Eastward Mantle Drag on the Caribbean Plate: Insights from Neotectonic Modeling

Thin-shell finite element techniques have been applied to study the neotectonics of the Caribbean plate. Model domain comprises the entire Caribbean plate and it is outlined by fault elements. Kinematic conditions applied at the lateral boundaries represent motion of Cocos, Nazca, South America and North America plates with respect to the Caribbean plate. The velocity of this plate was not fixed, so its predicted velocity was used as a test of the model quality. We also included major intraplate active or potentially active faults in order to provide some estimates of slip rates, which is relevant for their relative long-term seismic hazard evaluation. We have obtained different sets of models varying the lateral and basal boundary conditions. Velocity conditions applied at grid sides were computed assuming poles and rotation velocities deduced from global plate model NUVEL-1A and from a number of published geodetic studies. We have also considered different possibilities for shear tractions from lower mantle flow acting on the base of the Caribbean plate. We have tried the following basal boundary condition: (1) no basal traction; (2) a lower mantle static with respect to Africa; (3) an eastward lower mantle flow. Different models have been computed with systematic variation of the fault friction coefficient. The quality of the models has been evaluated by comparing their predictions of anelastic strain rates, vertically-integrated stresses and velocity fields to data on seismic strain rate, most compressive horizontal principal stress direction, horizontal velocity deduced from geodetic studies and estimates of fault slip rates. Best results are obtained assuming low fault friction coefficient and boundary conditions at model sides deduced from geodetic studies. Our validation datasets require velocities of North America and South America plates relative to the Caribbean plate higher than those of model NUVEL-1A. Moreover our results clearly indicate that the only successful models are those with a significant basal drag on the Caribbean lithosphere exerted by eastward lower mantle flow.