This scenario you mention about global warming is still an assumption that relies at least somewhat on probabilities. Many scientists believe global warming fro

This scenario you mention about global warming is still an assumption that relies at least somewhat on probabilities. Many scientists believe global warming from CO2 emissions is highly likely, while others don't (they tend to be labeled as skeptics). Though there are too many variables that we as humans don't fully understand. So even if it is true and it makes some land unusable, it may well make other areas such as existing desert usable. So I don't really know that this makes a tiny probability of a nuclear accident all that acceptable. That said, I'm not against nuclear energy, but the probability of accidents needs to be next to nothing. Build enough reactors near fault lines and the probability of one of them receiving an 8.9 earthquake would surely approach certainty.