A method for constraining the causes of active tectonic deformation using 2D finite element models in a Bayesian framework

Abstract:

Plate boundary deformation zones present a challenge in terms of understanding their underlying geodynamic drivers. We develop geodynamic models of the surface kinematics in selected plate boundary zones, and use Bayesian inference to determine the relative contribution of input parameters towards explaining the observations. Our models are 2D visco-elastic finite element models representing a vertical average over the lithosphere. Boundary conditions represent geodynamic processes: (1) far-field relative plate motions; (2) resistive fault tractions (we develop a new, efficient technique to impose constant-magnitude differential forces opposing fault slip); (3) forces from lateral variations in gravitational potential energy (GPE); (4) slab pull and trench suction at subduction zones; and (5) coupling between the lithosphere and the underlying asthenosphere. The magnitude of each of these is a parameter in a Bayesian analysis, which involves many forward models being run in a Metropolis-Hastings algorithm. The search yields a probability distribution over all parameters, allowing us to calculate mean/median parameter values, robustly estimate parameter uncertainties, and identify tradeoffs (i.e., parameter covariances).

We apply this technique to the Aegean Sea region (mainly Greece and western Turkey) and the Gibraltar region (from northwestern Africa to the Iberian Peninsula). The most important of the model input parameters appears to be the relative plate motions. Fortunately, plate motions are well constrained compared to the other parameters. Average resistive tractions on many major faults (e.g., the North Anatolian Fault) tend to be close to zero. In contrast, many of the smaller faults are effectively locked on a lithospheric scale. Pull forces associated with the Gibraltar slab are hardly transmitted to the overriding lithosphere in the western Mediterranean, whereas significant trench suction forces from the Hellenic slab act on the overriding Aegean Sea. GPE forces are necessary to explain local features in the velocity field, particularly in the Aegean Sea, but are less important for fitting the regional pattern of velocities.