For nearly 2 centuries, Dharahara, a nine-story observation tower in the center of Kathmandu, stood firm as the tallest building in Nepal’s capital. But on 25 April, it was reduced to rubble by the magnitude 7.8 Gorkha earthquake—even though many one- and two-story buildings escaped the shaking relatively unscathed.
Two factors caused the disparity, a pair of new studies shows. First, a basin underneath Kathmandu, filled with half a kilometer of soft sediments, amplified the shaking of taller buildings; and second, a gentle, slow onset to the earthquake’s rupture limited the types of waves that would have shaken shorter buildings. The peculiar nature of the earthquake was one reason it killed fewer than 9000 people, and it could have been far worse, says Jean-Philippe Avouac, a seismologist at the California Institute of Technology in Pasadena and leader of the studies, published today in Science and in Nature Geoscience.
The researchers analyzed data from 15 GPS stations, painstakingly deployed in the mountainous region, along with data from a single seismometer near Kathmandu. With the data, they constructed a dynamic picture of the earthquake, which was a rupture of a 20-by-140-kilometer patch along the Main Himalayan Thrust (MHT)—a 2200-kilometer-long fault zone that runs near the southern border of Tibet. “It’s really the first time we have such nice measurements,” Avouac says. “We’ve never been so close to the rupture itself.” Along this fault, the Indian subcontinent is being forced down underneath Asia, and the collision is the force pushing up the tallest mountain range in the world.
Earthquake shaking is caused by a mixture of seismic waves moving through the ground at different frequencies. High-frequency waves tend to cause the most problems for short buildings, whereas low-frequency waves are more damaging for taller buildings. Kathmandu’s basin of soft sediments helped amplify low-frequency waves, as seen in its destructive effect on taller buildings, says Youssef Hashash, an earthquake engineer at the University of Illinois, Urbana-Champaign, and the lead author of a report published last week that documented the quake’s damage to infrastructure. But the amplification effect, centered on low-frequency, long-period waves of 5 seconds, is somewhat puzzling and might not be explained entirely by the shape of Kathmandu’s basin and the characteristics of the soft sediments it contains, Hashash says. “We really haven’t seen something that amplifies at 5 seconds before,” he says. “It may have something to do with the nature of the slip itself.”
Hashash also notes that the earthquake’s “directivity” played a big role in shaking Kathmandu more than other parts of Nepal. The earthquake began 80 kilometers to the northwest of the city and ruptured eastward. Just as an ambulance’s siren screeches more shrilly as it approaches a listener, the earthquake’s waves piled up on one another as they propagated toward Kathmandu at 3.3 kilometers per second.
Video credit: Diego Melgar
But the earthquake seemed to lack the high-frequency waves that would have damaged smaller buildings—and it probably has something to do with its smooth and slow onset, Avouac says. “This earthquake was surprisingly gentle,” he says. The MHT fault is nearly horizontal, but to the north, the fault begins to dip downward into the earth. Slip along the fault there, under the higher temperatures and pressures found deeper in the crust, takes on a different character: Ruptures are more plastic than brittle. Avouac says the slip in the Gorkha earthquake occurred closer to this plastic zone than in other Nepalese earthquakes, and the difference may have contributed to its gentler qualities.
Chris Marone, an earthquake physicist at Pennsylvania State University, University Park, is skeptical that the rupture actually behaved this way. In the Science paper, the researchers derived a quantity related to the fault’s slow onset, called the slip weakening distance, and found it to be several orders of magnitude higher that laboratory experiments on rock samples would predict. “It’s a challenge to people who do numerical models of earthquakes, anybody who does theory of dynamic rupture,” Marone says. “I don’t buy it.” He says the GPS stations used in the study naturally favor the collection of low-frequency data, and thinks that if the team had had more seismometers in place, they would have recorded more high-frequency waves that would have altered the researchers’ interpretation of the rupture.
But Roland Burgmann, a geophysicist at the University of California, Berkeley, says the apparent slowness makes him want to listen even more carefully. “Could there have been an even slower precursor slip that we failed to detect? Nepal is not that well monitored,” Burgmann says. “This gives us more reason to look for that.”
One thing all the researchers agree on is that this earthquake was not ”The Big One.” The Gorkha earthquake relieved strain over a fairly small patch of the MHT fault. To the west, a huge swath of the fault remains locked and has been accumulating strain since 1505, when a much more massive magnitude 8.5 quake went off. “Now we are a bit worried about all the other areas that are locked in Nepal,” Avouac says.