Why the concept of '100-year events' in nature is wrong

Homes are reflected in flood waters in Minot, N.D. on Saturday.

As the Souris River in North Dakota moves as much as 5 feet above an 1881 flood record, and as the Missouri River Valley worries about flooding in weeks ahead, and faces its own fair share now, we continue to hear about "100-year-floods" and other 100-year events.

This is just wrong.

First, we don't know how much anthropogenic global warming is skewing the issue. But, here's a bit of insight by Penn State's Michael Mann:

"Even a couple degree warming can make a 100-year event a three-year event," Mann, the head of the university's earth systems science center, told AFP.

"It has to do with the tail of the bell curve. When you move the bell curve, that area changes dramatically."

I'm also wondering, per the Wall Street meltdown and the failures of the "quants" to know it could happen ... is that "tail" getting fatter until we hit a new equilibrium?

Second, beyond that, in much of the U.S., we've only been taking detailed weather measurements for about 100 years. To measure just 100 years of anything, and assume that represents an average out of a longer period in history, is simply wrong. It's intellectually stupid.

It's like looking at world history since 1914 and making assessments about a 100-year average of world violence.

It's unbelievable that people do this anywhere, just for the second reason. For both reasons, it's doubly unbelievable that professional climatologists and meteorologists still talk like this.

It makes the public think, even short of AGW, that we can predict severity of climate issues with more accuracy than we can. It's probably been a factor in continued building in floodzones that may not be 100-year floodzones, but 50-year floodzones for all we know.

Ditto on the flip side in overbuilding in the vicinity of Western river basins, as anybody who knows the history of the Colorado River Compact and the acceptance of a series of heavy flow years as "average."