Numerical Simulations of Lithospheric Mantle Delamination
Abstract
Sudden uplift, extension, and increased igneous activity are often explained by rapid mechanical thinning of the lithospheric mantle (e.g., Kay and Kay (1993)). Two main thinning mechanisms have been proposed, convective removal of a thickened lithospheric root or delamination at the favor of a crustal decoupling level. The latter mechanism was first defined by Bird (1979) as the whole mantle lithosphere peeling away from the crust by the horizontal propagation of a shear zone at the crust mantle boundary. We performed 2-D numerical simulations of convection using a viscoplastic rheology with an effective viscosity depending strongly on temperature, depth, composition (crust/mantle), and stress. The simulations develop in four steps. (1) We first obtain "classical" sublithospheric convection for a long time period (~100~Myrs), yielding a slightly heterogeneous lithospheric temperature structure. (2) At some time, in some simulations, a strong thinning of the mantle occurs progressively in a small area (~100~km wide). This process puts the asthenosphere in direct contact with the lower crust. (3) Large pieces of mantle lithosphere then quickly sink into the mantle, by the horizontal propagation of a detachment level away from the "asthenospheric conduit". (4) Delamination recesses or stops when the lithospheric mantle slab break off. We determine the parameters (crustal thicknesses, activation energies, and friction coefficients) leading to the initiation of delamination (step 2). We find that delamination initiates where the Moho temperature is the higher, as soon as the crust and mantle viscosities are sufficiently low. Delamination may thus occur on Earth in orogenic settings when the crust is thick and the Moho temperature exceeds ~8000C. Simulations preformed with variable crustal thicknesses show that delamination initiates at the base of the deeper crust, but does not propagate along the base of the shallower crust. A few 3-D numerical simulation have been performed using a Newtonian rheology with a viscosity depending strongly on temperature. A sharp viscosity increase is imposed within the lithosphere at a given depth to simulate a decupling level at the crust/mantle boundary. We then compare these 3-D simulations to 2-D simulations performed in the same conditions, to highlight differences between 2-D and 3-D delamination geometries. References: Bird, P., Continental delamination and the Colorado Plateau, \it J. Geophys. Res., 84, 7561-7571, 1979. Kay, R. W. and Kay, S. M., Delamination and delamination magmatism, \it Tectonophysics, 219, 177-189, 1993.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2002
- Bibcode:
- 2002AGUFM.T12F..09M
- Keywords:
-
- 8110 Continental tectonics: general (0905);
- 8120 Dynamics of lithosphere and mantle: general;
- 8164 Stresses: crust and lithosphere