See more stories
Submit your own story
So far, science hasn’t found a way to predict when earthquakes will happen. But Jordan’s team is working on the next best thing … being prepared for them.
Specifically, Dr. Jordan and his team at the Southern California Earthquake Center (SCEC) are developing methods to simulate how the ground will move in the event of an earthquake.
Their research is producing more accurate
seismic hazard information that communities can
use to prepare for earthquakes such as retrofitting
existing structures and supporting better building
codes for new construction projects.
National Center for Supercomputing Applications (NCSA)/Blue Waters
The Blue Waters supercomputer, housed at the National Center for Supercomputing Applications (NCSA) at the University of Illinois at Urbana-Champaign, is the fastest university system in the world. It’s used for a wide range of research problems, from predicting the behavior of complex biological systems to simulating the evolution of the cosmos.
Cray® XE™/XK™ hybrid supercomputer
13 PF peak performance
22,640 XE compute nodes
4,228 XK GPU accelerators
Oak Ridge Leadership Computing Facility
The Oak Ridge Leadership Computing
Facility (OLCF) is home to Titan, the
U.S.’s most powerful supercomputer for
Cray® XK™ hybrid supercomputer
27 PF peak performance
18,688 compute nodes
18,688 GPU accelerators
“We can’t predict earthquakes,” says Jordan.
“But what we’re doing is a much better job of
predicting what will happen when they occur.”
Key to making these predictions starts with
knowing how the ground will move in specific
locations. Thus, the linchpin of SCEC's
research effort is CyberShake -- a new
physics-based computational approach to
calculating how, and where, earthquake waves will
produce ground motion as they propagate through the
earth's complex crust.
Ground motion is notoriously difficult to model,
Jordan says, and getting an accurate view of
ground motion takes more than the standard
CyberShake takes those standard techniques
and builds on them, integrating advanced
physics, boosting the model frequency range,
and aggregating a vast amount of data and
wave motions to produce the most complete
earthquake models yet.
“The holy grail of hazard maps is to minimize
uncertainties [about the ground motions],” says
Jordan. “CyberShake lets us really understand
those ground motions in a systematic way.
And when we can compute ground motions
more accurately, we can make infrastructure
more resistant and society more resilient to
The CyberShake project’s success to date has
been supported by its access to two leading Cray
supercomputers — Blue Waters at the National
Center for Supercomputing Applications (NCSA)
and Titan at Oak Ridge Leadership Computing
“These calculations are so large in scale,” says
Jordan’s teammate and SCEC science director
Christine Goulet. “But they’re blazingly fast on
supercomputers.” Their latest run generated two
sets of seismic maps expanding from the original
Los Angeles basin area into central California and
covered 438 computation sites including public
utility stations, historic sites and key cities.
These calculations, says Jordan, are a “pretty
awesome thing” and representative of a giant leap
forward in earthquake science. Does it mean that
as computing advances we might one day be able
to actually predict an earthquake? Jordan pauses
and then answers: “We’re not holding our breath,
but I wouldn’t rule it out.”