Coseismic deformation associated with the November 1995, MW = 7.1 Nuweiba earthquake, Gulf of Elat (Aqaba), detected by synthetic aperture radar interferometry


  • Gidon Baer,

  • David Sandwell,

  • Simon Williams,

  • Yehuda Bock,

  • Gadi Shamir


The November 22, 1995, MW = 7.1 Nuweiba earthquake occurred along one of the left-stepping segments of the Dead Sea Transform in the Gulf of Elat (Aqaba). Although it was the largest earthquake along this fault in the last few centuries, little is yet known about the geometry of the rupture, the slip distribution along it, and the nature of postseismic deformation following the main shock. In this study we examine the surface deformation pattern during the coseismic phase of the earthquake in an attempt to better elucidate the earthquake rupture process. As the entire rupture zone was beneath the waters of the Gulf, and there is very little Global Positioning System (GPS) data available in the region for the period spanning the earthquake, interferometric synthetic aperture radar (INSAR) provides the only source of information of surface deformation associated with this earthquake. We chose four synthetic aperture radar (SAR) scenes of about 90 × 90 km each spanning the rupture area, imaged by the ERS-1 and ERS-2 satellites. The coseismic interferograms show contours of equal satellite-to-ground range changes that correspond to surface displacements due to the earthquake rupture. Interferograms that span the earthquake by 1 week show similar fringe patterns as those that span the earthquake by 6 months, suggesting that postseismic deformation is minor or confined to the first week after the earthquake. A high displacement gradient is seen on the western side of the Gulf, 20–40 km south of Elat and Aqaba, where the total satellite-to-ground range changes are at least 15 cm. The displacement gradient is relatively uniform on the eastern side of the Gulf, and the range changes are less than 10 cm. To interpret these results, we compare them to synthetic interferograms generated by elastic dislocation models with a variety of fault parameters. Although selecting the best fit fault parameters is nonunique, we are able to generate a group of simplified model interferograms that provide a reasonable fit to the coseismic interferogram and serve to constrain the location of the fault. The present analysis shows that if the rupture reached the Gulf-bottom surface, the mean sinistral slip along the fault is constrained to about 1.4 m. If surface rupture did not occur, the average sinistral slip is constrained to the range of 1.4–3 m for a fault patch buried 0–4 km below the Gulf-bottom surface, respectively, with a minor normal component.