Earthquake sensors could help predict rockfalls

Valerie Zimmer

Earthquake sensors could help predict rockfalls

According to one eyewitness, watching Yosemite Valley’s largest rockfall was like seeing a cliff “unfolding like the stairs of an escalator.” In total, 600,000 cubic meters of rock fell from the valley’s north side that day in 1987, burying part of the road beneath 3.5 meters of rock debris. Fortunately, the area’s roads and trails had been closed earlier, the park rangers forewarned by seeing a series of smaller rockfalls—and hearing loud cracking noises—coming from the cliff face. The possibility of picking up on similar early warnings has made the monitoring of rockfall patterns an important activity, but useful data is lost each time a fall goes unobserved. Now, researchers say they have hit upon a better way to track falls in the area—using the same tools that monitor earthquakes to listen for the seismic disturbances rocks make as they impact the ground. The team set up seismic stations—with geophones and accelerometers to detect ground movement and acceleration, along with infrasound sensors—within 1 kilometer of an active rockfall zone in Yosemite Valley, where the 1000-meter-high cliffs experience an average of 40 to 70 falls per year (see example, pictured). The study ran over two winters, when rockfalls in Yosemite are the most frequent. The scientists developed an algorithm to distinguish rockfalls from other sources of seismic noise, like rainstorms, avalanches, and earthquakes. Writing in Engineering Geology, the researchers report that the sensors were able to pinpoint small rockfalls (of less than 20 cubic meters) at distances of up to several hundred meters away—and identify not only individual impacts within falls, but also, in some cases, the signals generated as the rocks began to move. In total, the team detected 12 rockfalls, with six being related and forming part of a larger, ongoing instability in the cliff face. Four of the falls had not been observed independently and would otherwise have gone unnoticed—meaning that useful information on the pattern of falls would have otherwise been lost.