Luisa Rodriguez suggests the most important aspect of nuclear conflict (with respect to chance/severity of nuclear winter and therefore chance of x-catastrophe) is targeting strategy.
Countervalue, the targeting of cities, is really bad re. nuclear winter because cities produce a lot of black, sooty smoke when burnt.
Counterforce, the targeting of out-in-remote-areas nuclear missile launchers/platforms (e.g., silos),[1] is really not bad re. nuclear winter, because little sooty smoke is produced when silos and surrounding areas are detonated on.
(Counterforce is also not bad re. direct fatalities from detonations, because not many people are present out in these remote areas.)
Intervention for reducing nuclear x-risk: influence nuclear weapons states’ governments to do less countervalue.
(This talk of Luisa’s was based on her US-Russia nuclear war post sequence − 1, 2, 3, 4 - so it’s surprising that Luisa emphasises this point about targeting strategy much more in her talk than in her posts. Perhaps she updated on the importance of targeting between writing the posts and giving the talk.)
“You would get nuclear winter [...] at under 100 cities bombed, probably between 50 and 100. And I should also say that’s [number of] nuclear detonations, not necessarily that many cities. But if there were 50 to 100 bombs—ish—dropped on cities [...] That’s the level we’re talking about, leading to severe nuclear winter.” (21:15-21:50)
I’m actually more optimistic than Luisa on this. In my own Guesstimate modelling, I’ve found 200 as roughly the number of detonations on cities that’d result in severe nuclear winter and thus threaten x-catastrophe.
The severe nuclear winter definition I’m working with is: “Almost everyone outside of Australasia would die of starvation,[2] assuming minimal response in agriculture/food production methods (response examples include deploying hardier crops, farming edible algae, and substantially increasing farming in the tropics).”
The difference between my and Luisa’s numbers could just come down to us working with different definitions of “severe nuclear winter.”
Further comments
Shortly before giving this EA Global talk, Luisa wrote a great four-part series of posts on US-Russia nuclear war. These posts, in my preferred order of reading,[3] are:
Despite these posts being awesome,[4] I’m not going to include them in my Nuclear Risk 101 sequence because:
They’re already on the EA Forum, and
Luisa’s talk that I’ve summarized in this post is itself a summary of her US-Russia posts.
(Edited to add: I’ve since withdrawn my Nuclear Risk 101 sequence, choosing instead to just tag my nuclear-related posts and linkposts to the nuclear warfare topic page.)
[EAG talk] The likelihood and severity of a US-Russia nuclear exchange (Rodriguez, 2019)
Link post
Linkposting, tagging and excerpting in accord with ‘Should pretty much all content that’s EA-relevant and/or created by EAs be (link)posted to the Forum?’.
Will’s summary and takes
Luisa Rodriguez suggests the most important aspect of nuclear conflict (with respect to chance/severity of nuclear winter and therefore chance of x-catastrophe) is targeting strategy.
The two main types of targeting strategy are countervalue and counterforce.
Countervalue, the targeting of cities, is really bad re. nuclear winter because cities produce a lot of black, sooty smoke when burnt.
Counterforce, the targeting of out-in-remote-areas nuclear missile launchers/platforms (e.g., silos),[1] is really not bad re. nuclear winter, because little sooty smoke is produced when silos and surrounding areas are detonated on.
(Counterforce is also not bad re. direct fatalities from detonations, because not many people are present out in these remote areas.)
Intervention for reducing nuclear x-risk: influence nuclear weapons states’ governments to do less countervalue.
(This talk of Luisa’s was based on her US-Russia nuclear war post sequence − 1, 2, 3, 4 - so it’s surprising that Luisa emphasises this point about targeting strategy much more in her talk than in her posts. Perhaps she updated on the importance of targeting between writing the posts and giving the talk.)
“You would get nuclear winter [...] at under 100 cities bombed, probably between 50 and 100. And I should also say that’s [number of] nuclear detonations, not necessarily that many cities. But if there were 50 to 100 bombs—ish—dropped on cities [...] That’s the level we’re talking about, leading to severe nuclear winter.” (21:15-21:50)
I’m actually more optimistic than Luisa on this. In my own Guesstimate modelling, I’ve found 200 as roughly the number of detonations on cities that’d result in severe nuclear winter and thus threaten x-catastrophe.
The severe nuclear winter definition I’m working with is: “Almost everyone outside of Australasia would die of starvation,[2] assuming minimal response in agriculture/food production methods (response examples include deploying hardier crops, farming edible algae, and substantially increasing farming in the tropics).”
The difference between my and Luisa’s numbers could just come down to us working with different definitions of “severe nuclear winter.”
Further comments
Shortly before giving this EA Global talk, Luisa wrote a great four-part series of posts on US-Russia nuclear war. These posts, in my preferred order of reading,[3] are:
How likely is a nuclear exchange between the US and Russia?
Would US and Russian nuclear forces survive a first strike?
How many people would be killed as a direct result of a US-Russia nuclear exchange?
How bad would nuclear winter caused by a US-Russia nuclear exchange be?
Despite these posts being awesome,[4] I’m not going to include them in my Nuclear Risk 101 sequence because:
They’re already on the EA Forum, and
Luisa’s talk that I’ve summarized in this post is itself a summary of her US-Russia posts.
(Edited to add: I’ve since withdrawn my Nuclear Risk 101 sequence, choosing instead to just tag my nuclear-related posts and linkposts to the nuclear warfare topic page.)
Launchers include silos and (on submarines) tubes, rocket launchers, and artillery. Platforms include tanks, aircraft, and ships.
For further discussion on Australasia’s prospects in nuclear winter, see Aird (2021).
My preferred order is not chronological.
The fourth post is especially awesome: to the best of my knowledge, it’s the only end-to-end model of risks from nuclear winter in existence.