Dyke or pipe? Contributions of magnetic fabrics to the reconstruction of the geometry of an eroded subvolcanic body (Cadí basin, Pyrenees)

The reconstruction of the geometry of igneous bodies in inverted basins is often a difficult task, because most of their volume can be either below or above the topographic surface. Indirect methods, such as the determination of their internal fabric can give clues to constrain geometrical reconstructions. The Sant Salvador intrusion, located in the Late Carboniferous-Permian Cadí basin (Central-Eastern Pyrenees), is a good example for testing the application of magnetic fabrics to shallow igneous bodies because of (i) its characteristics in terms of crystal composition/orientation, showing a homogeneous petrofabric controlled by plagioclase crystals 0.4 mm in size, (ii) its good, though partial, preservation, (iii) its position in the stratigraphical succession, between volcaniclastic sediments and unconformably overlying red beds, and (iv) the many unknown factors related to the intrusion, including the geometry of the body and its depth and the amount of erosion after intrusion. Furthermore, the Sant Salvador intrusion provides opportunity for checking the correspondence between the petrofabric and the magnetic fabrics when the magnetic carriers of susceptibility (with an average value of 5217·10 SI) are mainly ferromagnetic. In this case the main magnetic carriers are magnetite and also hematite, the latter resulting from oxidation processes at the paleosurface. The results of Anisotropy of Magnetic Susceptibility show a bimodality in the K disposition after bedding restitution: vertical in the deeper areas of the intrusion and subhorizontal, E-W, in the uppermost stratigraphic positions with an intermediate necking zone in which transition from vertical to horizontal flow is interpreted to occur. Bands deflecting the mineral foliation at the micro scale have also a tectonic or thermal origin but do not seem to interfere with the magnetic ellipsoid axes. Columnar jointing is interpreted to be parallel to the long axis of the intrusion (and therefore to magma flo