Imaging central Pacific upper mantle using P-wave tomography and receiver functions

Abstract:

Several aspects of the oceanic lithospheric mantle remain unknown, largely due to the lack of dense local seismic instrumentation. In particular, we do not understand the nature of lithosphere-asthenosphere boundary, sparsely observed mid-lithospheric velocity gradients, and dynamic processes below the plate that may play an essential role in controlling Earth surface evolution. The Pacific OBS Research into Convecting Asthenosphere (Pacific ORCA) experiment included a ~1 year deployment of an 500x500 km^2 OBS array consisting of 30 stations in the central Pacific Ocean on ~40Ma seafloor northeast of the Marquesas Islands. Sub-lithospheric small scale convection (SSC) is a promising theory accounting for the observed gravity lineations at this area, although it has not been seismically imaged. Here we present results from the first year of ORCA data: 3-component seismic records from 13 stations, and pressure data from 19 stations. Numerous data issues required careful quality control through sophisticated noise analysis and data anomaly removal. We also computed interchannel transfer functions to remove oceanic noise from the vertical component.

Teleseismic events (30-90) over Mw6.0 were used to calculate P-s receiver functions (RF). After waveform selection using a multi-stage Principal Component Analysis method, the stacked RFs clearly show crustal conversion phases that are well fit by synthetics for normal oceanic crust. We used a filter to remove conversion multiples that occlude mantle structure and migrated RFs to image deep discontinuities using a CCP stack. The result revealed ~6km Moho and 40-50km negative velocity gradient (NVG) beneath our station array; this NVG is shallower than the ~85km predicted thermal LAB depth (to 1300˚C isotherm), assuming conductive cooling. We also measured differential travel times for P-waves from both vertical and pressure components using multi-channel cross-correlation. We combined these data and tomographically inverted for 3-D P-wave velocity structure. Resolution tests show that our data can resolve variations in mantle structure at 100-300km depth. We find alternating velocity anomaly bands oriented in WNW-ESE direction, which is nearly parallel to the gravity lineations, offering potential support for the presence of SSC.