New Zealanders living in the nation's second-largest city, Christchurch (population approximately 377,000) on the South Island's Canterbury Plains were hit hard Tuesday by magnitude 6.3 earthquake, an aftershock from September's magnitude 7.1 tremor. Prior to these two seismic events, Canterbury Plains likely had not experienced an earthquake in thousands of years. In fact, scientists did not even know there was a geologic fault there until it ruptured last year.

The latest Christchurch tremors were not as strong as the original earthquake, but they have caused considerably more damage and claimed dozens of lives. (No one died during the September quake). The Christchurch epicenter was only 10 kilometers outside of the city, whereas the 2010 event took place about 40 kilometers to the west, in an area that is mostly farmland. Adding to Christchurch's misfortune, the aftershock struck only about four kilometers in depth below the city, whereas September's temblor originated about 10 kilometers deep. Compounding these problems, Tuesday's quake hit during lunchtime when the city was buzzing with activity, whereas the earlier disaster occurred during early morning hours.

Scientific American spoke with Robert Yeats, a professor emeritus of geology at Oregon State University in Corvallis, about why earthquakes are so difficult to predict and what is being done to lessen the odds of surprise temblors.

[An edited transcript of the interview follows.]

The 7.1-magnitude earthquake in September caught the locals completely off guard. Why was that earthquake as well as Tuesday's aftershock such a surprise?
The earthquakes struck an area of New Zealand's South Island where sediments are deposited from the Southern Alps and from the nearby rivers. The sediment deposits reach all the way to the east coast on what is called the Canterbury Plains. The fault that ruptured in September had not done so in thousands of years, during which sediments had been deposited on top. I've been out there, and it's like driving anyplace where it's all flat. There are farms, but there's nothing that says, "Here's a fault." So, when the earthquake struck in September they were totally surprised. On the South Island, the Hope Fault and Marlborough Fault System are better known—there had been an earthquake in 1888 along the Hope Fault. Much less was known about faulting on the Canterbury Plains because no earthquake had happened in their historical record.

What defines an "aftershock" as opposed to an earthquake? Does a certain amount of time have to elapse for a seismic event to be considered an earthquake?
It takes many years before seismic activity can be considered an earthquake rather than an aftershock of a previous earthquake. That's a point of debate among seismologists. If you look at a map of southern California, you'll see quite a few little earthquakes south of Bakersfield in the San Joaquin Valley. Some people regard those as still aftershocks from the earthquake of 1952, which measured 7.3. That's not the consensus but it indicates that these aftershocks go on for decades. It takes quite a long time for everything to become quiet again. Now the San Andreas, on the other hand, had a large earthquake in 1857, just west of San Joaquin Valley, and it's quiet as could be. Same with the area of the 1906 San Francisco earthquake along the San Andreas. You can't paint all aftershock series with the same brush.

You mentioned that the aftershock was "shallow". What is the difference between an earthquake that takes place four kilometers below ground and one that takes place 10 kilometers below the surface, as the September earthquake did?
It's like how close you are to a bomb going off. If you're within a couple of kilometers, you're likely to get injured. If you're three or four times that distance, you not likely to. The waves are attenuating, or propagating, toward the surface. Christchurch was a very shallow earthquake, and that's a reason why the damage was much worse than the earlier one.

What do you look for when you investigate seismic activity in a particular area?
If I study a particular fault, I like to know its slip rate, how fast it's moving, whether it's a millimeter per year or a centimeter per year. As plates move, they're building up strain, and I estimate how much strain can build up before there is a rupture. New Zealanders have been good about trenching faults (digging trenches along fault lines to study previous seismic activity). The problem with this fault was that they didn't even know it was there. That tells me it's a pretty slow-moving fault but, nonetheless, when it builds up toward an earthquake of magnitude 7, then that's going to continue to produce aftershocks for a long time. It's not an exact science.

I'm working with the New Zealand Institute of Geological and Nuclear Sciences on a project funded by the Global Earthquake Model to map all of the active faults on Earth. This Global Faulted Earth project will include a global active fault and seismic source database, along with a book I'm writing. At some point in the not too distant future, if you hear about an earthquake in a place like Christchurch, you will be able to click on this database to find out what is known about it.

What can be learned from this week's aftershock in New Zealand?
We can map faults, and that's what we do, and we can use what we learn about those faults to establish some probability of an earthquake happening—but you can't map all faults. In the case of Christchurch, I'm not sure what they could have done differently. They could have said, let's do a seismic survey of the whole Christchurch metropolitan area just to be sure there's nothing going on underneath the city. But it was unlikely for the faults to extend as far south from the original Darfield earthquake site at as they did, so I can't fault them for not doing that.

You have to realize that New Zealand has some of the strongest building codes in the world, and those building codes are respected. That means you have loss of life, but it's in the dozens or maybe 100 or 200. If the same earthquake were to happen under a city of that size in a developing country, the number of deaths would be in the thousands, if not tens of thousands. Turkey, for example, had great building codes but that didn't keep tens of thousands of people from getting killed in the 1999 Izmit 7.6-magnitude earthquake because they weren't paying attention to those codes.