A new tomographic image of the Pyrenean lithosphere from teleseismic data Annie Souriau a, ⁎, Sébastien Chevrot a , Carme Olivera b a CNRS-UPS-UMR5562, Observatoire Midi-Pyrénées,14 Avenue Edouard Belin, 31400 Toulouse, France b Institut Geològic de Catalunya, Balmes 209-211, 08006 Barcelona, Spain abstract article info Article history: Received 26 March 2008 Received in revised form 30 July 2008 Accepted 19 August 2008 Available online 31 August 2008 Keywords: Seismology Seismic tomography Lithosphere Pyrenees A new tomographic model of the Pyrenean lithosphere is determined down to 200 km depth from teleseismic P and PKP travel times, with a lateral resolution of 0.25°. Compared to previous models, two important improvements are 1) a larger number of stations with a more even distribution, in particular to the west of the range, and 2) the introduction, before inversion, of crustal corrections inferred from previous refraction and reflection experiments. This last point is crucial because a strong Moho jump (up to 20 km) is present at the North Pyrenean Fault, the former boundary between Eurasian and Iberian plates. The comparison of the models obtained with and without crustal corrections reveals the strong contamination of the models by the crust down to 100 km depth. In the uncorrected model, a large strip with negative P-velocity anomalies, previously interpreted as subduction of lower crust, is observed. It disappears in the corrected model. Moreover, the introduction of crustal corrections allows us to reveal short wavelength heterogeneities which were hidden by the crustal signal. An attempt is made to relate the heterogeneities revealed by the tomographic model with the tectonic history of the Pyrenees, in particular with the Alpine orogeny. The Alpine phase includes an extensive episode with generation of the thin continental crust and possibly the opening of an oceanic sea floor, and then a compressive stage. In our model, no signature of an oceanic subducted slab could be detected all along the range, a result which rules out the opening of a large oceanic floor before the compressive stage. A subduction of continental crust is possible but, due to the transformation of lower crust into eclogite at depth, it can not be detected by seismological methods, whereas it was observed from electrical and gravity data. To the East of the range, large heterogeneities with low velocities are ascribable to the Neogene extension related to the rotation of the Corso–Sardo block and the opening of the Gulf of Lion. A prominent high velocity anomaly extending down to 200 km in eastern-central Pyrenees could possibly be interpreted as a detached piece of the Tethys slab. In north of Iberia outside the range, deep (down to 200 km) low velocity structures oriented N130°E are probably related to Hercynian orogeny. © 2008 Elsevier B.V. All rights reserved. 1. Introduction Knowledge of the lithospheric structure is a key information, toge- ther with geological records, for retracing the history of the building of a mountain range. In particular, geometry of the subducted materials may give constraints on the global convergence rate of the two plates before and during the build-up of the range. For the Pyrenees, which result from the convergence of the Iberian and Eurasian plates, this information is of great interest, because there are strong controversies concerning the global rate of shortening of the range, from 50 km to 400 km, the possible variations of this rate from east to west and the amount of oceanic lithosphere and con- tinental thinned crust which were subducted during the convergence (see reviews in Olivet, 1996; Sibuet et al., 2004; Capitanio and Goes, 2006). There are also some debates about the mechanisms of shor- tening after the beginning of continental collision, which could be either homogeneous crustal thickening, crustal stacking, or continen- tal subduction. A tomographic image of the upper mantle has been previously obtained from an analysis of P-wave teleseismic records at the Pyrenean stations (Souriau and Granet, 1995, hereafter noted SG95), but it suffered from the great heterogeneity and poor distribution of seismic stations, with a lack of stations in the center and at the western end of the range. Other limitations were the rather small number of records, and the poor quality of phase readings at some analog recording stations. Here, we propose a new tomographic model for P-wave velocities, which is based on a new data set of much better quality. It benefits from a redeployment of permanent digital stations in 1996–97 with a more even distribution on both sides of the range. It also benefits from a better knowledge of the crustal structure, which allows us to introduce crustal corrections of travel time data before inversion for subcrustal structures. Tectonophysics 460 (2008) 206–214 ⁎ Corresponding author. Fax: +33 5 61 33 29 00. E-mail addresses: Annie.Souriau@dtp.obs-mip.fr (A. Souriau), Sebastien.Chevrot@dtp.obs-mip.fr (S. Chevrot), colivera@igc.cat (C. Olivera). 0040-1951/$ – see front matter © 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.tecto.2008.08.014 Contents lists available at ScienceDirect Tectonophysics journal homepage: www.elsevier.com/locate/tecto