InSAR Quake Research
Using InSAR to Study Earthquakes in South America
by Matt Pritchard, Princeton University and Mark Simons, California Institute of Technology
The Nazca Plate is subducting eastward into the mantle below the western coast of South America at about 7 cm yr -¹. This collision causes high levels of volcanic activity and the long-term uplift of the Andes by compressional tectonics. On shorter timescales, this deformation manifests itself as large earthquakes along the entire length of the west coast of South America, including several earthquakes with magnitudes greater than 7.5 occurring in the past decade.
Earthquakes radiate seismic waves and cause permanent local crustal deformation that can be modeled to determine exactly which parts of a fault slipped (both at the surface and at depth) during that event. It is important to know which parts of a fault have slipped for many reasons. One common application of fault slip maps is to better characterize earthquake hazard. When one segment of a fault ruptures it can increase or decrease the stress on some of the neighboring segments of the fault, making an earthquake reoccurrence more or less likely in these regions.
In addition, if we know how the stress in the crust was modified by the earthquake, and can precisely measure how these stresses relax with time (by affecting surface deformation), we can constrain the mechanical properties of these regions. Using RADARSAT-1 and ERS radar data made available through ASF, we are combining InSAR observations with GPS measurements of deformation and seismographic recordings of the radiated seismic energy to construct fault slip maps.
The arid central Andes are an ideal location for InSAR, because the radar scattering properties of the ground change little between observations. Even though the InSAR observations are of high quality, it is not a straightforward matter to determine what parts of the fault slipped in these large earthquakes.
One particular problem with this area is that a large amount of fault slip occurs under water where, currently, no measurements of deformation can be made. Nonetheless, by combining the InSAR, GPS and seismic data, we can estimate the sites of fault slip and determine how these earthquakes might impact surrounding areas which may become the sites of future large earthquakes.
The image on this page shows the color contours of the satellite line-of-sight component of ground deformation from radar interferograms of three shallow thrust subduction zone earthquakes draped over shaded relief and bathymetry.
The white outlines enclose the approximate rupture areas of large earthquakes of the 19th century that will possibly re-rupture in the 21st century. Black lines show political borders, and the red line is the Peru-Chile trench. The region of interest is indicated in the reference map at the lower left. In the 1995, 1996 and 2001 earthquakes, South America moved to the west, but the events look slightly different because of the different locations of slip on the fault interface relative to the coastline and the size of each earthquake. Because the radar satellites measure primarily vertical deformation, we can interpret the gross features as portions of the ground that were uplifted or subsided.
Typically, for subduction zone earthquakes, we detect primarily subsidence on land, with uplift offshore. For the 1995 earthquake, a small region of dry land (the peninsula) was uplifted, and the closed contours in the interferogram are mostly caused by the on-land subsidence. For the 1996 earthquake, the slip was closer to land, so more uplift is recorded onshore, but the closed contours represent subsidence. Most of the fault slip from the 2001 earthquake was off-shore, so only subsidence is measured on land.