Whoever does it, and whichever fuel they use, the success of deep decarbonization will hinge on technological progress. Why assume that the know-how of 2018 is the best the world can do? Decarbonization will need breakthroughs not just in nuclear power but on other technological frontiers: batteries to store the intermittent energy from renewables; Internet-like smart grids that distribute electricity from scattered sources to scattered users at scattered times; technologies that electrify and decarbonize industrial processes such as the production of cement, fertilizer, and steel; liquid biofuels for heavy trucks and planes that need dense, portable energy; and methods of capturing and storing CO2.
The last of these is critical for a simple reason. Even if greenhouse gas emissions are halved by 2050 and zeroed by 2075, the world would still be on course for risky warming, because the CO2 already emitted will remain in the atmosphere for a very long time. It’s not enough to stop thickening the greenhouse; at some point we have to dismantle it.
The basic technology is more than a billion years old. Plants suck carbon out of the air as they use the energy in sunlight to combine CO2 with H2O and make sugars (like C6H12O6), cellulose (a chain of C6H10O5 units), and lignin (a chain of units like C10H14O4); the latter two make up most of the biomass in wood and stems. The obvious way to remove CO2 from the air, then, is to recruit as many carbon-hungry plants as we can to help us. We can do this by encouraging the transition from deforestation to reforestation and afforestation (planting new forests), by reversing tillage and wetland destruction, and by restoring coastal and marine habitats. And to reduce the amount of carbon that returns to the atmosphere when dead plants rot, we could encourage building with wood and other plant products, or cook the biomass into non-rotting charcoal and bury it as a soil amendment called biochar.96
Other ideas for carbon capture span a broad range of flakiness, at least by the standards of current technology. The more speculative end shades into geoengineering, and includes plans to disperse pulverized rock that takes up CO2 as it weathers, to add alkali to clouds or the oceans to dissolve more CO2 in water, and to fertilize the ocean with iron to accelerate photosynthesis by plankton.97 The more proven end consists of technologies that can scrub CO2 from the smokestacks of fossil fuel plants and pump it into nooks and crannies in the earth’s crust. (Skimming the sparse 400 parts per million directly from the atmosphere is theoretically possible but prohibitively inefficient, though that could change if nuclear power became cheap enough.) The technologies can be retrofitted into existing factories and power plants, and though they are themselves energy-hungry, they could slash carbon emissions from the vast energy infrastructure that is already in place (resulting in so-called clean coal). The technologies can also be fitted onto gasification plants that convert coal into liquid fuels, which may still be needed for planes and heavy trucks. The geophysicist Daniel Schrag points out that the gasification process already has to separate CO2 from the gas stream, so sequestering that CO2 to protect the atmosphere is a modest incremental expense, and it would yield liquid fuel with a smaller carbon footprint than that of petroleum.98 Better still, if the coal feedstock is supplemented with biomass (including grasses, agricultural waste, forest cuttings, municipal garbage, and perhaps someday genetically engineered plants or algae), it could be carbon-neutral. Best of all, if the feedstock consisted