On the Benefits of Pumping Tons of Sulfur into the Stratosphere
What happens when we inject a bunch of sulfuric acid in tiny watery droplets (1/1000 the size of a human hair) 20km above the Earth?
95% of the ideas here are not mine but belong to David Keith. I took them from his 2013 book, A Case for Climate Engineering (there’s no affiliate link, I swear). This post is a brief overview on what I retained from reading his lovely book in one day.
TL;DR
Injecting sulfuric acid droplets in the stratosphere would cause those droplets to scatter sunlight back into space, reducing the warming effect we’re getting from climate change. Why does it need to be sulfuric acid? If we used something like water, it would evaporate too quickly. Sulfuric acid would probably last us about a year before we need to replenish it. Gradually, we would need to ramp up the amount of sulfur we put into the stratosphere. 25,000 tons for the first year. 50,000 tons for the second year. By 2030, we would need to be pumping 250,000 tons of sulfur into the stratosphere annually, if carbon dioxide continues to accumulate.
So why is the dispersal so easy? We just need to use Gulfstream jets to inject to sulfur, which is technology that we already have. It’s cheap, only costing $1 billion annually. This doesn’t do anything to help ocean acidification and the accumulation of greenhouse effects, but it mitigates their warming effects on the planet. Oh yeah, and it damages the ozone layer and contributes to air pollution when it descends into the lower atmosphere. It reduces damage in most regions of the world while making it worse in others.
But at the end of the day, if we continue to increase emissions at the current rate, we would have no choice but to resort to geoengineering to fix our mistakes.
A brief overview of Solar Radiation Management
Keith calls it geoengineering in his book, but geoengineering is a really broad term. What he actually means is solar radiation management (SRM) — a fancy term for “blocking out the sun”. More specifically, stratospheric aerosol injection — blocking out the sun with tiny particles suspended in the stratosphere. To avoid typing such long words, I’ll just use SRM to refer to stratospheric aerosol injection.
The amount of carbon dioxide (CO2) in the atmosphere is about 40 times the amount of garbage in our landfills. Humans move carbon 100 times faster than natural processes (volcanic vents). The amount of carbon is important, but not as important as the rate at which it’s being moved.
The Earth receives sunlight from the Sun, which influences climate. Normally, the Earth would absorb the sunlight and radiate the heat back into space, but the greenhouse gases have an insulating effect on the Earth so that we continue to absorb energy from the Sun, but it’s much harder to radiate it back into space.
SRM intends to change the intensity of sunlight that the Earth receives so that we don’t absorb as much energy and trap more heat.
No amount of SRM is going to change the sunlight intensity so that it’s exactly opposite the effect we get from CO2. We can bring down the temperature, but the climate won’t be the same as one without so much CO2.
Most climate impacts would be local. CO2 suppresses precipitation, but warming increases precipitation. Because the suppression from CO2 is stronger than the increase in precipitation from warming, the net effect would be an overall decrease in precipitation.
So yeah, using SRM to cool the planet would result in a climate with less precipitation, but Keith argues we can just use less SRM to increase rainfall if we need to.
But what about the effect of SRM on monsoons? If used recklessly, SRM can cause the Asian monsoon to fail. However, if used appropriately, SRM can slow or stop harmful increases in precipitation that result from climate change. Keep in mind that reducing precipitation does not mean we’ll cause droughts. Increases in CO2 cause less precipitation and less evaporation.
Extreme climate events such as those droughts depend on the strength of the water cycle. An increase in temperature increases the strength of the water cycle, which increases the number of extreme events. SRM weakens the water cycle, lessening its impact on the climate. We can even use SRM to increase food supply by reducing heat-stress during the early growing seasons.
Uncertainty + Inertia = Danger
Uncertainty
The amount of warming we experience is roughly proportional to how much emissions we pump out. If we begin emitting double the amount of carbon, we warm the planet by a factor of 2. However, we can’t accurately predict warming based on emissions due to uncertainties in the emissions and temperature histories. We’re even more uncertain about the link between carbon emissions and their effects.
Inertia
If we cut all carbon emissions to zero by the end of 2030, there’s still little noticeable reduction in global warming, because all of the carbon from the past century is still in the atmosphere, warming the planet.
This is what Keith is describing when he talks about inertia. Newton’s first law. An object will remain at rest or in uniform motion unless acted upon by an external force. A rolling ball is going to keep rolling unless some outside force is strong enough to stop it.
Cutting emissions to zero would mean cutting out a huge chunk of society’s infrastructure. Coal is the worst offender. It’s terrible for the environment, but so much of society relies on it. Because the inertia of the carbon cycle and human society is so great, we won’t see any benefits from dramatic efforts to decarbonize until late in the century. The ball is too heavy and rolling too quickly to stop.
So why put Sulfur in the Stratosphere?
Keith’s favored method of SRM is stratospheric aerosol injection. By increasing the amount of sulfuric acid aerosol in the stratosphere, we can increase how much sunlight gets scattered, because aerosol particles scatter light. The amount of light scattering is greatest if those aerosols are tenths of a micron in diameter (1/1000 the size of a raindrop). Small particles in aerosols also have the added benefit of falling to earth more slowly than larger ones.
But why are we injecting it into the stratosphere?
We call it the stratosphere because it’s the layer of the atmosphere that’s highly stratified. There’s vertical separation, similar to how stratified rocks are separated into layers as you go up. Air in the stratosphere mixes horizontally, but not vertically.
Compare that to the troposphere, the weather layer. The troposphere mixes all over the place because warm and cool air rises and sinks to give us the weather we all know and love. Because the troposphere is moving so much, aerosols would remain in the troposphere 100x shorter than in the stratosphere, which naturally makes the stratosphere a better place to dump sulfuric acid (it’s also 7–15 km above our heads). Placing the aerosol in the stratosphere also comes with the added benefit of being 100x cheaper (because the aerosols last longer), and reduces the risk of acid rain and pollution. Keith claims that a few tons of aerosol are enough to offset the effects of sunlight from 1 million tons of carbon in the atmosphere.
It’s risky
The first major complaint is ozone loss.
SRM would reduce the number of nitrogenous oxides in the stratosphere. This is bad because nitrogen oxides act as a shield against harmful chlorine molecules that damage the ozone layer. The nitrogen oxides bind to the chlorine molecules to form chlorine nitrate, which doesn’t have any effect on destroying the ozone layer.
Why would SRM even reduce the nitrogenous oxides in the first place? Sulfates turn those nitrogen oxides back into regular old nonreactive nitrogen, which doesn’t do anything to prevent the chlorine from attacking the ozone. Think of it like covering a sword (chlorine) in bubble wrap (nitrogen oxides). Less bubble wrap means there’s a greater chance the sword can break something.
Then there’s also the fact that we’re putting tons of sulfur into the atmosphere.
The 50 million tons of sulfur pollution in the lower atmosphere already contributes to 1 million annual deaths. The good thing about the sulfur in SRM is that there will actually be less pollution because the sulfur is spread out worldwide, instead of concentrated around factories. But it’s still contributing to pollution.
Keith presumes that SRM would save more lives from climate effects than the pollution from sulfur would kill. But still. Not the ideal scenario.
Basically, the benefits of reducing climate risk outweigh the harmful side effects of SRM (acid rain, ozone depletion, pollution)
How can we deploy SRM?
- Theory and lab work
Use computers to model the potential impact, and understand chemical reactions in the lab.
2. Experiments in the atmosphere
Understand how aerosols would affect the stratospheric chemistry. Use small-scale experiments to test processes, not large-scale responses to SRM.
3. Minimal deployment.
Let’s say we get good results from phases 1 and 2. Then it’s time to bring out the smallest scale deployment on which we can detect some kind of response. The purpose is to find any unexpected problems before they have the ability to cause damage. At this stage, we’ll need some kind of rational, independent governance (blockchain, anyone?). We’ll also need new small-scale experiments and lab work to understand the results we get from the minimal deployment.
4. Gradual deployment
The key word is gradual. We want a gradual ramp up if the results from the first three phases prove to be beneficial. Keith estimates it’ll realistically take us till 2035 to get to stage 4.
Of course, the hard technical part isn’t necessarily the dispersal. It’s developing sensors and observation tools to monitor the effects of SRM and making sure we don’t end up accidentally ruining the planet.
Taking inspiration from volcanoes
To inject sulfur dioxide into the atmosphere, it must first oxidize to form sulfuric acid so that the molecules are small enough to condense and form liquid droplets of sulfuric acid that form an aerosol.
Volcanoes inject sulfur dioxide into the atmosphere to create an aerosol layer. Those volcanic particles are about 1/10 of a micron in diameter — very effective at scattering sunlight.
If you straight up pump sulfur dioxide into the stratosphere, we won’t be able to get those droplets small enough. The particle would just combine with the existing particles in the atmosphere and form large droplets that won’t be as effective at scattering sunlight.
Solution: release sulfuric acid as a vapor from an aircraft.
As a vapor, the particles would be less likely to combine with existing droplets, because the concentration of sulfur is so high in the vapor that it just condenses into tiny droplets.
It’s not that expensive to deploy
We need to inject a buttload of sulfur into the bottom of the stratosphere in the tropics, about 20 km above the surface of the Earth.
Gulfstream business jets are able to get that high. Retrofit them with low-bypass military engines so that they’re able to carry the sulfur payload, and we’ve built our deployment mechanism!
We can get a fleet of 20 for about $1.5 billion. Most of the technology for deploying SRM doesn’t exist, but they’re really easy to build quickly.
If we’re in the full deployment phase, we can build a small fleet of custom aircraft (basically the business jet with much longer wings).
One proposed deployment technique is to attach a hose to a stratospheric balloon and use that hose to spray sulfur into the stratosphere. This method is cheaper, but hoses don’t disperse well horizontally. Cost isn’t really a problem either, as it’s not as socially relevant as understanding risk and efficacy.
What about naval guns and rockets? They’re awesome, but also much more expensive than planes, so we might as well just use the planes.
Total cost: $1 billion annually.
That’s chump change for the world’s richest people. Costs are so low that it shouldn’t be a major factor in decision making.
Limitations
SRM doesn’t do anything to help with ocean acidification and the accumulation of carbon in the atmosphere. Right now, the best solution (according the Keith) seems to be a combination of SRM and cutting emissions.
Sulfates are also a crude and cheap way of dealing with warming. Hopefully, we can cheaply engineer something better than sulfates in the future.
We can use nano-structured particles to scatter more sunlight and engineer them to adjust how much light they use to influence the climate. Maybe we can even use them to avoid some of the negative side effects of sulfates. But they’re so expensive to develop and maintain. Keith has an interesting idea of leveraging the photophoretic effect to keep the engineered particles in the air, but you can read about that in the book.
Marine cloud brightening is a really promising technique. The basics involve taking a bunch of ships into the ocean and making them spray sea salt into the sky to make the clouds whiter and reflect more sunlight.
However, we don’t have any accurate models of how sea salt would affect the clouds. It might actually decrease reflectivity by breaking up the cloud deck. And we need to keep those ships out there 24/7 for years if we want those clouds to stay. The good thing is that marine cloud brightening is easy to control. Shut it off and the clouds go away in a few days.
Speaking of control, how should we control SRM?
Should we have a slow or a rapid response?
Sulfates would persist in the stratosphere for about a year. That means if we make a decision to shut them off, it’ll take a year for those sulfates to get out of the atmosphere. On the other hand, if we used something like space reflectors, those can be turned on and off in less than an hour.
Using a slow response might be better, as fast responses won’t work well with current politics. Politicians would have a tendency to overreact and turn on an untested system too quickly, or overreact to the side effects of a SRM system and turn it off too quickly, increasing climate risk.
We can’t really control it regionally either, because local climates are interconnected. And that brings a whole passage about international dynamics that you can read about in the book.
Some thoughts on ethics and politics
Keith describes something called moral hazard. If an individual is shielded from bad outcomes of a risky but desirable activity, they take on more risk than they would accept if they were paying the full cost. It’s like getting a free pass, except others are paying the price.
Some people say we should avoid SRM in the short term because of its risks. However, society is actually worse off in the long run, because the costs of climate change would outweigh the risks of holding off on implementing SRM.
An important factor that should be considered is risk compensation. Risk compensation is a change in behavior towards increased risk exposure after that risk has been reduced by some technical fix. This is happening now. We anticipate that we’ll eventually solve climate change in the future, so we have no problem dumping more carbon in the atmosphere.
We increase carbon emissions (change in behavior towards risk exposure), because we believe that we’ll eventually solve climate change in the future (risk reduced by some technical fix).
An argument against SRM would be to forgo the SRM Band-Aid altogether and allow society to tough it out while focusing on long-term solutions, like cutting emissions.
Cutting emissions reduces risk for the future, but doesn’t do anything for us right now. It’ll take at least a century until cutting emissions will have any noticeable effect. On the opposite end of the spectrum is SRM. It reduces risk right now but doesn’t do anything in the long run.
Keith has an interesting theory about social classes and climate change.
Most emissions come from rich people. The burden of climate change would mostly fall on poor people, as they are the most vulnerable to changes in the climate.
Therefore, the poor would benefit the most from SRM, and yet the people who argue against SRM in favor of emissions cuts and social change are… rich people.
Basically, fixing climate change, whether it be through cutting emissions or geoengineering, means extending our moral compass out towards people in the future that we don’t know, can’t see, and who probably won’t care about us. No amount of technical fixes will completely solve climate change without collective action from humans.
Keith goes into much more detail than I do in his book. If you have nothing to do, I recommend at least checking it out. It’s a fun, lightweight introduction to solar radiation management through stratospheric aerosol injection.
