A huge
earthquake in Chile in May 1960 created a tsunami that killed thousands of people
and sent waves across the Pacific Ocean, with effects felt as far away as the
Philippines. The earthquake was the largest recorded on modern-day seismometers
— and should not have happened when it did, according to seismologists’
theories on earthquake cycles. Earlier notable events, though smaller, should
have released stress being stored on the Chilean fault. But new research documenting
tsunami deposits in the region is resetting the seismic clock.
Researchers in Chile used soil profiles
to check historic earthquakes, reported by Spanish conquistadors and local settlers,
to see how they fit with the region’s devastating 1960 earthquake, the
largest ever recorded, at magnitude 9.5. Image courtesy of Marcos Cisternas.
The Nazca plate dives underneath the South American plate, creating a subduction
zone several thousand kilometers long, and a huge fault line where the plates
converge. Four major earthquakes within the past 250 years were greater than
magnitude 8 according to the historic record, and presumably released quite
a bit of stress from Chile’s subduction fault. Seismologists thought that
an 1837 earthquake in particular, which generated a tsunami that washed ashore
Hawaii with a 6-meter-high wave, was large enough to have relieved some of the
tension on the fault.
But the 1960 earthquake measured magnitude 9.5, rupturing 1,000 kilometers of
the fault’s length — an event that came as a complete shock. The conundrum
puzzled seismologists: 123 years clearly was not enough time to build that much
stress on the Chilean fault.
To address that problem, a team of researchers led by Marcos Cisternas of the
Pontificia Universidad Católica de Valparaíso, Chile, searched
for evidence of how big the previous earthquakes were. They documented signs
of about a dozen earthquakes at sites along Chile’s south-central coast,
particularly noting tsunami deposits, and checked written reports of shoreline
changes or subsidence made by Spanish conquistadors and later settlers. In some
cases, such as a huge 1575 earthquake, the team found evidence for local tsunami
events. But for the substantial 1737 and 1837 earthquakes, no tsunami evidence
was found in local Chilean deposits, according to the research published in
Nature on Sept. 15.
This absence of evidence has implications for the nature of magnitude-8 earthquakes
versus magnitude-9 and higher, and for the earthquake cycle of subduction zones
such as Chile’s — similar to the Sumatran fault that set off the Indian
Ocean tsunami last December, and with some parallels to the Cascadian subduction
zone in the Pacific Northwest. Stress building on that part of the Chilean subduction
zone needs about four centuries “for incubating a giant” earthquake,
Cisternas says, leaving that part of Chile relatively safe for now.
Seismologists “tend to assume that stress alleviated is related to size”
of an earthquake, says Alan Nelson of the U.S. Geological Survey in Lakewood,
Colo. The new study, which Nelson calls “very detailed” with “excellent”
data, shows “it’s more complicated.”
In Sumatra, for example, “people were very surprised” by the magnitude-9-plus
earthquake that struck last December, he says. A magnitude 8 would have been
unsurprising, judging from the last big earthquake there, about 120 years ago
— coincidentally the same as Chile’s seismic history. But as more
extraordinarily large events occur and are documented, expectations have been
slowly changing. For example, as scientists further elucidate the Pacific Northwest’s
seismic and tsunami history, Nelson says, they continue to reevaluate the seismic
hazard for the region.
Another conclusion drawn by Cisternas’ team is that magnitude-9 events
on subduction faults occur on separate but parallel cycles from magnitude-8
earthquakes. Cisternas suggests that this evidence could change geophysical
models of subduction fault zones. But Hiroo Kanamori, a seismologist at Caltech
in Pasadena, Calif., says that the seismological community has long known that
earthquakes do not unfold like clockwork. “Any sequence can’t be very
regular because dynamic relations between different patches” create chaotic
event patterns, Kanamori says. The main value of the new research, he says,
is that it provides solid field evidence finally confirming suspicions that
the 1837 earthquake did not have as much of an impact on the fault as the historic
record indicated.
Naomi Lubick
 |
Geotimes Home | AGI Home | Information Services | Geoscience Education | Public Policy | Programs | Publications | Careers  |