Earth's surface features predict earthquakes

Earthquakes sometimes seem like the rumblings of invisible giants — predicting the areas which are likely suddenly to shift requires an intimate understanding of what goes on deep inside the crust. New Scientist’s Catherine Brahic writes that now a team of researchers says seismologists could make better use of the surface features of mountains to detect the troubles which lurk beneath. Such analysis of the landscape could highlight regions that conceal powerful and active faults below. In the 1990s, Eric Kirby of Penn State University and colleagues trekked out to the mountains above Chengdu in western China and made detailed observations of the steepness of riverbeds. Their observations led them to suggest that of all the numerous faults that run along the length of the mountains, the southern extent of the Yingziu-Beichuan fault was most likely to suddenly shift. With hindsight, their study seemed inspired — this is the very fault which shifted by five metres on 12 May 2008, flattening entire villages and killing 70,000 in the Sichuan earthquake. How did the researchers arrive at their conclusion? “You can identify patterns of active deformation in a landscape,” says Kirby. “Mountains are formed by tectonic plates pushing against each other and crumpling the crust, but mountain ranges can be vast and identifying the zones that are most likely to rupture is a challenge, especially in remote and rugged ranges.”

Thanks to new radar technologies, high-resolution maps of the surface can now be obtained from planes and satellites. These maps can help distinguish sudden changes in the slope of riverbeds, which can indicate where the watercourse crosses an active fault line. “If you were sitting in Chengdu, and could lift all the clouds and smog, you would see an incredibly steep mountain front rising almost 6000 metres above you just 50 kilometres to the west,” says Kirby. This is where the tectonic plate beneath the Chengdu basin pushes up against the Eastern Tibetan plateau. Rivers flow out of the range into the Chengdu basin, perpendicular to a series of fault lines (see map, right). Kirby and his colleagues noticed that the riverbeds were in parts far steeper than elsewhere, as though forming steps down the valley. “If you kayaked down one of these rivers, you would hit a point where it falls down steeply, along a series of rapids,” says Alex Densmore of the University of Durham in the UK, who has also studied the region’s faults. To Kirby and his colleagues, this sudden change in gradient indicates a region that is rising faster than the surrounding landscape - a fault that could slip. So having pinpointed this steep rise, they proposed in 2003 (Journal of Geophysical Research, DOI: 10.1029/2001JB000861), that the southernmost part of the Yingziu-Beichuan fault was more active than surrounding faults and was most likely to slip. Although their suspicions were verified only five years later by the 2008 Sichuan earthquake, the observations of Kirby’s team could not have predicted the precise timing of the event. The researchers also say that the connection between mountain shapes and earthquake risks are not always easy to interpret. “It is not as simple as saying ‘Aha! the mountains are steeper here so they must be rising faster, and this is where we’re going to see an earthquake’,” says Kirby. “Rocks, for instance, are not always of the same strength, they give in more or less to erosion, which is the other factor that affects the slope of a riverbed.” He adds: “You could say it is a quick and dirty reconnaissance tool. Topography can highlight regions that have the potential for active deformation, but then you have to get on the ground and dig around, and look at history of earthquakes in region.”.

To Densmore, this is as far as the method can go. “But that in itself is a huge step compared to nothing,” he says. “Seismologists have limited time and funding to do field studies, so if you can direct the selection of sites and go to areas that are most likely of being at risk, then you have done something worthwhile.” Kirby fantasizes that the method could be pushed further. He is now analysing his readings of topography maps in the Andes and Himalayas against erosion and slip data. “In an ideal world, we could one day come close to pointing to a steep slope, in granite, that receives so much precipitation and say something about how much the fault is rising,” he says.

-read more in Eric Kirby, Kelin Whipple, and Nathan Harkins, “Topography Reveals Seismic Hazard,” Nature Geoscience (20 July 2008)