Transverse ridges are regions of anomalously high uplift that run parallel to many of the transform faults at mid-ocean ridges. Previous models of the formation of these ridges generally fail to explain the magnitude of uplift (between 1 and 8 km) or account for geophysical observations (which indicate that the ridges are dynamically uplifted by neutral or heavy, but not buoyant, material). However, the potentially large Maxwell relaxation time of the lithosphere and the high strain rates and finite deformation of transform faults imply that nonlinear viscoelasticity may play an important role in the formation of these ridges. We present simple theoretical and experimental models of viscoelastic flow beneath a transform fault to show that the purely horizontal motion of the fault can generate vertical uplift typical of transverse ridges. The theoretical model is of an infinite half-space of fluid driven from above by two plates moving parallel and in opposite directions to one another. The constitutive relation for the viscoelastic rheology is for a second-order fluid; i.e., nonlinear viscoelastic effects are treated as perturbations to Newtonian, purely viscous flow. The theoretical model not only predicts the two transverse ridges that are often observed on either side of the transform fault, but the central transform valley that separates them as well. For typical lithospheric flexural rigidity and deformation associated with transverse ridges, the viscoelastic effect provides sufficient vertical stress to produce the observed uplift.

Further, the dynamic uplift is a purely mechanical effect (i.e., it does not involve buoyant material) which corresponds to geophysical observations. The experimental model is comprised of a viscoelastic fluid (2% aqueous solution of high molecular weight carboxymethylcellulose) driven by a rotating plate. This is analogous to a transform fault because the boundary of a spinning circular tectonic plate (surrounded by stationary plates) is one continuous transform fault. The experiment produces a large ridge of fluid along the edge of the plate with a slight trough outside of the ridge; this result is qualitatively described by a cylindrical version of the above theoretical model. In conclusion, both experimental and theoretical models show that when nonlinear viscoelasticity is accounted for, uplift characteristic of transverse ridges can be generated from purely horizontal motion typical of transform faults. ¿ American Geophysical Union 1992