Supervolcanoes and Geo-engineering

Super-volcanos, geoengineering and global warming, but what about global cooling?

The latest IPCC report sets out a stark assessment of what needs to be done by 2050 to avoid a 1.5C temperature rise. Not only do we need to completely decarbonise all sectors of our economy, we also need to put carbon emissions into reverse, potentially by as much as 10 Giga-tons/annum, that is 10 times the amount of CO2 emitted by global aviation!

So the gauntlet has been laid down for carbon capture and there are some really promising developments in this space. For instance Carbon Engineering, a start-up backed by Bill Gates, claims to have a pathway to deliver carbon removal at a cost of $100/ton at scale. This would be the equivalent of paying a $220 dollar surcharge for heating a home with gas in the UK each year to remove its CO2 production.

However, there are serious question marks around our ability to scale these initiatives with sufficient pace. Particularly as economies continue to lock in future carbon emissions. For example, in 2021, China, had 176GW of new coal power plants under construction. That's more than twice the generation capacity of the UK's entire power grid. Meanwhile in India, where half of the country's electricity comes from coal, there were 28 coal power plants under construction (~40GW) as of last year.

These challenges bring the question of geo-engineering into focus. One notable idea in the world of geo-engineering is Solar Radiation Management (or SRM). The principle is quite simple. We inject aerosols from an aircraft into the atmosphere to reflect sunlight back into space. This idea has in part been inspired by the observation of reduced global temperatures in the aftermath of eruptions from super volcanos. A recent example is Mount Pinatubo, which in 1991 erupted, throwing ash into the atmosphere causing global temperatures to drop 0.6 degrees C for nearly two years due to global dimming. A 2018 Nature Communications paper estimates that to avoid catastrophic climate change, we might need to dim the skies by anywhere between 1W/m2 and 5W/m2 depending on which carbon reduction pathway the planet goes down (see Figure 3).

To put that into perspective, the average solar radiation across the Earth is 1,361W/m2. So reducing the brightness of the sky by ~0.3% shouldn't do much harm should it? Well, on the one hand SRM could buy us time to avoid excess deaths from heat-strokes in developing countries, forest fires and rising sea levels; on the other, it likely that the net effect on agricultural yields would be neutral. In other words we'd still lose the same amount of crop because reduced sunlight would lead to lower yields which would offset the plants saved by avoiding higher temperatures. This was the conclusion of another interesting paper which can be found in the following link. We would also fail to deal with the problem of ocean acidification which poses a threat to marine life, in particular animals with exo-skeletons such as crabs and corals. This problem ushers in yet another frontier of geo-engineering: 'Ocean Ph Control'. Though a very nascent field of research, a comprehensive assessment (in the following link) suggests that trillions of tonnes of CO2 could be sequestered into the ocean whilst taking the Ph back up to pre-industrial levels by accelerating the weathering of carbonate and silicate rocks whilst pumping lime into the ocean.

If we therefore assume that a geo-engineer can buy us time with solar radiation management whilst offsetting the harm of CO2 emissions by dispensing giant Rennie tables into the sea, the only problem that remains is maintaining food production levels in the face of reduced sunlight levels. If, in the worst case, 5W/m2 of artificial LED lighting is required. Then we can estimate the amount of additional electricity production to offset the reduced sunlight levels under the worst SRM scenario. There's around 25 million km2 of agricultural land in the world. Assuming we don't change our diets, we'd need a huge amount of additional electricity to maintain today's crop outputs. Most likely it would be more than all the electricity currently consumed today. Alternatively, we would need to convert more land into agriculture through the reversal of desertification and terraforming places which are currently inhospitable.

Figure 1: Yellowstone Supervolcano

Figure 2: IPCC carbon pathways to keep within 1.5C

Figure 3: How much sunlight do we need to reflect across different emission reduction scenarios (https://www.nature.com/articles/s41467-018-05938-3)

Figure 4: Offsetting ocean acidity by encouraging increased weathering of carbonate and calcite rocks and discharging lime into the oceans

So in summary, geo-engineering is not a simple fix. The above considerations gloss over many other problems, particularly around the politics of getting consent throughout the world on how to control 'Earth's thermostat'. SRM is likely to cause significant changes to rainfall patterns and worse still, these changes might be inherently unpredictable. There are however, in my view, two key reasons justifying further research into geo-engineering and perhaps even experimental trials:

1. Not withstanding the side effects, there is a real risk that if we don't do geo-engineering, people could end up enduring severe heat stress to the point of a humanitarian catastrophe

2. As a species, global warming isn't the only existential risk that might need an answer in the form of geo-engineering.

On this latter point, we only need to think about the potential fallout (pun intended) from another volcanic eruption in Yellowstone national park. The last time this happened was over 70,000 years ago. If this happened again, it is estimated that 100,000 people would be instantly killed before the Earth is thrown into a mini ice age until the dust settled (another pun intended). If we're thinking about geo-engineering to avoid global warming, do we also need to consider ways in which we could rapidly warm the world under such circumstances whilst sustaining crop production in dim twilight? This would take considerably more electricity than offsetting SRM. Perhaps we need to ramp up nuclear power production as an insurance policy!