The German magazine "Der Spiegel" has a catastrophe map, depicting which areas would become uninhabitable in case of nuclear accidents: https://www.spiegel.de/wirtschaft/soziales/atomkraft-diese-i...

It's significantly more than a few km, according to that map. It's based on the research of Viennese climate scientists, according to the article.

Dams are a good analogy, here. Dam failures can be catastrophic and far, FAR worse in life lost than even worst-case nuclear failures. The 1975 failure of the hydroelectric Banqiao Dam killed between 100,000 and 230,000 people and destroyed 11 million homes. https://en.wikipedia.org/wiki/Dam_failure

Nuclear power is worth the risk when faced with the threat of climate change. At very least, we must keep existing plants operating until we decommission all fossil fuel power plants.

Compared to the land lost with a 4°C global temperature rise that's a pretty good tradeoff.

I can't read German, but I can read maps and do know a fair amount about radiation.

The first thing to note is that concentration dramatically decreases as radiation is pushed by the wind. We should also note that we're REALLY good at detecting radiation (it has been a great way to spy on other countries' use of atomic weapons, but since we can't always get into their country we need pretty sensitive devices).

But models are one thing, we have real world data. So let's look at a map of Japan, which has active monitoring[0]. The legal public dosage is 1mSv/yr. So to convert that into a number we can read on this map our hourly dosage needs to be below [(1e-3(mSv/Sv)/365(days)/24(hrs)1e9(nSv/Sv)]= 114nSv/hr. You'll actually find that the vast majority is below that. But this doesn't leave much room for other types of radiation (take note of this point).

But let's talk about that 1mSv limit. Why is it there? Is it where danger starts? Radiation workers have a limit of 20mSv/yr (2283nSv/hr, which is only a handful* of places on the map). Why that? It used to be 50mSv/yr (5707nSv/hr). So let's look it up [1] (PDF, see chapter 3). We see a lot of talk about 100mSv/yr (11,415nSv/hr, nowhere on map) or in a single dose being completely safe. We even see mention of no detectable effects at 0.5Sv (3.3 page 56, page 57 even says little evidence of below 1Gy). Well, the thing is we use a model called the Linear No-Threshold (LNT). We use this because we'd rather be safe than sorry (note the use of "assumption" in the paper). The limits given to just workers has an included safety factor. Then it was decided later than the public should have even lower limits (more factors of safety).

Is having such low thresholds a good idea? Definitely. I agree with the attitude of better safe than sorry. So include factors of safety, because we don't know how many chest x-rays you're getting a year or if you live next to a coal plant. But should we take those legal limits as "you're going to get cancer if you exceed them?" Definitely not. We can realistically see that there is little evidence to support that idea. So is it exactly right to say that these places are uninhabitable? I don't think so, but it is safe to say that there is increased risk. But we should also recognize how much risk that is. (The above might also explain why there's a lot of wildlife in Chernobyl)

[0] https://jciv.iidj.net/map/

[1] https://www.researchgate.net/profile/Hanno_Krieger/post/Why_...

Thanks for the link, I had not seen that before. This shows very well how large regions, which are highly populated would be affected by a large nuclear incident in Europe. Even if one assumes, that every endangered citizen could be evacuated without suffering from any direct damange, it would mean that far to many people would lose their housing and that the food production in the affected regions would suffer.