Fault identification for buried strike-slip earthquakes using InSAR: The 1994 and 2004 Al Hoceima, Morocco earthquakes

The 1994 M-w 6.0 and 2004 M-w 6.5 Al Hoceima earthquakes are the largest to have occurred in Morocco for 100 yr, and give valuable insight into the poorly understood tectonics of the area. Bodywave modelling indicates the earthquakes occurred on near-vertical, strike-slip faults with the nodal planes oriented NW-SE and NE-SW. Distinguishing between the primary fault plane and auxiliary planes, using either geodetic or seismic data, is difficult due to the spatial symmetry in deformation fields and radiation pattern of moderately sized, buried, strike-slip earthquakes. Preliminary studies, using aftershock locations and surface observations, have been unable to identify the orientation of the primary fault plane for either earthquake conclusively. We use radar interferometry and aftershock relocation of the earthquake sequence to resolve the ambiguity. For the 2004 earthquake, inverting the interferograms for a uniform slip model based either of the two potential nodal planes results in similar misfits to the data. However, the NE-SW best-fit fault plane has an unrealistically high fault slip-to-length ratio and we conclude the NW-SE striking nodal plane is the primary fault plane and slip was right lateral. We carry out tests on synthetic data for a buried strike-slip earthquake in which the orientation of the fault plane is known a priori. Independent of geometry, missing data, and correlated noise, models produced assuming the auxiliary plane to be the fault plane have very high fault slip-to-length ratios. The 1994 earthquake had a smaller magnitude and comparisons of model misfits and slip-to-length ratios do not conclusively indicate which of the nodal planes is the primary fault plane. Nonetheless, the InSAR data provides valuable information by improving the accuracy of the earthquake location by an order of magnitude. We carry out a multiple event relocation of the entire earthquake sequence, including aftershocks, making use of the absolute locations for the 1994 and 2004 main shocks from our InSAR study. The aftershock locations are consistent with a NW-SE orientated fault plane in 2004 and suggests that the 1994 earthquake occurred on a NE-SW fault; perpendicular to the fault which ruptured in 2004. Previous tectonic models of the area proposed a bookshelf model of block rotation with NNE-SSW left-lateral faults. This model requires modification to accommodate the observation of right-lateral slip on a NW-SE fault plane for the 2004 earthquake and we prefer to interpret the fault orientations as due to a zone of distributed shear with a right-lateral fault striking at similar to 115 degrees and conjugate, clockwise rotating, left-lateral faults striking at similar to 25 degrees.