SPOT imagery shakes up earthquake evaluation

One way seismologists study an earthquake is by measuring the ground displacements observed near the fault that slipped. The problem is that conventional geodetic techniques yield useful, but sparce information. Satellite imagery could soon become an attractive alternative because it regularly provides comprehensive and detailed images of the ground with high radiometric and geometric quality.

While satellite techniques that combine synthetic aperture radar (SAR) images with either interferometry or offsets have shown promise, they generally do not provide measurements near enough to the faults. The interferometric option has severe limitations related to data decorrelation and signal saturation. The technique of offsets has pixel size and speckle noise limitations. Rémi Michel and colleagues at the Laboratoire de Détection et de Géophysique (Bruyères le Châtel, France) think a more effective solution may be optical satellite imagery, namely Systéme pour l'Observation de la Terre (SPOT) panchromatic imagery.¹

The SPOT technique relies on two high-resolution visible push-broom optical devices onboard SPOT satellites to produce panchromatic images of a 60-km-wide ground area with a resolution approaching 10 m. Each satellite's optical device includes a Schmidt-like telescope (f = 1.082 m, F/3.3), a tunable mirror that allows selecting the location of the scene center, and a Diviseur de Ligne. This device connects four 1500-pixel charge-coupled device arrays to provide a 6000-pixel line of imagery as a satellite moves along its orbit. The SPOT images are composed of 6000 x 6000 pixels, and digitization can reach one byte (the effective dynamic may be lower). The optical center of images is at the center of the gravity of the platform that is submitted to roll, pitch, and yaw. Absolute location of the scene can be determined with an accuracy approaching 1 km.

Working with images from the Landers earthquake (magnitude 7.3, 1992 in California), the French research team produced near-fault measurement of ground displacements with an accuracy of about 10 cm and low-frequency measurement with an accuracy of approximately 1 m. According to Michel, the SPOT technique worked so well because it was combined with a method that compensated for other sources of image deformation and a dedicated subpixel correlator with an accuracy better than 20 cm rms. The scientists chose to study the 80-km-long Landers earthquake because it has already been studied extensively.

The images shown indicate ground displacement induced by the earthquake as measured from two separate SPOT images. These East-West and North-South components of the earthquake's displacement fields show a discontinuity that locally follows the map of the surface fault ruptures with an accuracy of 80 m. According to Michel, the measurements correlate well with an elastic modeling of the ground deformation that was constrained from numerous geodetic data. In addition, profiles in the offset field yield the measurement of the two-dimensional displacement on the fault with an accuracy of about 20 cm.

REFERENCE

• N. Van Puymbroeck et al., Appl. Optics 39, 3486 (July 10, 2000).