From a reader:
This is nonsense.
A 500-year flood is not a flood that is supposed to happen only once every five hundred years. It’s a flood that has a probability of 0.002. You could have a 500-year flood several years in a row, just like you can flip heads three or four times in a row. Each occurrence is an independent event. After you get three heads in a row, the fourth toss still has a 50-50 chance of being heads. Floods from year to year are not completely independent (given global warming, dams, levees, irrigation, deforestation, etc.) but a single 500-year event is not in itself proof of anything. Even the most idiotic trend requires at least two data points.
Me: This sounds like a good point, and I’m hardly versed on how engineer types (like this reader) measure these things. But I should say I’m entirely open to the idea that these floods are to some extent man-made (or man-exacerbated, if that phrase passes muster), I’m just dubious about the global warming part. See here, for one of many such explanations.
Update: From another engineer reader (there are wayyyy too many of you):
Your reader response is quite correct. Even check our “wonderful” government bureaucracy of FEMA and search for the definition of a 500 year storm event or flood. Also, FEMA has changed (and rightfully so) the name of the storm event or flood to not use a term of years. A 500-year is now a 0.2% chance event. A 100-year is now a 1% chance event. This was done to prevent bad claims that a 500-year event is happening more often do to one reason or another.