Hacker News Reader

Katrina was a "once a hundred year" storm. Likewise Sandy. Now Harvey is a "once in 500".

All within 12 years. I think we should stop measuring against data from a different global climate.

A "hundred year storm" doesn't mean we expect one per hundred years globally, but one per hundred years locally.

Even that is a little misleading (given how bad our intuition about statistics is). A '100 year flood' is flood that has a 1% chance of occurring in any given year (it's independent of what happened 1, 5, or 50 years ago).

Calling it a '100 year flood' makes people think 'oh, good, now we shouldn't have one of those for another ~100 years'. In reality, the odds of a flood of the same magnitude occurring next year are exactly the same as they were this year.

Grady (of the Practical Engineering YouTube channel) has a great video on this topic: https://www.youtube.com/watch?v=EACkiMRT0pc

(Grady is in the Austin area, so hopefully he hasn't too much 'practical' experience from this storm...)

The other thing to remember is a 5 year flood and a 100 year flood are often not very different in size.

Reply on news.ycombinator.com