Gravity and seismic constraints on plate flexure and mantle rheology along the whole Louisville Ridge

Louisville Ridge in the southwest Pacific Ocean is a ~4200-km-long chain of submarine volcanoes generated at a hotspot presently located between the Heezen and Tula Fracture Zones, ~550 km northwest of the Pacific-Antarctic spreading ridge. Swath bathymetry surveys reveal the Louisville Ridge comprises seamounts, a number of which are guyots and so were once ocean islands. Seamount age increases progressively along the ridge, such that the youngest (unnamed) seamount is near the Pacific-Antarctic ridge while the oldest, Osbourn (~77-81 Ma), is located near the intersection of the ridge with the Tonga-Kermadec trench. Plate kinematic studies show a) the smooth trend of the ridge is copolar with the Hawaiian-Emperor seamount chain in the northwest Pacific Ocean, b) the ages at the main bends in the two chains are similar (~47 Ma), c) the difference in distance between same age seamounts in the two chains and the expected distance based on their present hotspot separation is small (±2^°) and, d) the Pacific plate as a whole has behaved rigidly for at least the past 50 Myr as it migrated northwest over fixed the Hawaii and Louisville hotspots. Studies of plate rigidity immediately beneath the Louisville Ridge, however, have yielded conflicting results. Previous studies suggest the elastic thickness, T_e, a proxy for the long-term flexural rigidity of the plates, is relatively high north of the main bend (~20-22 km) and relatively low (~16-18 km) to the south. However, seismic refraction data acquired north of the main bend along a ‘dip’ line during SONNE cruise SO195 at the 27.6^° S seamount yielded a low T_e (~10 km). Here, we use seismic refraction data acquired north of the bend along a ‘strike’ line, Profile C, during SONNE cruise SO215, together with ~1900 estimates of T_e derived from gravity data, to show that T_e is indeed low (6-10 km) at the northern end of the Louisville Ridge and then increases to ~26 km in the vicinity of the main bend at distance ~1309 km. These observations are consistent with the hypothesis that T_e is dependent on age, and hence thermal structure of the Pacific plate, at the time of volcano loading. However, the isotherm that controls T_e (276±10^oC) along the whole ridge is lower than at the Hawaiian-Emperor seamount chain (336±18 ^oC) and, interestingly, the bend-fault region of the proximal Tonga-Kermadec trench – outer rise system (342±35^oC). We examine here the implications of a ‘weak’ zone within an otherwise rigid Pacific plate for deformation models of brittle and ductile flow at lithospheric conditions based on extrapolations of data from experimental rock mechanics and for subduction initiation models where large downward flexures (up to 3.7 km) of oceanic and mantle crust may extend some thousands of km from a trench almost to a ridge.