Super pollutants are trendy, but we should be careful how we use them
Substituting short lived climate pollutants like methane for CO2 as carbon offsets risks breaking the math of net zero
“Super pollutants” – short-lived climate pollutants like methane (CH4) and some refrigerants (halocarbons) – are having a moment. There were numerous sessions on the topic during the recent New York Climate Week, and a number of companies are exploring investments in reducing these emissions as part of their climate goals.
Reducing emissions of short-lived climate pollutants (SLCPs) is, by itself, an unambiguously good thing. Methane in particular is responsible for around a third of all warming to-date from well-mixed greenhouse gases in the atmosphere, and reductions in emissions can have a rapid cooling effect on the planet.
It is when methane (or other SLCPs) are used to offset or neutralize CO2 emissions – to make a claim that the climate effects of CO2 can be counterbalanced by methane – that the problem becomes much, much thornier. As Ray Pierrehumbert explains, “It is useful to reduce methane, but it’s not going to really help us towards net zero. The only real solution to the climate crisis is to get carbon dioxide emissions down to as close to zero as we can.”
Stocks vs flows
The question of how to compare methane and CO2 is one that has long interested me. I wrote a paper a decade ago on how to compare the climate impacts of coal and natural gas (back when talk of a “natural gas bridge” was in vogue), and authored the chapter on methane and other short-lived climate pollutants for Greta Thunberg’s Climate Book.
At its core, the difference in climate impacts between CO2 and methane comes down to the fact that CO2 is a “stock pollutant” and methane is a “flow pollutant”.
CO2 is an extremely stable molecule that accumulates in the atmosphere over time with constant emissions; while a portion of CO2 can be absorbed by land and ocean sinks in the form of organic or inorganic carbon, it does not naturally degrade. The warming that results from CO2 is – to a first order approximation – a largely time-invariant function of cumulative emissions. If CO2 emissions increase, the world warms faster; if they stay constant the world warms at a constant rate; if emissions decline, the world warms more slowly. But even if CO2 emissions get to zero, the world does not meaningfully cool back down for centuries to come; the only way to cool the planet through CO2 is to go net-negative – remove more CO2 from the atmosphere than we are adding.
Methane, on the other hand, is an unstable molecule that interacts with hydroxyl radicals in the atmosphere and oxidizes into CO2 and water (after a long series of chemical reactions). This instability means that methane does not accumulate in the atmosphere over longer time periods unless emissions increase. Rather, the concentration of methane in the atmosphere – and its resulting warming – is a function of the rate of emissions rather than the cumulative emissions.
This has a few important implications. First, we are stuck with warming from CO2 more or less forever; we would have to remove CO2 we previously added to cool down the climate. We can, however, undo past warming from methane emissions by simply cutting emissions. This makes methane both a powerful lever to have a strong near-term climate effect, but also a potentially dangerous distraction if we prioritize it over CO2 as we trade short-term benefits for longer-term harms.
Why cows are like closed power stations
To better understand the difference between CO2 and methane, I’m going to borrow an example from Dr. Michelle Cain of why cows are like closed power stations. It is an imperfect but interesting example of the important differences between CO2 and methane. Let’s take a look.
Let’s assume there is a rancher named Jane. Her family has had a herd of 1,000 cows for the past 30 years. Each day these cows happily graze, eating grass and burping out methane that mixes with the atmosphere. However, the methane in the atmosphere is constantly oxidizing and breaking down. The average lifetime of the methane emitted from her cows is around 10 years.
This means that while Jane’s herd produces around 100 tons of methane per year (0.1 tons per cow), a similar amount of methane emitted by their predecessors is breaking down, and the amount of atmospheric methane remains unchanged as long as the herd size stays constant. The total amount of atmospheric methane from Jane’s cows over the long term depends on how much is emitted each year – not the sum of emissions over time.
If Jane adds a new cow to the herd atmospheric methane will increase by one ton (each cow’s annual emissions of 0.1 tons remains in the atmosphere for ~10 years, so 0.1 tons / year * 10 years = 1 ton). If Jane removes a cow atmospheric methane will decrease by 1 ton.
Now, Jane’s town has a very small 1MW coal powerplant that powers the 500 or so homes. This coal powerplant generates 10,000 MWh of electricity and emits around 10,000 tons of CO2 each year. It turns out that 10,000 tons of CO2 has approximately the same warming effect as 100 tons of methane – if both remain in the atmosphere (as methane is a bit more than 100x stronger than CO2 while it’s in the atmosphere). So are Jane’s cows as bad for the climate as the coal fired powerplant, as this simple math would suggest?
As long as Jane’s herd isn’t growing, the methane emitted is perfectly balanced out by previously emitted methane breaking down in the atmosphere. The same is not true for the CO2 emitted from the coal powerplant, however! Each year about half the CO2 emitted by the coal powerplant remains in the atmosphere – with about half being absorbed by land and ocean sinks over time. So while Jane’s cows add 0 additional methane to the atmosphere, the coal plant adds 5,000 tons of CO2 every year.
The warming effect of the coal powerplant is the same as if Jane were adding an additional 50 cows to her herd each year. This is because 50 cows increase atmospheric methane by 50 tons, and methane is ~100x stronger than CO2, so it’s approximately the same as 5,000 tons of CO2. The next year the town decides that it would be cheaper to generate their electricity with zero-carbon solar + storage. They close down the old coal power plant. However, the carbon that was previously emitted by the coal fired powerplant remains in the atmosphere.
This means that the closed coal powerplant is perpetually warming the planet just as much as Jane’s 1,000-cow herd – even though it’s no longer emitting any CO2! If Jane were to decide to get out of the ranching business (and her herd all became hamburgers), their methane emissions would fall to zero and all the methane they had ever emitted would be gone from the atmosphere in a decade or two.
This story highlights the critical distinction between CO2 and methane: CO2 accumulates in the atmosphere, and once we emit it we are stuck with it (barring actively sucking it out of the atmosphere). Methane does not accumulate over the long term; the amount of methane in the atmosphere depends on the rate of emissions rather than the total amount that has ever been emitted. Both are important greenhouse gases, but they behave in very different ways that we need to account for when planning how to mitigate each.
The problem of simple metrics
The problem with methane and CO2 is that they are not directly comparable. Any framework that adds them together necessarily comes with a set of value judgements – what metrics to use to compare warming impacts, and what timeframe to compare them over.
Governments, corporations, and a large part of the scientific community have settled on a framework known as “global warming potential” (or GWP) to compare CO2 to methane (and other greenhouse gases). GWPs are defined as the amount of CO2 needed to result in the same average radiative forcing as a ton of a different greenhouse gas over some specified time horizon – generally 20, 100, or 500 years, with 100 years (abbreviated GWP100) being the most common.
For methane, a GWP100 approach results in an equivalency of between 28 and 34 (e.g. a ton of methane is equivalent to 28 to 34 tons of CO2), with the lower end of the range reflecting the direct radiative forcing of methane and the upper end including effects of methane on atmospheric chemistry and the lifetime of other greenhouse gases like N2O. Using a shorter GWP20 timeframe gives you an equivalency of around 84x CO2. Greenhouse gases are then added together into a single unit of CO2-equiviant (or CO2e), which is used to report emissions and assess progress on climate goals.
However, CO2e does not have any physical meaning in the real world. It does not tell you how much your emissions – or mitigation – will influence global temperatures or other climate impacts. And it conflates near-term warming with longer-term warming.
To illustrate this, let’s look at the climate impact of one million tons (one megaton) of CO2 emitted today, and compare it to the GWP100-equivalent amount of methane (e.g. 29.4 kilotons of methane).
The math of GWP assumes that the red line and the grey line represent the same climate impact. It should be pretty obvious that they do not. If a company or country were to decide to offset a ton of CO2 emissions with a ton of methane abatement using a GWP100 approach, the actual temperature effect would be to trade short-term cooling of the climate for long-term warming, as shown in the figure below:
Global warming potentials are, effectively, a way to sneakily discount the future, to prioritize near-term benefits relative to longer-term harms. There are some fascinating papers in the economic literature that very explicitly work out what a particular GWP time horizon implies as the discount rate used – and find that GWP20 translates to a discount rate of around 14% while a GWP100 translates to a discount rate a bit above 3%.
There is an argument that future damages should be discounted, though even the most fervent proponents of discounting the future would be reluctant to use a rate as high as 14%. And the idea of discounting in this context is fundamentally inconsistent with the approach taken in the Paris Agreement, which seems to stabilize global temperatures at well-below 2C rather than to maximize net present value at the cost of potential future harms. If we want to argue for discounting, we should be more explicit about it rather than hiding it behind misleading equivalency metrics.
Methane, what is it good for?
If methane and other short lived climate pollutants are fundamentally different from CO2, how should they be used in the context of mitigation?
This has been an area of active debate over the past decade, with groups proposing (and critiquing) alternative metrics like GWP* that try and more accurately compare the temperature impacts of different gases. Ultimately the community has largely settled on the idea of setting separate targets for CO2 and SLCPs, avoiding the need to conflate the two directly.
However, there are still a number of ways for private sector companies to pursue impactful actions on super pollutants that avoid substituting short-term cooling for long-term warming.
The most straightforward is following the idea of “like for like” developed by Prof Myles Allen and colleagues: either directly reduce emissions within your supply chain, or compensate for methane emissions in your supply chain that cannot be easily mitigated by purchasing high-quality additional SLCP abatement credits.
Another option is to support high quality SLCP abatement without making a specific offsetting or neutralization claim. Reducing near-term warming has real quantifiable benefits, and is worth doing for its own merits and not only as a substitute for reducing CO2 emissions.
Finally, offsets using SLCP abatement are not problematic if their short-lived nature is explicitly taken into account (e.g. we don’t pretend they are the same as CO2). This could take the form of “horizontal stacking” where a single ton of CO2 is compensated by a continuous reduction of methane emissions, though it is challenging in practice to create a system that guarantees ongoing activities for the millennial-scale lifetime of CO2-induced warming.
Another option – which companies like Google are exploring – is to use methane and other SLCP abatements as a short-term bridge toward permanent carbon removal. This would work provided that the permanent removal is guaranteed – money is set aside upfront – and the amount of methane is calibrated such the combination of the two has a warming impact equal to or better than immediate permanent carbon removal.
But if the current excitement around super pollutants causes companies to use them as carbon offsets based on problematic GWP conversations, it risks resulting in greater long-term warming. As Pierrehumbert cautioned back in 2014 “Any earlier implementation of SLCP mitigation that substitutes to any significant extent for carbon dioxide mitigation will lead to a climate irreversibly warmer than will a strategy with delayed SLCP mitigation. SLCP mitigation does not buy time for implementation of stringent controls on CO2 emissions.”
I’d be grateful if you could hit the like button ❤️ below! It helps more people discover these ideas and lets me know what’s connecting with readers.
Zeke, this is a great article and comparison between CO2 and CH4 and their time horizons. I think it highlights that time horizons matter, not just for greenhouse gases, but for all aspects of environmental sustainability issues. Thank you!
Thank you for this useful comparison of CO2 and CH4. It's appreciated.