 |
Giampiero Iaffaldano Department of Earth and Planetary Sciences, Harvard University |
| Research |
Understanding the dynamics of global plate motion is one of the most important problems in geophysics today. Mantle convection is commonly accepted as the driving force for plates, but while the kinematics of plate movements is well known from increasingly-detailed geodetic and paleomagnetic observations, we still lack a rigorous description of the dynamics of the coupled mantle convection-plate motion system.
Over the past 10 years there has been much progress in developing and exploring computational high-resolution 3-D mantle convection models. These geodynamic earth models solve the conservation equations (mass, momentum, energy) for a highly viscous (Stokes) fluid and provide a first-order estimate of buoyancy forces driving plate motions. While these advances in mantle convection modeling are encouraging, the main shortcoming today is their lack of a realistic treatment of the lithosphere, and in particular of faults which arise from the brittle nature of the uppermost cold regions of plates.
Independently of advances in mantle convection models, there has been substantial progress in the development of global lithospheric models. Based on stress-equilibrium and sophisticated, empirical rheology laws these models include finite element formulations to account explicitly for surface topography, lithosphere structure and geological faults through the use, for example, of contact elements. In the past these models have been used to compute global plate velocities and stress distributions only from parameterized shear tractions at the base of the lithosphere.
The logical step to take is merging these two independent classes of models in order to gain a comprehensive modeling ability of the coupled plate tectonic/mantle convection system. During my Ph.D. I have focused on coupling two of the most advanced numerical models of mantle flow and lithosphere dynamics, such that I am in possession now of a dynamically consistent model for lithosphere and mantle dynamics, where forces driving and resisting plate motions account for mantle shear tractions exerted at the base of plates as well as deviatoric stresses originated by regional variations in lithosphere structure and topography.
The global mantle convection code I am using is TERRA developed by Dr. Baumgardner (Los Alamos National Laboratory) and Prof. Bunge (at the Geophysics Section of LMU Munich). The code is based on an icosahedral grid of finite elements and an efficient multigrid algorithm to solve the elliptic problem arising from the momentum equation. The code is fully parallelized and performs well on PC-Beowulf clusters.
The lithosphere motion code I use is SHELLS developed by Prof. Bird (UCLA), based on 2-D (thin shell) triangular global grid it is capable to provide surface velocity and stress maps.
With the aid of such joint numerical models and by exploiting the enormous amount of plate motions data available today through both paleomagnetic and geodetic techniques, I direct my research toward a better understanding of forces acting upon plates and causing their motion changes. The ability to consider past as well as present plate motions is in fact an incredible source of information to understand the dynamics of lithosphere. The principle of inertia tells us that a change in plate motion must necessarily be related to a change in one or more forces acting upon plates. I have recently proved the potential and the effectiveness of such joint approach by predicting the convergence rate reduction between Nazca and South America plates over the past 10 m.y. following the growth of the Andean mountain belt. The results from my models are confirmed by the record of plate motions available through paleomagnetic as well as geodetic data. By accurately reproducing the observed convergent motion I demonstrate that surface topography generated at convergent margins is a key factor controlling the long term evolution of plate motion. Specifically, the topographic load of large mountain belts and plateaus consumes a significant amount of the driving force available for plate tectonics, by increasing resisting forces between downgoing and overriding plates.