Share this story
Close X
Switch to Desktop Site

A new theory about frequency of fire, flood, quake

A geophysist applies 'chaos theory' to natural disasters - and surmises

Throughout time, man has been trying to make sense of the seeming whims of nature.

And as the world cleans up after a series of devastating earthquakes, hurricanes, and floods, one increasingly persistent question recurs: Are these events really as rare as we think they should be?

About these ads

Donald Turcotte, a geophysicist at Cornell University in Ithaca, N.Y., doesn't think so. He is getting fresh attention for what he believes is a more valid way of predicting the frequency of certain dramatic natural phenomena - such as forest fires, landslides, and floods. A headline in the Independent newspaper of London this month touted this new pattern as "the answer to life, the universe, and everything."

Well, maybe. Professor Turcotte himself begs to differ. But recently it has come out that this analysis of rare phenomena seems to apply to a much wider array of inquiries than previously thought, from species distribution in the forest, to magnetism, to aircraft turbulence.

If his thesis on floods is correct - and he acknowledges this is still under debate - it poses fresh questions about federal flood policy, home and business owners with properties in or near flood plains, and the strength of dams and levees. Because big floods are so expensive, virtually no private companies will insure against them. Since 1968, the federal government has been the main insurer for floods.

Mr. Turcotte's system uses the relatively new field of chaos theory - in particular, focusing on fractals, or "self-similar" objects - to show that in flooding, rare events actually occur more frequently than the labels would indicate. For example, in some instances, what might be called a 1,000-year flood should more appropriately be called a 100-year flood. (A 100-year flood means that there is roughly a 1 percent chance that that size flood will happen in a given year.)

Look in any flood-gauge station in the US, he says, and it will describe the strength of a so-called 100-year flood.

"We are seeing 100-year floods much too often, so either the statistical theory is wrong, or global warming has created a lot more floods," says Turcotte. "Those are two competing ideas. The insurance companies in some ways prefer global warming, because it doesn't mean they're doing their statistics wrong. We think the latter."

So far, says a senior official with the US Geological Survey, US flood policy is not about to be revised. But he agrees that the current system is less than perfect.

About these ads

"I've seen figures from North Carolina, Delaware, and New Jersey on the flooding from this past week, and we've got a lot of places where we're saying it's maybe a 500-year flood," says Bob Hirsh, chief hydrologist for the US Geological Survey. "Well, maybe it really isn't a 500-year flood. Maybe it's a more frequent event, but our theories didn't point us in that direction."

He says the current system for predicting the frequency of large floods, based on a curve called "log-Pearson type 3," was adopted many years ago because of a need for both accuracy and consistency. Over the years, the consistency has remained, but the accuracy has come into increasing question.

If the US government were to consider changing its system, a variety of competing theories would battle for supremacy.

Bob Hartwig, an economist for the Insurance Information Institute in New York, agrees that the growing frequency of certain rare weather phenomena has forced

'We are seeing insurers to recalculate their risk. After hurricane Andrew in August 1992, which caused $16 billion in insured losses, the industry learned it was much more exposed than previously thought - and has invested heavily in catastrophe modeling that predicts the track and intensity of hurricanes.

But he points out that many factors can affect the frequency of rare weather events, such as the earth's wobble and tilt, which vary over the years. He also notes that during the 1980s and '90s, the US has been in a period of heightened hurricane activity, which follows a few decades of relative calm.

One thing, though, is certain: Regardless of how frequent such natural phenomena actually are, they will appear to be happening with greater frequency. The earth's growing population means more and more people will live in harm's way. And with 24-hour-a-day cable news, big events will be evermore apparent.

(c) Copyright 1999. The Christian Science Publishing Society

Follow Stories Like This
Get the Monitor stories you care about delivered to your inbox.