Flexural modeling of circum-Pacific trench – outer rise systems and its implications for mantle rheology

Abstract:

Oceanic flexure studies suggest that the long-term strength and effective elastic thickness, T_e, of oceanic lithosphere increases with age due to thermal cooling. A recent study at trench – outer rise systems (Bry and White, 2007, JGR), however, has questioned whether such a relationship exists. In order to reassess this issue, we model trench-perpendicular, ensemble-averaged profiles of satellite-derived free-air gravity anomalies at the trench - outer-rise of circum-Pacific subduction zones. A broken elastic plate model is used with an inverse approach, in which we iterate for T_e and invert for the applied vertical force and bending moment at the trench using damped least squares. We first model the profiles using a plate of constant T_e. Results show that, to first order, oceanic lithosphere does indeed strengthen with age. However, there is scatter, which we attribute to variations in plate curvature and axial loading, and the effects of hotspot volcanism. Comparisons between our results and those of Bry and White reveal that the discrepancy can be explained by the different ways the two studies treat the regional long-wavelength gravity field, the position of the plate break and the applied bending moment. In many subduction zones, a constant T_e plate cannot fully explain the amplitude and wavelength of the bulge and the high curvature of the outer trench slope. We therefore model the profiles with a plate that is allowed to weaken trenchward of the outer-rise. The weakening is attributed to inelastic yielding, evidence of which is manifest in swath bathymetry and seismicity data. In an attempt to constrain experimentally derived low temperature flow laws for olivine, we compare our inverted T_e values with the predictions of viscoelastic-plastic yield strength envelopes derived from recently published flow laws. We test six flow laws, finding that Mei et al (2010, JGR) is too strong to fit our inverted T_e values. This is in accordance with the findings of Zhong and Watts (2013, JGR), who modelled Hawaiian Island loading using a 3D multilayered viscoelastic plate, suggesting that their result maybe generally true for oceanic lithosphere.