Coseismic and postseismic deformation associated with the Mw = 6.4 February 24, 2004 Al Hoceima earthquake (Morocco) from InSAR: Evidence for triggered postseismic slip on neighboring faults

Abstract:

The Al Hoceima region of Northern Morocco is located within the east-west trending imbricated thrust-and-fold system of the Rif Mountain range that results from collision between the African and Eurasian plates. The transpressive tectonics and existence of a complex fault with thrust, normal and strike-slip faulting in the Rif probably reflect the rapidly changing local tectonic regime with block rotations during the Neogene and Quaternary (Meghraoui et al., 1996). The Al Hoceima earthquakes of the May 26, 1994 (Mw=6.0) and February 24, 2004 (Mw=6.4) are the largest seismic events in the last century that occurred on conjugate strike-slip faults trending approximately NNE-SSW and NW-SE on the Rif Mountain range (Akoglu et al., 2006). Models of the 2004 earthquake based on InSAR data suggest different fault ruptures with mainly NW-SE trending planar or curved shaped right-lateral strike-slip fault (Cakir et al., 2006; Biggs et al., 2006; Tahayt et al., 2009), consistent with the aftershocks distribution on surface and in deep (Van der Woerd et al., 2014). In this study, the postseismic deformation is studied using the Synthetic Aperture Radar Small Baseline (SBAS) technique. InSAR time series calculated from 15 Envisat ASAR images using Stanford Method of Persistent Scatterers (Hooper, 2008) reveal the subtle ground movements on the Al Hoceima region between 2004 and 2010 where a remarkable coseismic displacement was observed after the earthquake. SBAS analysis shows up to 4 cm cumulative line-of-sight movement towards and away from the satellite in the region of coseismic surface deformation, which is in good agreement with right-lateral strike-slip motion. The spatial pattern of the postseismic deformation field however requires not only afterslip on the coseismic fault but also aseismic slip on a fault that splays from the coseismic rupture to northeast. Reactivation of the fault splay is most likely caused by postseismic afterslip on the coseismic fault plane itself. On the basis of these new observations we revised the coseismic model using the same data set (one ascending and one descending interferogram) used by Cakir et al., (2006). The results suggest that the earthquake rupture took place on a fault with listric geometry and a synthetic fault on the hangingwall was reactivated during the postseismic period.