Crustal and Upper Mantle Structure of the Louisville Ridge Seamount Chain at Its Intersection with the Tonga-Kermadec Subduction Zone

Friday, 19 December 2014
Adam Hackett Robinson1, Christine Peirce1, Wanda Rose Stratford1, Anthony Brian Watts2 and Ingo Grevemeyer3, (1)University of Durham, Durham, United Kingdom, (2)Univ Oxford, Oxford, United Kingdom, (3)GEOMAR Helmholtz Centre for Ocean Research Kiel, Kiel, Germany
The Louisville Ridge Seamount Chain (LRSC) is a 4300 km-long chain of Cretaceous-Cenozoic seamounts. The oldest extant seamount (~77-79 Ma) intersects the Tonga-Kermadec subduction zone at ~26S, coinciding with the divide between the Tonga forearc morphology to the north and the contrasting Kermadec forearc to the south.

Multichannel seismic reflection (MCS), wide-angle (WA) refraction, gravity, magnetic, swath bathymetry, and Parasound data were collected along an ~725 km profile over the LRSC and the overriding plate during R/V Sonne cruise SO215 in 2011. The objectives of this study are to determine: 1) the structure of the LRSC and pre-subduction crust, identifying any along-ridge variation in crustal structure or magmatic underplating, 2) how seamounts interact with the overriding plate during subduction, and 3) how along-ridge variation in topography and crustal structure relate to post-collisional forearc uplift. Interpretation of these features will be complemented by previous profiles acquired perpendicular to the LRSC and adjacent parts of the trench.

We apply a forward ray-tracing approach to the WA data obtained by 55 OBS to develop a crustal and upper mantle velocity model. The initial model is developed using MCS data to determine the sediment thickness and velocity, and this is underlain by average thickness and velocity oceanic crust and upper mantle. The resulting best-fit velocity model will be used to re-stack the MCS data, and tomographic inversion of travel times will be performed to test uniqueness. The velocity model will be converted into a density model and the gravity anomaly calculated and compared with the observed free-air gravity anomaly. Here we present an in-progress LRSC model, which will be used to characterize seamount structure prior to subduction at the trench and guide investigations as to whether any intact seamounts older than 77-79 Ma are present on the subducting plate interface.