How many more once in a lifetime storms are we going to have this year?, page-66

I don't think that our communications were that good until the seventies. We didn't have comms satellites until then, or weather satellites. And the big causes of high death rates were floods and drought - I don't think building codes would have had much impact there.

"Everything gets reported and recorded these days but many events could have slipped through unrecorded at that time." I agree that a lot of events would be because reporting is much better - and that people are living (but not dying) where events happen. We had an extreme event a few years ago, very small in area affected but several cars died, quite a lot of damage to housing and it would certainly have been a blip on the insurance companies radar. Thirty years ago it was just bushland and nobody would have known about it. So the reporting of events almost certainly understates the situation a hundred years, but if you look at Rememdium's chart you'll see that the absolute death rate from natural disasters collapsed from 500,000 p.a in 1900 to about 15,000 in 2010. In spite of the world population more than quadrupling in that time.

If anyone's got a better explanation for such an improvement in deaths due to natural disasters I'd be interested. I can't see any other explanation that the disasters are nowhere near as severe as they used to be. rememdium blames the loss of life on climate change (his second post) but if climate change has anything to do with natural disaster death rates, it can only be for the phenomenal good of the population, not to its detriment.