ON THE ORIGIN OF DEVIATORIC STRESSES IN THE LITHOSPHERE

Mantle circulation models capable of predicting satisfactory plate velocities and geoid are tested for their ability to generate adequate lithospheric stresses. The model Earth mantle is driven by mass heterogeneities defined by seismic tomography and slab distribution. Its surface is divided into 11 freely moving plates. The mantle circulation produces shear stresses at the base of each plate. These are used to compute a dynamical component of the lithospheric stresses which is added to another component induced by variations of the Moho depth inside the lithosphere itself. The extensional state of the continents caused by their thick crust is more than compensated by a compressional tendency related to their high seismic velocity roots, interpreted as dense roots. Strong upwellings beneath equatorial regions with geoid highs could also generate ammalous topography and extensional stresses. The long-wavelength stress pattern proposed by our models cannot easily be compared with available data: the internal loads are still poorly defined, the tested Earth model may be too simple, and the observations are unadequately distributed.