What triggers earthquakes? Are they predictable? What makes volcanoes erupt? These and related questions are what drive EarthScope, a set of experiments that will use the US as a laboratory to explore the structure, evolution, and dynamics of continents.
“We are looking at the deformation and structure of the North American continent at all scales,” says Gregory van der Vink, director of EarthScope, which is headquartered in Washington, DC. One goal is to figure out how the continent was assembled over geologic history, he adds. Seismometers across the US will image Earth’s interior; geodetic sensors will monitor deformation of the crust around the boundary of the North American and Pacific tectonic plates; and a deep hole drilled into California’s San Andreas fault will provide an up-close look at fault zone composition and behavior before, during, and after earthquakes.
Although most of the instruments in EarthScope are proven workhorses, using them on such a massive scale is new and is feasible thanks largely to advances in moving and crunching big data sets. The integration of seismic, geodetic, and in situ earthquake data is also new—as is drilling into an active fault, says van der Vink. Collaborations that would extend some EarthScope measurements into Canada and Mexico have been initiated, he adds.
“EarthScope will revolutionize Earth sciences. It is like constructing a telescope that looks into the Earth. It will tell us how Earth’s engine works,” says Jim Whitcomb, a section head in Earth sciences at NSF, which, under its major research equipment and facilities construction (MREFC) account, is funding the project with $219 million over five years. NSF will also help fund operations and data analysis through grants to investigators. The first data are expected within a year, and the project will run for at least 15 years. The data will be freely and immediately available via the Internet. Among EarthScope’s practical applications, Whitcomb says, are determining how and where mineral deposits are formed and understanding and mitigating the hazards of earthquakes and volcanoes.
Velocity tomogram
Vibrations from earthquakes around the world will be recorded by USArray one component of EarthScope. Some 400 seismometers, sensitive to ground motion in the frequency band 0.01–15 Hz, will be arranged in a grid with 70-kilometer spacing. The grid will start off as a swath of roughly 8 by 50 stations along the West Coast, and gradually move eastward, visiting some 2000 spots in the lower 48 states and Alaska over 10 years. Individual seismometers will spend an average of 18 months at any given location.
By factoring in when and where specific earthquakes occurred, the data can be used to create a map of seismic velocity, akin to a medical tomogram. Each seismometer gives an average velocity of a seismic wave, says Göran Ekström, a geophysicist at Harvard University and chair of the executive committee of Incorporated Research Institutions for Seismology, the university consortium that runs USArray. “But with an array, you can tease out the three-dimensional distribution. For example, if a signal took a long time to go through Earth’s crust, the array data can tell you if the crust is thick or if the wave was slow.” Time-series recordings show the topography of internal boundaries, Ekström adds.
Composition and temperature can also be teased from the seismic velocity map. For example, ultrasonic wave propagation in rocks under pressure or as a function of temperature will help interpret the velocity data. And fluids can be pinpointed in combination with conductivity data garnered from electric and magnetic field measurements with magnetotelluric sensors. Says Stephen Park, a geophysicist at the University of California, Riverside, “Graphite films are conductive, but they do nothing to the velocity. For partial melt [of rocks], on the other hand, the velocity will be lowered and the conductivity will be increased.”
USArray will have another 2400 seismometers for dispatching to specific targets. They might be set up where, for example, a massive earthquake rattled Missouri nearly two centuries ago, a volcano erupts, or controlled explosions are set off to probe local interior structure. The USArray suite of instruments will further include fixed-location seismometers to both calibrate the grid as it makes its way across the US and augment global earthquake monitoring.
Deformations over time
Contrasting the abruptness of earthquakes is the slow drift of tectonic plates. “Along the margins [of the plates] are deformations going on on all time scales,” says William Prescott, president of UNAVCO, the university consortium in charge of the Plate Boundary Observatory. The PBO is EarthScope’s project to study surface deformation across the boundary of the North American and Pacific plates. With an array of global positioning system receivers and strainmeters, the PBO aims to tackle questions about earthquakes, volcanic activity, and the evolution of plate-boundary zones.
Nearly 1000 GPS receivers and 180 strainmeters will pepper the country from the Rocky Mountains westward as part of the PBO. “We put the GPS stations out and leave them in the same place and watch them move day by day or hour by hour,” says Prescott. “A typical instrument near a major fault would move at about 3 centimeters a year, and you would see an offset in meters for a big earthquake, and then an exponential decay as it gradually relaxes back to the normal rate of movement.” Some GPS receivers will target specific transient phenomena, such as a slow earthquake in the Pacific Northwest that recurs every 14 months and lasts about 10 days, Prescott adds. A hundred of the receivers will be portable, at the ready to study crust movement near volcanic eruptions and other tectonic crises.
The strainmeters will be clustered along active faults and in volcanic areas. About 3 meters long and 10 centimeters in diameter and cemented into rock about 200 meters below the surface, they measure strain in three dimensions. When a rock squeezes or relaxes, the strainmeters sense changes in their diameter to a part in a billion. “They are sensitive to signals we can’t see with GPS,” says Prescott.
From the rate and degree of movement, scientists can map how surface deformations change with time. Constraints on the viscosity can be solved for and then related to the strength of the crust and mantle and to the stresses that produce the deformations to begin with, says Paul Silver, a scientist in the terrestrial magnetism department at the Carnegie Institution of Washington and chair of UNAVCO’s board. The viscoelastic response also determines the redistribution of stress, which makes future earthquakes “more likely to occur where the change in stress has brought a region closer to failure,” he says. “We want to understand the connections between aseismic deformation and earthquake occurrence. We want to figure out what triggers earthquakes in the first place and what correlations exist between one earthquake and the next.”
A biopsy
“So far, we have been forced to look through miles of rock which obscure and distort signals. To really understand earthquakes, we need to see up close the processes that cannot be observed from Earth’s surface,” says Bill Ellsworth, chief scientist of the earthquake hazards team at the US Geological Survey (USGS). Such an up-close view is the aim of a third EarthScope project, the San Andreas Fault Observatory at Depth (SAFOD), a 4-kilometer-deep hole that will host instruments that measure fluid pressure, local deformation, temperature, and seismic waves, and will provide core samples from the fault.
The borehole will be halfway between San Francisco and Los Angeles, near Parkfield, where magnitude-6 earthquakes have repeatedly occurred, most recently in 1966. The site was chosen because it’s the best-studied part of the San Andreas fault and because it exhibits both earthquakes—a magnitude 2.2 recurs about every two years—and creep, slower movements that don’t radiate seismic energy. About 2 kilometers down, the bore hole will bend at a 50° angle and cross the fault zone. It’s like taking a biopsy, says Ellsworth. “We will get to sample the entire fault zone.”
SAFOD scientists, like those working with the PBO, hope their experiment will shed light on what triggers earthquakes. Do they start slowly and accelerate, or do they start fast? Does creep trigger earthquakes? What controls how they propagate and stop? “Whether earthquake prediction is possible is very controversial and points to our lack of understanding of the initiation process,” says USGS geophysicist Stephen Hickman.
A related question is why plate boundary faults are weaker—slipping with less shear stress—than faults in plate interiors. “Is it because the fluid pressure is elevated? The role of fluids is a very big question,” says Mark Zoback, a geophysicist at Stanford University, which, with the USGS, runs SAFOD. “We have no idea whether fluids have anything to do with the mechanics of faulting. We don’t know where the fluids come from. Are they meteoric? Are they freed from minerals in dehydration reactions?” Other possible explanations for plate boundaries having slippery faults involve localized weak minerals, chemical reactions, and the dynamics of rupturing.
“We will be able to exhume fault zone materials and measure composition, strength, deformation mechanisms. We will hold these materials in our hands for the first time,” says Zoback. Lab tests on the sample cores, combined with the long-term observations in the borehole, he says, “will help us answer specific questions about the physics of faulting. Whatever we find, it will help us begin to constrain the hypotheses.”
Synergies
All elements of EarthScope are tied together by the compelling science questions, adds NSF’s Whitcomb. “The aim is to be able to solve the dynamic equation for Earth as a whole.”
An example of the different projects working together would be the use of portable GPS receivers and seismometers around the San Andreas fault to get a broader view than from the borehole alone. “There are issues of how stress and strain move along and between faults,” says Hickman. “SAFOD will have instruments in the borehole, and the PBO instruments will give us a bigger picture of the deformation field.” Similarly, he adds, “seismic images from USArray can tell you there is an area with low velocity—it could be soft rock or it could be hard, fractured rock with high fluid pressure. It’s not until you drill with SAFOD that you can distinguish. By combining the tools, you can scale up from the borehole to get a full, three-dimensional picture of the San Andreas fault and its environment.”
In its full vision, EarthScope includes a fourth component: An inter-ferometric synthetic aperture radar (InSAR) satellite for imaging surface deformation. The satellite would sweep over the same area perhaps every eight days to get deformation measurements accurate to a centimeter roughly every 30 meters. The high spatial resolution “would complement the PBO remarkably well,” says Bernard Minster of the Scripps Institution of Oceanography. So far, though, NASA has repeatedly declined to fund such a satellite. Still, says Minster, “I am reasonably optimistic that it will happen.” An InSAR satellite would cost about $350 million or $400 million, he adds. “That’s dwarfed by the costs associated with earthquake risks.
EarthScope also plans to involve the public. High schools could, for example, host the traveling USArray seismic stations. “That’s a great opportunity,” says Roberta Rudnick, a geochemist at the University of Maryland, College Park, and a member of the EarthScope science and education committee, which advises NSF and serves as a liaison between the agency and the scientific community. “But it’s only a start. The interesting thing is, the MREFC is paying strictly for equipment. Just like the science, the education and outreach is going to come from proposals from the community. The vision is that EarthScope will become widely known in the US—and the whole world.”
Seismic stations such as this prototype (above) are to be set up in a grid pattern that will move gradually across the US. Deformations in Earth’s surface will be monitored by global positioning system receivers such as one being set up on Mt. Olympus in Washington (right).
Seismic stations such as this prototype (above) are to be set up in a grid pattern that will move gradually across the US. Deformations in Earth’s surface will be monitored by global positioning system receivers such as one being set up on Mt. Olympus in Washington (right).
A worm’s-eye view of the San Andreas fault: As part of EarthScope, a hole will be drilled next to the existing pilot hole (yellow line). Partway down, the hole will bend toward the fault plane (red curtain) to penetrate a seismically active area (bubbles represent the actual locations of past microearthquakes).
A worm’s-eye view of the San Andreas fault: As part of EarthScope, a hole will be drilled next to the existing pilot hole (yellow line). Partway down, the hole will bend toward the fault plane (red curtain) to penetrate a seismically active area (bubbles represent the actual locations of past microearthquakes).
