Feature
 |
Still image from Zareh Gorjian animation depicting an artist's concept for a dedicated Interferometric Synthetic Aperture Radar (InSAR) mission. |
"Large earthquakes are always disconcerting," she said. "Being a geophysicist I was immediately interested in how large the earthquake was and where it had occurred."
Within minutes, news reports confirmed that Los Angeles' San Fernando Valley had taken a direct hit from an earthquake comparable in size to the damaging 1972 San Fernando earthquake. More than 60 people were killed in each earthquake and thousands were injured. The latter event became one of the costliest natural disasters ever to strike the United States. Only the pre-dawn time of day and the fact that it was a holiday kept the death toll from being much higher.
Less than two months before that fateful day, Donnellan and colleagues from the Massachusetts Institute of Technology had published a landmark paper in the journal Nature on ground distortion north of LA's San Fernando Valley. Six years of relatively sparse data from a fledgling network of Global Positioning System (GPS) deformation monitors, that had been developed and installed around Southern California by scientists at JPL and other organizations, had detected that Earth's crust was being squeezed closed across the Ventura Basin. The data showed the area's faults were accumulating strain, and they gave the scientists clear indications of the style and relative size of an earthquake that might strike there, even though the faults there do not all break the surface. They placed no time frame on when such a temblor might occur, however.
California's San Andreas Fault
"The Northridge GPS measurements solidified in many scientists' minds how valuable data from space-based instruments could be for collecting precise measurements of Earth's crustal movements," said Donnellan. "We knew that something was up because an earthquake had not occurred there historically and yet a large amount of strain needing to be released had accumulated. After the earthquake, additional GPS data made it possible to rapidly and uniquely determine where the fault ruptured and to measure how the earthquake had deformed Earth's surface."
In the decade since Northridge, a high-tech, GPS-based ground deformation network was installed within Southern California. Called the Southern California Integrated GPS Network, it provides a continuous measurement of ground deformation at 250 locations with a precision of a few millimeters, measuring the slow buildup of deformation along faults. In addition, advances in satellite-based radar Interferometric Synthetic Aperture Radar (InSAR) and lidar are now used in combination with the GPS measurements to provide images of ground deformation for the entire Southern California earthquake region. These new technologies, coupled with powerful new computer modeling capabilities, have revitalized research in understanding earthquakes and earthquake processes. The new technologies will substantially refine earthquake hazard maps.
"We've confirmed through space observation that Earth's surface is constantly moving, periodically resulting in earthquakes, and that we can measure both the quiet motions preceding earthquakes and the quakes themselves," Donnellan said. "These technologies are allowing us to pursue lines of data and research we didn't know existed a few years ago."
Donnellan said scientists are particularly excited about the possibilities of InSAR, a technique that compares satellite radar images of Earth taken at different times to detect ground movement. "Last year, a study by NASA's Solid Earth Science Working Group concluded that the highest priority mission for the solid Earth science community is a dedicated InSAR satellite that can continuously monitor surface deformation," she said. Such a mission is a key component of EarthScope, a jointly led initiative by the National Science Foundation, NASA and the U.S. Geological Survey to study the North American continent's structure and evolution and the physical processes that control earthquakes and volcanic eruptions.
Dr. Brad Hager, a Massachusettes Institute of Technology professor and co-author of the 1993 Nature paper with Donnellan, said precise Earth surface movement data from space can measure strain in the Earth's crust and can provide a good first approximation of where earthquakes are likely. "In California, patterns of ground deformation are complicated by complex interactions between fault systems," he said. "Interpreting this data requires computer models that can estimate how much deformation has accumulated and to identify regions where strain should be released, but hasn't been."
University of California, Davis, Professor Dr. John Rundle said the complexity of earthquakes requires that we study them as part of the full Earth system. "Most natural events result from interrelated Earth processes with scales that are measured in lengths and times. "These processes have variables that can't be readily observed, so understanding them requires computers. A good analogy is the weather and climate simulations that meteorologists use to make our nightly weather forecasts and to forecast phenomena like the El Nino-Southern Oscillation."
Rundle said QuakeSim, a joint project between NASA and several other institutions, is developing such a forecasting methodology. QuakeSim's tools simulate earthquake processes and manage and model the increasing quantities of data available.
"We're focusing on observing and understanding earthquakes in space and time and on developing methods that use patterns of small earthquakes to forecast larger ones," Rundle said. "New simulations of earthquakes on California's active faults are providing considerable insight, showing that earthquakes tend to "cluster" in space and time due to their interactions: that is, an earthquake on one fault section can turn them on or off on nearby fault sections, depending on the relative orientation of the faults. So simulations have led researchers to conclude that fault system geometry determines earthquake activity patterns."
So just how close are we to being able to accurately predict when and where damaging earthquakes will strike? If the preliminary results of an ongoing NASA/Department of Energy-funded research project are any indication, perhaps not that far. Rundle's team set out to develop a 10-year forecast of damaging earthquakes in southern and central California for the years 2000 to 2010. The research uses mathematical methods to forecast likely locations of earthquakes above magnitude 5 by processing data on earthquakes of about magnitude 3 during the prior decade. The high-risk regions identified in the forecast map are refined from those already identified by the government as susceptible to large earthquakes. Since the research was completed, five earthquakes greater than magnitude 5 have occurred. All have been located within the high-risk regions identified in the forecast. Rundle said the odds of that occurring randomly are about 1 in 1,000.
As data from this project and other space-based technologies are validated, the technologies will be transferred to end users such as the United States Geological Survey. "Such data and models will improve our understanding of earthquake and volcanic processes, substantially refining seismic hazard maps and resulting in more appropriate, quake-resistant construction codes and more targeted retrofitting strategies," said Dr. Wayne Thatcher, a senior research geophysicist at the U.S. Geological Survey in Menlo Park, Calif. "Ultimately those improvements will help to mitigate damage from future earthquakes and save lives."
"We don't have all the answers," said Donnellan. "But we're making progress at a rapid pace, and I believe that it won't be that many years before we're able reduce the forecast probabilities from hundreds of miles and fifty years, to tens of miles and five to ten years. We can't stop earthquakes from happening, but with that kind of forecast accuracy, we'll be much better prepared for them when they do because we should be able to develop targeted and prioritized retrofitting strategies. It's an exciting time for our science."
NASA's Jet Propulsion Laboratory