftermath of a large northern eruption.
Climate geoengineering programmes that work by reducing sunlight would have to be designed to take account of the fact that there are occasionally large volcanic eruptions. This work strongly suggests that one of the ways they should be taken into account is by the development of contingency plans aimed at balancing out their hemispheric effects, if any. If such a balancing mechanism existed, there seems to be a real chance that it could avert a major drought in the Sahel next time a big volcano pops off in the north.
That is not in itself an unanswerable argument for such a scheme. There are other ways of dealing with droughts (though they didn’t work all that well in the Sahel in the 1980s). And people might well think there are issues that should weigh more heavily in deliberations about climate geoengineering than a theoretical ability to counteract some sorts of natural disaster.
But if it is not an unanswerable argument in favour of such a scheme, it seems to me that it is a new and interesting one. A world with geoengineering in it often seems, in our imaginations, a more fragile one. In this particular respect, though, such a world might have a new resilience.
* There are other factors at play, including the possibility that cooling of the northern hemisphere by industrial aerosols was playing a role in the 1980s; there’s also a possible role for El Nino. But the first order story seems pretty plausible.
**Some may think at this point that there could be an argument for a permanently asymmetric geoengineering scheme designed to cool the south, push the ITCZ north, and make the Sahel wetter and more productive. There may indeed be attractions to such an idea. But there would be worries, too. In the Met Office models the sort of changes that bring moisture to the Sahel in this way reduce rainfall in north-eastern Brazil. Though it is my understanding that Met Office models have a history of being more prone than others to predict drought in north-eastern Brazil, that is something one would want to worry about a lot. Moving the ITCZ would also, I think, make a difference to the formation of hurricanes in the mid Atlantic. In general, many people might find find an intervention aimed at keeping the ITCZ in its customary place more acceptable than one aimed at permanently moving it to a specific group’s advantage and the possible detriment of others.
4 Comments so far
Leave a comment