Vertically Integrated Rheology of Deforming Oceanic Lithosphere

The tectonics of the oceans have traditionally been modeled in terms of rigid plates interacting at narrow boundaries. The now well-documented existence of diffuse oceanic plate boundaries, across which relative motion is distributed over hundreds to thousands of kilometers, demonstrates the need for a different approach to understanding the tectonics and geodynamics of a substantial fraction of oceanic lithosphere. A model that has usefully been applied to diffuse zones of continental deformation is that of a thin viscous sheet of fluid obeying a power-law rheology. The model has few adjustable parameters, typically a power-law exponent, n, and the Argand number [England & McKenzie, 1982], which is a measure of the size of buoyancy forces caused by the deformation, and which can be neglected for deformation of oceanic lithosphere. In prior investigations of a thin sheet of power-law fluid for continental regions, most studies have found that the most appropriate power-law exponent is ≈3 [e.g., England & Molnar 1991, 1997], but a value as large as ≈10 has been recently suggested by Dayem et al. [2009]. Because the rheology of oceanic lithosphere differs significantly from that of continental lithosphere, the most appropriate exponent may be larger than 3, and should in some sense be an appropriately weighted average between the properties of the upper lithosphere, which deforms brittlely and semi-brittlely, and for which the power-law exponent is n → ∞, and the lower lithosphere, which deforms by dislocation glide [Goetze 1978; Evans & Goetze 1979; Ratteron et al. 2003; Dayem et al. 2009; Mei et al. 2010], which obeys an exponential law, and by dislocation creep for which n≈3 [Sonder & England, 1986]. To estimate the appropriate power-law exponent consistent with laboratory experiments we determine strain rate as a function of applied end load on the lithosphere for various ages of lithosphere. We find that a power-law fluid well approximates the vertically integrated rheology of oceanic lithosphere determined from laboratory experiments and that the best-fitting power-law exponent for the vertically integrated rheology is insensitive to strain rate. We also find that, except for very young lithosphere (< ≈10 Ma old), the best-fitting power law exponent is insensitive to the age of the lithosphere, with the value of the exponent being between 14 and 16 when failing for thrust faulting for the flow laws of Kohlstedt et al [1995] and between 15 and 19 for more recently published flow laws. These results support the application of thin viscous sheet models to diffuse oceanic plate boundaries, such as the ones accommodating motion between the India, Capricorn, and Australia plates in the Indian Ocean.