Hacker News Reader

Comment by TuringNYC | Hacker News Reader

I was disappointed to hear the term "100-yr flood" with the latest storem because I remember at least two other such floods in the area in my life -- one was in 2012 (Hurricane Sandy) And one was ~1992 (a Nor'Easter.) So that is three in thirty years. Is our distribution incorrect, or is "100-yr flood" just a media term? Or is it a way to make something sound rare so we can justify unpreparedness?

Of course, all three of these were different things -- one a Nor'Easter, one a Hurricane, and the latest was the effects of a Hurricane down south.

View on HN

It's kind of an abuse of the term. The technical definition is a storm that according to historical records has a 1% chance of happening in any given year. So you'd expect to see it at least once in 100 years. As climate change increases the severity of all types of weather those level of storms will get more common than in the past in a hard to predict way. The storm percentages are also based on the historical record instead of future predictions so it will inherently lag behind the expected increasing storm frequency/intensity.

A probability of 1% per year means 1 - .99^100 = 63.4% probability of at least 1 storm in a hundred years.

The expected value of how many years until such a storm occurs is 100 years.

"A common misunderstanding is that a 100-year flood is likely to occur only once in a 100-year period. In fact, there is approximately a 63.4% chance of one or more 100-year floods occurring in any 100-year period." [1]

Which is different than "You'd expect to see it at least once in 100 years."

However, yes, the expected value of the number of 100-year floods occurring in any 100-year period is 1.

[1] https://en.wikipedia.org/wiki/100-year_flood#Probability

I wish they'd talk about it in these terms. It's easier to think about the cost/benefit of investing in mitigations when faced with "you have an n% chance of incurring m billions in damage this year" vs "x billion dollar damage storms should only occur once every y years".

We do it all the time with e.g. cloud SLAs.

I think the rain in Newark was so unprecedented it broke all expectations of what was possible. It wasn't a one-in-anything flood.

I realize it was rain (and not a hurricane) but we had very similar flooding back in 2012 (just 9yrs ago) with Hurricane Sandy. https://en.wikipedia.org/wiki/Effects_of_Hurricane_Sandy_in_...

I'm not as familiar with NJ as I am with NYC -- but in NYC we had something similar in 2012 and again in ~1992 with a Nor'Easter.

I'm just wondering if the risk distributions were' using are unrealistic?

The flooding from Sandy was due to extreme high tide, not due to rain. Lots of places away from the sea (like Queens) were bad this year but fine in Sandy, while places close to the sea shore (like Staten Island) were devastated with Sandy but no problem this year. So multiple threats which is bad.

Reply on news.ycombinator.com