Using Satellites to Forecast Earthquakes and Volcanoes
by Andy Dow
Violent earthquakes and volcanic eruptions have the power to alter features of our visible environment. Yet, they result from forces silently at work beneath the earthÕs surface. While many scientists look deep into our planet's core to understand these phenomena, some also see the benefits of taking a step back, a 500-mile step back, that is.
 |
| InSAR satellites scan particular points on the ground at separate times and measure displacement by referencing the phase (or position within the wave cycle) of the second returning wave against that of the first. The resulting InSAR image is a color-coded map displaying "fringes" or quantitative bands of deformation across the observed surface. Each cycle off color marks 28 mm of near-vertical displacement. |
How InSAR Works
InSAR satellites measure changes in the earthÕs crust by bouncing radar waves off the planetÕs surface and ÒcatchingÓ the return signal. Each satellite transmits a pulse of known wavelength to a particular point on the surface, where the wave is reflected back toward space. The satellite records variations in the intensity and phase of the returning pulse. By collecting this information for the same surface location at different times, the data precisely measures even the slightest deformation that has occurred at each point since the previous pass of the satellite. Measuring the phase of the returning waves proves a useful calculation tool: because electromagnetic radiation travels at the speed of light, scientists can use the wave phases and wavelength to infer changes in distance traveled and therefore the changes in uplift at the earthÕs surface. This allows for the high measurement sensitivity needed to detect centimeter-scale changes in the earthÕs crust.
The power of InSAR lies in its ability to generate maps (interferograms) spanning hundreds of kilometers with centimeterscale resolution of deformation. Yet in spite of the technologyÕs scope and remarkable detail, several factors limit its applicability. Vegetation poses the largest problem. In densely covered areas, changing surface features due to plant growth appear in the satellite measurements. ÒWe have no way of separating movement of plants on the surface from the soil underneath,Ó explains Zebker. ÒThat tends to obscure any underlying geophysical approach.Ó
Overcoming Surface Interference
To sift through this interference, Zebker and his research team have applied a method known as PS analysis, first developed by teams in Italy for studying changes in urban environments due to heat expansion. The team adapted this method to the study of natural surfaces, investigating areas that cannot usually be measured due to vegetation. PS exploits the lack of uniformity in a vegetated area. Because not all of the areaÕs fl ora is prohibitively dense, in a typical image of a vegetated region consisting of millions of pixels Ð each representing an InSAR radar point Ð a small fraction of the pixels, perhaps ten thousand, reflect signals that reached the surface, Zebker explains.
By identifying the points that actually measure surface motion, researchers can set up accurate networks of data samples to greatly refine models of deformation patterns. ÒYou measure enough points at [a high] enough density to infer the gross motion on the surface,Ó says Zebker. Stanford researchers have achieved success in PS-generated interferometry, primarily over volcanic regions. ZebkerÕs team has applied PS to vegetated regions including Long Valley in California, Mount St. Helens and the Galapagos Islands.
Although PS enables new levels of accuracy in seismic regions previously unsuitable for InSAR analysis, the process is data intensive and thus, very time-consuming. The technique must be applied to several years of monthly observations of a geographical location in order to extract accurate patterns. In addition, each area under study has different surface features that affect the radar signal received by the satellite. Researchers analyzing data must tailor the PS to each regionÕs site-specific conditions. Zebker noted that much research involves determining how to optimize data for specifi c regions by accounting for differences in vegetation and in rates and distribution of deformation with appropriately modified algorithms.
Using InSAR to Analyze Earthquakes
Well-known earthquake regions such as the San Andreas fault demand this kind of accurate InSAR analysis. To better understand earthquake behavior, scientists need a consistent series of points along the fault to track deformation patterns during times of mounting tectonic tension. The goal is to precisely monitor levels of deformation accompanying strain thresholds so that geophysicists can recognize patterns that portend seismic release. ÒPersistent scattering will help with that,Ó says Zebker, Òbecause weÕll have a good spatial map of where strain is accumulating and where itÕs not.Ó
 |
| An interferogram showing deformation during the August 1999 earthquake in Izmit, Turkey. The black and red lines indicate the lines of fault rupture where the greatest amount of strain occurred, as shown by the closeness of the displacement fringes. |
One hundred years after the 1906 San Francisco earthquake, InSAR research methods elicit new hope in the ability to forecast the time and location of future quakes with sufficient warning. Zebker points out that we are much closer to developing this ability for volcanoes than for earthquakes, primarily because the basic geophysics driving earthquakes are not as well understood. Even highly accurate InSAR data for fault regions only provide insights related to surface effects rather than underground causes. While scientists can note correlations between strain accumulation and seismic activity, ÒWhat we canÕt do is relate that strain to true probability that an earthquake will happen because we donÕt really know what happens ten kilometers down when the earthquake decides to go,Ó Zebker explains.
InSAR analysis can therefore best be viewed as a valuable component in deciphering the complex workings of the earth. Satellite interferometry with Stanford-affiliated progress in PS and InSAR will continue to complement research into physical seismic models. By Òstepping backÓ to record the earthÕs surface with enhanced satellite technology, scientists can now observe remarkable details of our dynamic planet and move closer to understanding its powerful geophysical processes.