Geoengineering down under: Is Stratospheric Sulphate Injection Completely Reversible?

This guest post is by Simon Terry, Executive Director of the Sustainability Council of New Zealand. The risk rating on stratospheric sulphate injection went up another notch on the basis of material presented at a recent geoengineering symposium in Australia organised by the Australian Academy of Science, while the existing climate change risks did not get any better. The event made a useful contribution to the understanding down under of so called ‘geoengineering’ and delivered some perspectives that will be useful internationally, including a review of sulphate injection that raised a new issue: is it completely reversible? More on that below.

While not exactly the “southern hemisphere perspective” that was billed (as the contributors barely exceeded Australia’s borders), it nonetheless delivered strong presentations and discussion — partly as a result of most speakers being specialists in the field related to each technique reviewed but not technique proponents themselves.

The first result of this was that efficacy received strict treatment in most presentations and ocean fertilisation got hammered early on this count. The level of benefits initially claimed for ocean schemes have been called into question by a number of studies and the symposium highlighted the scarcity of places in which the technique would work significantly, while noting that even in these areas success would often be thwarted by the vertical mixing of ocean layers.

A critical review of biosequestration opportunities in Australia provided a reminder of the care that needs to be taken to distinguish technical potentials from what can actually be delivered once socio economic constraints are applied. While the continent’s technical potential was estimated by Australian government scientists at double the nation’s rate of emissions (around 500 Mt/yr), in practice they expected the amount of land that would actually be made available would mop up only about a tenth of those annual emissions.

Carbon capture and storage (CCS) did have a proponent in the speaker slot, but even in the land that is the world’s largest coal exporter, the notion that CCS was going to deliver in anything like the time required was far from obvious to the participants. Capturing emissions as a part of the projected explosion of “CCS Ready” power stations appeared more like a justification for business as usual from an industry scarcely acknowledging alternatives than a proposal to address dangerous climate change. All the more given the recent work by a Melbourne collective, Beyond Zero Emissions, that conservatively calculates Australia could get all its power from renewable sources within ten years at a cost of 3% of GDP additional to the fossil alternative — and 60% from solar no less.

The question of how close to the UNFCCC idea of ‘dangerous climate change’ we have already come was largely to the side as a result of the symposium’s heavy focus on technique assessment. Yet without an idea of the extent of risk already expected on the climate change side of the ledger, there is little context for assessing the new risks solar reflection or certain sequestration projects would present and thus the overall merit of any of these activities that have been loosely bundled under the banner of ‘geoengineering’. Governments’ polite hand wringing about humanity being unlikely to stay below a 2ºC rise provides no meaningful guidance as to how they will confront a series of ugly choices on the horizon, including solar reflection options.

The symposium did give an hour to the critical issue of governance but the presentations barely scratched the surface on how decisions over the experimentation or deployment of various techniques might best be made.

A thoughtful presentation by Greg Bodeker did however put a new twist on sulphate injection — one of the sharpest ugly choices and a particularly difficult governance challenge — it being the current default means of accessing cheap and brutally effective cooling. Bodeker, an independent atmospheric physicist and former NIWA scientist, first reviewed the well-catalogued plusses and minuses expected if sulphate injection were to be employed.

On the positive side he listed: effective, affordable, reversible, timely, more efficient photosynthesis, and tuneable. On the negative side were: ozone depletion, regional climate change (particularly rainfall pattens), continued ocean acidification, whitening of the sky, reduced solar power generation, governance challenges, along with the need for constant maintenance and problems if the system broke down for any length of time.

And slipped in the middle of this review under ‘reversibility’ was the following: “can we be sure that there is no hysteresis in the system? Will the path coming back from [this technique] be the same as the path going out?”

This question arose for him from the understanding that the climate system responds to rates of temperature change as well as to the magnitude of the change. Thus a winding down of sulphate injections in concert with reduced greenhouse gas concentrations could produce a different climatic response to the gases being reduced on their own.

However there seemed more to this line of thinking so I contacted him to test out the idea that the end state of the climate system after sulphate injections ceased might not be the same as when they began, other things being equal? While it was a little different to what Bodeker had been thinking, he said that the hysteresis concept did extend to the idea that the climate system could carry some form of memory which if ‘overwritten’ by an aerosol injection pattern, might not return to the prior state.

The first point to note here is that although there are some parallels in natural systems to this sort of memory effect, the above is simply raised as a theoretical possibility for exploration. Further, the process we are already engaged in of raising greenhouse gas emissions and (hopefully) then lowering them could in any case have a similar (though perhaps different) effect.

Finally, letting temperatures rise without such interference will definitely cause climatic changes and the outcome from this could easily be a lot worse than any not-entirely-reversible component of sulphate injection that might be identified. Nonetheless, the implicit question this line of thinking poses is: could the fix leave things in a sufficiently damaged state at the end that it would alter choices about how or whether to use this or other solar reflection techniques?

Overall, what this opens up is a subtle “unknown unknown” that has the potential to shift how we think about experimentation as much as deployment. While it is clear that the physical flow of sulphates that would drive the reflection process is fully reversible (as the aerosols will fall out of the stratosphere in a year or so), whether the effects of having introduced them at all will be completely reversible is a separate question. If the technique kept the temperature rise down as planned, could the process nonetheless leave a scarring that carries meaningful ongoing costs once the injections cease? Bodeker’s key point is that we simply do not know enough to reliably tell.

The possibility of such effects takes on more significance when full reversibility would otherwise stand as the ultimate safety assurance that could be offered. For instead of being able to acknowledge that sulphate injection has the capacity to alter rainfall patterns and then state “but at worst it can all be switched off and will return to present conditions”, that back pocket safety assurance could end up being a more qualified one.

Bodeker’s companion points show our chances of getting clarity on this new theoretical potential and similar issues are not good. This is due to both the complexity of the climate system and the relative lack of sophistication of computer models used to simulate it. On the latter he notes that:

  • Many of the models used to assess geoengineering actions do not have a fully resolved stratosphere.
  • Few include coupling to a dynamic vegetation model.
  • Models show surprisingly large differences in the pattern of radiative forcing for the same stratospheric aerosol distribution.
  • Changes in precipitation — choose your model for the pattern that you want.

With respect to assessing the potential for incomplete reversibility of sulphate injections, having subsequently spent time thinking about how this could be tested for, Bodeker’s conclusion is that:

“You really need a fully coupled Earth System Model, and while these are currently available, there is no computer powerful enough to execute them and certainly no computer on which an ensemble of simulations could be performed (which is what would be required to establish statistical robustness)”.

To what extent incomplete reversibility could apply to the more flexible and less brutal alternative of reflection through cloud whitening is another area for exploration.

Bodeker’s overall observation is that a high degree of system understanding is required to assure that increased intervention leads to decreased impacts, and that we do not have that understanding. Moreover, he does not see this changing any time soon as “The demand for policy-relevant scientific advice on the risks and benefits of various proposed methods of geoengineering is outstripping the supply”.

Although a great deal more research is taking place to assess the risks from climate change in absence of large-scale intervention, the unpleasant surprises that have emerged in recent years are clear reminders of how little is also known about some of those risks. How long will the arctic ice cap last under projected conditions, how many severe droughts can the Amazon basin withstand end on end, and so forth? And on both sides, we are still learning about the questions.

Abstracts of the conference presentations can be accessed at:

1 thought on “Geoengineering down under: Is Stratospheric Sulphate Injection Completely Reversible?”

Leave a Reply