Summary
The present study aims at building a 3‐D density model of a whole collisional mountain belt at the lithospheric scale. The geometry of internal structures is constrained by seismic studies, while density values are imposed by petrological considerations. The model must fit both the Bouguer and geoid anomalies. This approach is applied to the Pyrenees, for which numerous geophysical data are available. The crustal structure is constrained from the results of seismic profiles. The most important crustal feature is the large Moho jump which coincides with the surface expression of the North Pyrenean Fault, considered as the suture between the Iberian and Eurasian plates. The Eurasian crust is about 30 km thick, whereas the Iberian crust is 50 km thick in the central part of the range. Thick sedimentary basins at the Piedmont, and two large heavy intracrustal bodies located in the western and central parts of the range, are also important features of the crustal model. At the lithospheric scale, a low‐velocity strip under the North Pyrenean Fault down to 80–100 km depth, interpreted as the subduction of the Iberian lower crust, has been revealed by tomographic models. The density of this structure has been constrained using petrological arguments, which assume that the crustal material is transformed into eclogite at about 12–15 kbar (40–50 km depth). An eclogite‐type density value, much higher than the mantle density, is thus imposed from 50 to 100 km depth.
Theoretical Bouguer and geoid anomalies are computed for this 3‐D density model using a summation of density block anomalies. A good agreement is found with observations, giving strong support to the proposed model. The crustal structure, as seen by seismic profiles, explains most of the observed long‐wavelength negative Bouguer anomaly, as well the local anomalies related to the intracrustal bodies. A possible mantle origin of these bodies is ruled out. On the other hand, the transformation of the subducted lower crust into dense eclogite provides a successful fit to the long‐wavelength positive geoid anomalies. In addition, the water released during eclogite formation may help to explain the high conductivity zone detected down to 80 km south of the North Pyrenean Fault. The 3‐D model proposed here thus reconciles petrological considerations and geophysical data.