We are testing geoengineering, and that is a good thing

Geoengineering is the “deliberate large-scale manipulation of the planetary environment to counteract anthropogenic climate change”. (1) It has also been called “the worst possible way to address climate change that we need to take seriously.” It commonly refers to techniques for reducing the amount of solar radiation hitting the planet, also known as “solar geoengineering”, but there are many other forms. (2) Solar geoengineering works by reflecting some of the incoming radiation, for example by injecting reflective aerosols into the stratosphere or by increasing cloud cover. (3) We are starting to test some forms of it and I think that is a good thing. But many others do not, and for good reasons. I had a fascinating discussion about this topic with Thomas Ackerman, Professor Emeritus of Atmospheric Sciences at the University of Washington, in the context of a solar geoengineering technique called marine cloud brightening. I will give a quick overview of that approach and then dig into some of the testing considerations.

Before that, though, it is important to understand that solar geoengineering does not aim to solve the problem of climate change. At best, it buys us time. If deployed, the planet will heat up more slowly (good!), but greenhouse gases will keep building up and our oceans will continue to acidify. The climate will get distorted in undesirable ways (e.g., areas may get drier or wetter), and if there is a gap in the solar geoengineering bandaid, our highly insulated planet will heat up very quickly. Solar geoengineering is risky, with difficult ethical and governance problems. On top of that, it is not a cure. It must be at most a bit player to the main act of reducing greenhouse gases.

But we should understand the option, so… What is marine cloud brightening? Over 90% of the heat that is absorbed by the Earth ends up in the oceans. (4) The warming oceans are bleaching coral reefs, melting sea ice, and endangering marine life. If we can reflect more of the sun away from the oceans, it will help cool the oceans and ultimately the atmosphere. Various ideas have been considered, with the most progress being made on using ocean spray to brighten marine clouds, aka “marine cloud brightening”. Marine clouds are low-lying stratocumulus clouds frequently found off the west coasts of Africa and the Americas. (You may have seen and heard of “the marine layer” off the coast of California.) The basic idea of marine cloud brightening is to spray a fine mist of ocean water into the air, where the mist particles will serve as condensation nuclei for cloud formation. As water condenses around those particles, the marine layer will incorporate more, smaller droplets, resulting in more reflective and longer-lasting clouds. (5)

A sea spray nozzle in action at the back of a boat. Source: Reuters

A real-world analog for this effect is a ship track, which is like an airplane contrail but for a ship. The pollution particles emitted from a ship’s smokestack serve as nuclei around which water condenses, forming an enhanced cloud behind the ship.

Ship tracks in the East Atlantic. Source: National Weather Service San Diego

Teams researching the potential of marine cloud brightening have studied ship tracks to develop models of how marine clouds form, persist, and reflect. But researchers are clear that experimental data is needed to validate and improve those models. A proposal to do small-scale testing off the coast of Monterey to refine their understanding of the technology has been postponed for more than ten years, largely due to lack of funding but also in part due to opposition to geoengineering. I found this perplexing. Why wouldn’t we want to better understand the tools we have at our disposal to fight climate change, particularly when the tests are low-cost and low-impact? So I spoke to Ackerman, who has given the issue of testing considerable thought.

Ackerman started off by saying that there are different kinds of tests, some riskier than others, and described a testing taxonomy. But even beyond that, there are people and organizations who object on principle to testing or even research into solar geoengineering. The Christian Science Monitor quotes Carroll Muffett, head of the Center for International Environmental Law, summarizing this attitude as follows: “If there’s widespread recognition that this technology is not going to solve the climate crisis … it makes little sense to invest in experimentation.” We know that these techniques can only buy time, but some fear they will do much worse if they give people a false sense of security and delay the more necessary work of greenhouse gas reduction. (6) Testing can be seen as a slippery slope towards instilling that false confidence, so there is some objection to any testing on principle. One such critic describes a very limited proposed test of stratospheric aerosol injection as “mostly a stunt to break the ice and get people used to the idea of field trials.”

Others who are more supportive of testing will say that our need to understand our options balances those risks, and we should work to control both the risk of the testing and the risk of false security. Environmental scientist Alan Robock, who has compiled an often-cited list of 27 risks of geoengineering (see Table 1 in this comment), nevertheless thinks we need to study it. (7) It can also be argued that we are putting ourselves at some risk if other countries work to understand these technologies and we don’t. “A defensive posture requires research,” Ackerman points out.

Ackerman begins his testing taxonomy with a level that he refers to as “process testing”. These tests have essentially no measurable impact; their purpose is to vet aspects of the technology. For example, how do particles disperse in the stratosphere? Can you even find them to measure them? (8) Can we achieve a sufficiently fine spray in real-world conditions? Can we measure the cloud parameters accurately enough? Ackerman characterizes this type of experiment as “an engineering project, not a climate project. Does the process work, and can we measure it?” The actual impact of these tests would be much less than we would see from a single international flight or transatlantic ship. In his view, these are not “ethically challenging” experiments, and he notes that this kind of experiment is frequently done for air pollution research.

The second phase of testing is a localized, short-duration experiment designed to impact an accessible metric, probably one indirectly related to climate. (The climate itself would not be measurably impacted due to the small scale of the experiment.) For example, a boat might spray salt water over a small area for a few days, and the reflectivity of that area would be measured from satellite and compared to that of other days with similar atmospheric conditions. Ackerman and others have questioned whether such a small-scale test is feasible with stratospheric aerosol injection. Once the particles are aloft in the stratosphere, they stay there for weeks or months, dispersing around the globe. It is not clear how you could achieve a measurable effect on reflectivity without conducting a somewhat large-duration, large-scale test. The question is, can you design and execute a test that is highly localized and has no climate impact, yet able to give meaningful information about the climate impacts of a potential larger-scale deployment?

The third phase of testing is a larger-scale experiment designed to impact a climate metric like temperature or precipitation. This is likely to be a longer and/or bigger experiment because the metrics are harder to impact and measure reliably. For example, ocean temperature is a 3D variable (unlike reflectivity), and requires a whole network of sensors to measure. Both precipitation and temperature are also subject to many variables, so experiment time is needed to account for differing conditions. Some of the technologies would allow for an experiment to be cancelled part way through, which helps to mitigate risks. For example, you could stop spraying ocean water, or stop pumping water onto ice sheets. You might also be able to reverse the effect of an experiment, for example by melting ice that you had created. But some things are hard to undo (e.g., removing iron from the ocean) and effects underway could take time to play out.

It is certainly the case that we are already doing this type of climate experiment at scale, albeit unintentionally and poorly. We are operating coal plants that emit radiation-reflecting pollution, we are flying planes that enhance our cirrus cloud cover, we are warming and acidifying our oceans, and so forth. Given the crisis we are facing from this uncontrolled “experimentation”, you might reason that these designed experiments are surely better, since they are attempting to undo the damage we are causing. Or do the ethical and governance problems render geoengineering techniques so problematic that we should bypass them and focus exclusively on fossil fuel reduction? (9)

Not all geoengineering techniques are the same. One of the reasons I am intrigued by marine cloud brightening is that it seems modular, relatively innocuous at least at small scales, and testable. It uses only ocean water (and ships) over the ocean. It can be tested over a small area, with the reflectivity being measured fairly easily and unobtrusively from satellite, aircraft, or drones. It can be deployed regionally and is easy to stop. These are all significant advantages over stratospheric aerosol injection. In fact, a form of ocean spray engineering (without marine clouds) was tested just a month ago over the Great Barrier Reef, where about half of the coral died in 2016 and 2017 due to unusually warm water. But marine cloud brightening is less ambitious than stratospheric aerosol injection, more limited in scope, and more complex in some ways. The fact that it is most effective in certain locations means that modeling is tricky. Ackerman likens it to a hammer hitting a bell and “ringing the bell”. It is complicated to understand how those discrete impacts affect the whole system. (10)

Ackerman highlighted his concern that we aren’t giving enough attention to geoengineering, despite or even because of the difficult issues that it poses. He would like to see a more dedicated effort around risk management, process definition, and governance. He refers, for example, to the belief people hold that we would deploy some form of geoengineering when we are really in trouble, when we are “going over a cliff”. Not only are we not ready to deploy anything, but it’s not even clear that we will know when we are going over a cliff. “How bad does it need to get?” he wonders. The Australian fires weren’t enough. The shrinking ice cap isn’t enough. What if the Thwaites Glacier goes? Is that the cliff? We have no framework for thinking about these things, and it worries him.

It is true that solar geoengineering will likely do more harm than good if it is not secondary to a concerted effort to control our greenhouse gases. It is also true that there are difficult ethical and governance issues involved. But we cannot reason about what we don’t understand. We cannot improve our techniques if we do not learn about them. Given the rate at which we are reducing our use of fossil fuels and the low cost of early experiments, I believe we should aim to better understand these options.

Notes and References
0. I’d like to thank Professor Emeritus Thomas Ackerman for taking the time to talk with me. It was really interesting to learn from someone who thinks intuitively and knowledgeably on a planetary scale, but also has a nuanced grasp of the ethical issues involved with geoengineering.

1. The source for this definition is this introduction to a special issue on geoengineering in the journal Climatic Change from November 2013.

2. Geoengineering is a pretty broad term. The category typically includes techniques that are designed to reflect incoming radiation (“solar geoengineering”), such as injecting aerosols into the stratosphere or using sea spray to brighten marine clouds. Solar geoengineering can take many forms. Other examples include generating microbubbles on the ocean’s surface (like a ship’s wake) to make the ocean more reflective, pumping water over the ice caps to fortify them and enlarge their reflective surface, and distributing tiny reflective glass beads over eroding ice sheets to keep them cool and intact. The term “geoengineering” is used less frequently to refer to techniques that remove CO2 from our atmosphere, such as enhanced rock weathering, iron fertilization of the ocean, or direct air capture. It is generally accepted that some form of carbon removal will be necessary to hit our climate goals, though we don’t know yet how to do it at scale. Geoengineering typically does not refer to widespread planting of trees, though that also fits the definition of a “deliberate large-scale manipulation”.

3. Ackerman mentioned that reflecting radiation is the primary but not theoretically the only way to reduce incoming radiation. For example, incoming radiation could be reduced by moving the Earth ever so slightly farther from the Sun. We can all be grateful that no one is working on that yet…

4. You can see a great graph of where heat is absorbed in the fifth section of this earlier blog post.

5. The increased reflectivity comes from two effects. The first is cloud formation — a larger, brighter, and longer-lasting cloud reflects more sunlight. The second is reflection off of the ocean spray particles themselves. In areas where there is less cloud cover, such as over the Great Barrier Reef or the Caribbean, this spraying technique may still have a useful impact on temperature, though Ackerman is dubious due to the very short lifetime of the aerosol particles at the marine boundary layer.

6. Stephen Gardiner, a professor of Philosophy and Ethics at the University of Washington, warns that this delay could even be intentional: “If catastrophe may be coming quickly, then even a buck-passing generation has reason to do something, and short-term fixes (i.e., those good for a generation or two) are sure to be attractive, especially if they appear to have low startup costs and to impose most of their risks on others (e.g., in the further future or in other parts of the world).”

7. You can find a good interview with Alan Robock on the topic of geoengineering here.

8. It can be surprisingly hard to find material in the stratosphere once it has been released. As Ackerman explains “It’s hard to find a contrail of a plane you are in. The wind blows. It’s like dropping something in the ocean. The atmosphere is 3D, so it’s a lot harder than looking on the surface of the ocean.”

9. I recommend Stephen Gardiner’s paper “ The Desperation Argument for Geoengineering” to understand some of the moral and ethical dilemmas involved. It is a lengthy response to a quip that Nathan Myhrvold made. Gardiner questions whether impacted countries can meaningfully consent to these experiments, asks whether an “arms race” of geoengineering might not be worse than climate change itself, and emphasizes that the countries responsible for getting us into this mess should work hard to clean it up, rather than justify a potentially harmful geoengineering bandaid based on the evolving climate emergency. Here is an example he gives from that paper:

“Suppose Josef sets up a lab in his house and pumps the waste into the house next door. He then discovers that the waste is toxic. Josef recognizes that this poses a moral problem and that he has a number of options to address it. For example, he could cease his experiments, redirect his ventilation system, filter his emissions, and so on. However, Josef is unwilling to do any of these things and offers no adequate reason for his unwillingness. Unfortunately, his victim, Karla, can do very little about it. Josef is much stronger, richer and better armed than Karla, and the local police force and courts are in Josef’s pocket. Eventually, Karla becomes sick and desperate, and appeals to Josef for help. Josef says he will help, but this requires Karla agreeing to become an official subject of Josef’s experiments. As such, Karla must put himself completely in Josef’s power, and allow Josef to directly administer whatever treatment to his body and mind that Josef sees fit. In my view, this situation is morally horrifying, and desperation arguments do not help.”

10. The ENSO climate pattern illustrates this kind of complexity, where sea surface temperatures in the eastern Pacific can impact precipitation (via La Niña or El Niño) in areas as far flung as Indonesia, Peru, and Alaska.

11. Some of you may be familiar with cloud seeding to encourage precipitation, for example with silver iodide. Ackerman distinguishes it from marine cloud brightening, indicating that it is less reliable because rainmaking works on heterogeneous, unstable cloud systems, and also because it is harder to affect precipitation than it is to affect reflectivity.

Current Climate Data (April 2020)
Global impacts, US impacts, CO2 metric, Climate dashboard (updated annually)

Comment Guidelines
I hope that your contributions will be an important part of this blog. To keep the discussion productive, please adhere to these guidelines or your comment may be moderated:
– Avoid disrespectful, disparaging, snide, angry, or ad hominem comments.
– Stay fact-based and refer to reputable sources.
– Stay on topic.
– In general, maintain this as a welcoming space for all readers.