NPR's Bryant Park Project features a story about last week's M 5.4 earthquake in Illinois. And, well, they should have interviewed Maria instead, because her post explaining earthquakes in Illinois is really good. And NPR's story, ummm, isn't.
What did NPR get right? Well, yes, there have been big earthquakes on the New Madrid seismic zone. Really big earthquakes. And the 1811-1812 quakes (which supposedly rang church bells in Boston) are worth telling people about.
The California activity — along what is known as the San Andreas fault — is a result of the North American plate riding over the Pacific plates and continually building stress, Gibson says. The New Madrid fault, on the other hand, is an "ancient scar" where two very old plates are slowly pulling apart.
Ok. Problem #1: California. California is on a strike-slip plate boundary (as Julian's photos show very nicely). The Pacific plate slides past the North American plate. Mostly. Except where the boundary isn't oriented in quite the right direction. But it's not a subduction zone, as implied by "riding over." (You need to go north, near the Oregon border, to find one of those. And up there, North America slides over the teeny-tiny Juan de Fuca plate, not the Pacific plate.)
Problem #2: New Madrid as an active rift? The failed rift that is now the New Madrid seismic zone tried to tear North America apart 700 million years ago. It's not an active rift now. It failed. No ocean basin. No separate continents. Just lingering effects like low elevations and lots of old faults, just waiting to be reactivated. Maria's analogy is perfect: they are old injuries that react to unrelated events. In this case, I would guess that the fault was driven by some east-west compression of the North American plate, between the Mid-Atlantic ridge on the east and the San Andreas fault on the west. (The focal mechanism says that the earthquake involved strike-slip motion with east-west compression - consistent with reactivation, but not with active rifting.)
So. Seismic hazard? Yes. (And so are many other areas with old normal faults: one of the other big earthquakes of the 1800's was in Charleston, South Carolina.) Future beach-front property? Not without a lot of sea-level rise.
And should we care? Well, if the New Madrid seismic zone were actively building stress, we might have a better handle on the risks there. We could use continuous GPS to monitor the ground deformation. We could make a decent guess about the likely size of future ruptures, and the size of future earthquakes. And then it might be possible to decide just how crucial it would be to retrofit old buildings. But as it is, we've mostly got history, rather than theory, as a guide. And in most places, human history is too short to give a good feel for earthquake risk.