Stanford University Home

Stanford News Archive

Stanford Report, March 6, 2002

Silent earthquake in Hawaii offers clues to early detection of tsunamis


A slow-moving earthquake recently observed on Hawaii's Kilauea Volcano could become a model for predicting catastrophic tsunamis in the Pacific, according to a new study by geophysicists from Stanford and the U.S. Geological Survey (USGS).

Writing in the journal Nature, the researchers explained how they were able to detect a "silent" or "aseismic" earthquake on Kilauea's southern flank in November 2000 using data from the Stanford/USGS global positioning system (GPS) network on the Big Island of Hawaii. The magnitude 5.7 quake was a relatively slow-moving event that lasted about 36 hours and caused the southern flank of the volcano to slide nearly 3.5 inches (8.7 centimeters) into the sea.

Peter Cervelli of the U.S. Geological Survey installs a global positioning system receiver on the south flank of Hawaii's Kilauea volcano -- one of 17 permanent GPS instruments operated by Stanford and the USGS to monitor volcanic activity on the Big Island. A steam plume caused by lava entering the ocean is visible on the horizon (far left). Photo courtesy of Peter Cervelli

"We call them 'silent' earthquakes because the ground doesn't shake, and they produce no seismic waves," said Paul Segall, a Stanford professor of geophysics and co-author of the Feb. 28 Nature study.

Paul Segall

"We don't know what triggers them, but we do know that the Kilauea event was the first silent earthquake ever observed in a volcanic environment," he added.

"Fortunately, no one felt the earthquake on Kilauea," noted Peter Cervelli, who joined the USGS Hawaiian Volcano Observatory last fall shortly after receiving a doctorate in geophysics from Stanford.

"In fact, without instruments like our GPS observers, we never would have known that the Kilauea event occurred," added Cervelli, lead author of the Nature study.

In a companion feature in the Feb. 28 issue of Nature, research geophysicist Steven N. Ward of the University of California-Santa Cruz estimated Kilauea's southern flank to be nearly equal in size to a half-mile-thick slice of Rhode Island. Had that massive chunk of land suddenly collapsed into the ocean instead of sliding just a few inches, it could have generated an enormous wall of seawater -- or tsunami -- powerful enough to threaten coastal cities as far away as California, Chile and Australia, according to Ward.

A catastrophic flank collapse of an oceanic volcano happens somewhere in the world every 10,000 years on average, Ward added, but none has been caught in its early stages until now.

Controversial rainstorm

In their study, Segall and his colleagues suggested that the Kilauea quake might have been triggered by a torrential storm that dumped nearly 3 feet (1 meter) of rain on the Big Island on Nov. 1, 2000.

In the 1960s, geologists showed that injecting fluid deep into the ground can cause a fault to fail -- thus triggering a small earthquake.

Computer simulation of the tsunami waves that might set off in a collapse of Kilauea's southeast flank. The simulation assumes that a block 40 km long, 20 km wide and 2,000 m think (blue, inset) slides for 60 km at 70 m/s. The tsunami wave fronts are pictured at two-hour intervals from two to 14 hours. Red and blue contours are wave elevations and depressions, respectively, and the numbers are sample wave heights in meters. Illustration courtesy of Nature

According to the Nature study, the intense, 24-hour rainfall that inundated the Big Island on Nov. 1 may have had a similar effect on the Hilina/Holei Pali fault system located some 2.8 miles (4.5 kilometers) beneath Kilauea Volcano. GPS data revealed that the fault suddenly began to rupture on Nov. 8 -- one week after the storm.

"When water penetrates the relatively dry upper regions of the Hawaii crust, it eventually gets to the water table and causes it to rise," observed Cervelli. "This raises the pressure of the water throughout the volcanic edifice."

He noted that rainwater might have produced a "pressure pulse" that percolated through the crust for seven or eight days until it reached the fault zone.

"What we're hypothesizing could have occurred -- although we're certainly not wedded to this conclusion -- is that, when the pressure pulse began in early November, it propagated slowly down the fault zone and raised the pressure there," Cervelli said. "This had the effect of opening the fault a little bit and bringing it closer to failure."

The mechanism is very similar to what happens when a hot glass is placed upside down on a wet counter top, Segall explained.

"The heat from the glass heats up the air inside the glass," he noted. "This raises the pressure and can cause the glass to slide around on the counter."

Cervelli conceded that linking the Kilauea earthquake to the heavy rainfall event is controversial.

"It's not inconceivable that the rainfall event had a triggering role in the slip event," he noted, "but before we go out on a limb and say definitively that it is our opinion that rainfall of this magnitude can and has triggered silent earthquakes, we would like to first flesh out our model a bit more and, second, to observe multiple events so that we can correlate them with rainfall."

Tsunami warning

"We don't know how common silent earthquakes are because, up until now, we haven't had the capability or tools to measure them," Segall explained.

People in Hilo, Hawaii, flee a powerful tsunami that struck the Big Island on April 1, 1946. A 7.3 earthquake in Alaska generated the tsunami, which killed 159 people and caused millions of dollars in damage. Major tsunamis also struck the Hawaiian Islands in 1960 and 1975. Photo courtesy of Cecilio Licos, Yasuki Arakaki collection, Pacific Tsunami Museum

He pointed out that detecting the silent quake on Kilauea would have been impossible a few years ago, before Stanford and the USGS established a permanent network of instruments capable of monitoring millimeter-sized movements on the volcanic surface on a daily basis.

"Now that we have the networks in place, we're finding that silent earthquakes are popping up in all kinds of surprising places -- like volcanoes -- that we didn't know about before," Segall added. "This event did not produce a tsunami, but if we can detect potentially catastrophic ground motion in its early stages, we might be able to issue tsunami warnings in the future."

Ward agreed, noting that the silent earthquake detected by Segall and his colleagues could be interpreted as the early stage of a catastrophic flank collapse that may occur one day on Kilauea.

"People should not lose sleep over large but rare natural hazards," Ward wrote. "They should not run blind either, particularly when a useful eye exists. The world's oceanic volcanoes are stages best not left unwatched. For now, GPS provides one of the sharpest views."

Other co-authors of the Nature study are Stanford graduate student Kaj Johnson; Michael Lisowski of the USGS Cascades Volcano Observatory in Vancouver, Wash.; and Asta Miklius of the USGS Hawaiian Volcano Observatory in Hawaii National Park. The study was funded with grants from the USGS and the National Science Foundation.