Are the seismological and geological observations of the Al Hoceima (Morocco, Rif) 2004 earthquake (M = 6.3) contradictory? J. Galindo-Zaldívar a, ⁎, A. Chalouan b , O. Azzouz c , C. Sanz de Galdeano d , F. Anahnah a , L. Ameza c , P. Ruano a , A. Pedrera a , A. Ruiz-Constán a , C. Marín-Lechado e , M. Benmakhlouf f , A.C. López-Garrido d , M. Ahmamou b , R. Saji g , F.J. Roldán-García e , M. Akil b , A. Chabli b a Departamento de Geodinámica, Universidad de Granada, Spain b Département de Géologie, Faculté des Sciences, Université Mohammed V—Agdal, Morocco c Département de Géologie, Faculté des Sciences, Université Mohammed premier—Oujda, Morocco d IACT, CSIC-Universidad de Granada, Spain e Instituto Geológico y Minero de España, Spain f Département de Géologie, Faculté des Sciences, Université Abdelmalek Esaadi, Tetouan, Morocco g Département de Géologie, Faculté des Sciences, Université Caddi Ayyad, Beni Mellal, Morocco abstract article info Article history: Received 25 March 2008 Received in revised form 22 October 2008 Accepted 25 November 2008 Available online 3 December 2008 Keywords: Coseismic deformation Seismogenic faults Recent tectonics Crustal detachments Western Mediterranean Seismic hazard is associated with recent and present fault activity in mountain ranges. In the Betic-Rif alpine mountain chain, tectonic activity started in the Cretaceous, and topographic uplift continues since Tortonian times as a consequence of the NW–SE oblique convergence between Africa and Eurasia. The deformation is active and produces seismicity that sometimes has catastrophic consequences. The Al Hoceima earthquake (February 24, 2004), considered one of the largest earthquakes ever recorded instrumentally in the westernmost Mediterranean (M = 6.3), caused great damage in the region. Seismological studies agree that the main shock was situated on land, at the limit between the External and Internal Zones of the Rif, at a depth of 10–14 km. The focal mechanism points to a strike-slip solution with a NW–SE oriented P axis, quite similar to those of the significant 1994 earthquake swarm located to the north. The epicenter aftershocks distribution would signal the presence of a N–S oriented sinistral fault, activated by the NW–SE regional compression associated to plate boundary convergence. In this setting, the seismogenic fault ruptures related to these seismic events are expected to have reached the Earth's surface. However, detailed field work carried out 1 month after the earthquake does not evidence any N–S strike-slip coseismic fault in the epicentral area. The main observed effects were landslides, damages to constructions, and locally open cracks indicating an unexpected NW–SE extension. Scarce N–S faults are normal, the main ones being located several kilometers away from the epicentral area. To explain this apparent contradiction between geological and seismological observations, we propose a decoupled tectonic model with crustal detachments that separate a deep brittle crust from an upper crust undergoing uplift, and the development of large folds and normal faults. This geological setting, common to internal zones of cordilleras, may need to be taken into account in future paleoseismicity studies and in the assessment of seismic hazard. © 2008 Elsevier B.V. All rights reserved. 1. Introduction Fault activity with associated seismicity is an eminent natural hazard. To understand the seismic activity of a region, active tectonic studies combine geological and seismological researches. The geolo- gical observations attempt to identify the main Quaternary tectonic structures and their recent evolution in order to predict as far as possible the depths and magnitudes of the largest earthquakes that may occur in a region and finally to asses the building plans of a territory. Most active tectonic methods started by analogy to the detailed studies surrounding the San Andreas Fault, responsible for the San Francisco earthquake (1906, M =7.9). Yet because it is a transform fault, developed in a particular tectonic context featuring parallel and orthogonal main stresses (Sylvester, 1988), the characteristics, applied techniques and results cannot necessarily be extrapolated to other regions. Active tectonic studies combine surface geological and geophysical observations with geodetical and trenching techniques. Paleoseismological studies provide quantitative data on individual faults such as fault slip rate, coseismic displacement, earthquake recurrence interval, and time elapsed since the last large earthquake. Tectonic plates move continuously, and in non-cohesive rocks the active faults may slip progressively, by creeping, with no accumulation of Tectonophysics 475 (2009) 59–67 ⁎ Corresponding author. Fax: +34 958248527. E-mail address: jgalindo@ugr.es (J. Galindo-Zaldívar). 0040-1951/$ – see front matter © 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.tecto.2008.11.018 Contents lists available at ScienceDirect Tectonophysics journal homepage: www.elsevier.com/locate/tecto