“500-year flood”
Hint: it has more to do with journalism than with rivers
I often see news stories that say, “Scientists estimate that last week’s deluge was a 500-year flood”. But that doesn’t make much sense; there was a 500-year flood in 1993 and we’re having another one now. What gives?
Chris Rowan at Highly Allochthonous says big surprise, the scientists never said “500-year flood’, or at least they’re learning not to after the media grabs their conclusions and twists them into sensationalistic headlines.
Here’s the short version: events that are both extreme and semi-random are impossible to predict in a given year, so we express them as a probability. As in, “the probability of a river flow matching this rate is .002 in any given year.” News creatures hear that statement, think back to junior high math and think; “That’s one in 500! That means a flood like this one occurs every 500 years!”
No it doesn’t, any more than it means that every other flip of a coin will be ‘heads’. And - as data becomes available, extrapolations give way to observations and the probability estimates are revised. The long version, and the comments following, are well worth reading as an antidote to the chattering journalistic class.
Jon Stewart’s take on the flood
Posted by Lucas on 06/21/08 at 12:50 AM
Next entry: Suicide and religion
Previous entry: Larry Gonick's new comic(s)

