Construction and subduction of the Louisville Ridge, SW Pacific—insights from wide-angle seismic data modelling

The Louisville Ridge is a ca. 4000 km-long chain of seamounts in the SW Pacific that is currently being subducted at the Tonga-Kermadec trench. The Pacific Plate, on which the chain sits, is subducting obliquely beneath the Indo-Australian Plate. Combined with the oblique strike of the chain relative to the margin, this results in the southward migration of the ridge-trench intersection and leads to significant along-trench variation in forearc morphology as a result of tectonic erosion processes. To understand how the subduction of such large-scale plate topography controls forearc deformation, knowledge of the structure of the seamounts themselves and the crust upon which they lie, and how these seamounts are deformed prior to and on entering the trench is required. The TOTAL (Tonga Thrust earthquake Asperity at Louisville Ridge) project aimed to address these questions by undertaking a multidisciplinary geophysical study of the ridge-trench intersection and surrounding region, as part of which multichannel and wide-angle seismic, gravity and swath bathymetry data were acquired along a similar to 750 km-long profile extending along the Louisville Ridge and into the adjacent Tonga forearc. We show that each of the largest, single edifice seamounts (called Osbourn and 27.6 degrees S) imaged has a discrete core of elevated seismic velocity (V-p >= 6.0 km s(-1)) and density (2600 kg m(-3)) relative to the adjacent Pacific oceanic crust, reaching to within 1.0-1.5 km of the seabed at their summits. However, there is no evidence of significant crustal thickening associated with individual seamounts, or that the crust beneath the Louisville Ridge Seamount Chain as a whole is significantly thicker than the surrounding oceanic crust of the Pacific Plate. Despite significant forearc deformation, we find no evidence to suggest that the most recent seamount of the Louisville Ridge to have been subducted, was subducted intact. The degree of plate bend-related faulting being experienced by the next seamount to subduct (Osbourn) suggests that they may instead be disarticulated to a size smaller than the imaging resolution in the trench region. In addition, distinguishing between seamount flank and intraseamount saddle material based on seismic velocity alone is not possible. Therefore, determining how, and where, already subducted seamounts are located beneath the forearc of the overriding plate is entirely dependent on imaging any high velocity core, and that core having remained relatively intact.