4 ms·
A U.S. nuclear warplanner from around a decade ago (I can link you to him if you're interested) recently stated that he'd suspect that a typical full exchange b
by chakalakasp 5y ago
A U.S. nuclear warplanner from around a decade ago (I can link you to him if you're interested) recently stated that he'd suspect that a typical full exchange between the U.S. and Russia would kill 95% of the people living in both countries within 18 months of the attack. Main causes of death would be complete collapse of infrastructure; no supply chains, no food production, no fuel, no heat, no power, no water, no medicine or health care. Most major cities would be destroyed, many cities along rivers would be under fast-flowing as water dams would be targeted to cascade fail most hydrology management, anything downwind of spent fuel storage for nuclear plants (which, in the U.S., are stored above ground near the actual reactors) would be uninhabitable for months to decades, etc. It'd be a very messy 18 months, too, as those left alive immediately after that attack and fallout, which would probably be around 2/3rds of the population of each of the countries, competed for limited resources. And his estimates assumed that nuclear winter was not a real thing. It goes downhill from there if it is.
- justinpowers 5y agoPlease do provide that link
- chakalakasp 5y agoThis is the guy: https://www.linkedin.com/in/dmteter https://www.linkedin.com/in/dmteter He's an engineer who's job was, in the past, to be one of the people who developed the U.S. nuclear warplans against Russia. His open source attempt at using his understanding to build hypothetical target laydowns for Russian attacks against the U.S. is here: https://github.com/davidteter/OPEN-RISOP https://github.com/davidteter/OPEN-RISOP