Shear Velocity Structure of the Crust and Uppermost Mantle across the Mariana Trench from Ambient Seismic Noise

Abstract:

We investigate the shear wave structure of the crust and uppermost mantle across the Northern and Central Mariana trench using ambient seismic noise recorded by land and ocean bottom seismographs (OBSs). The main goal of this study is to constrain velocity variations resulting from possible serpentinization of the incoming plate or the forearc mantle. Vertical component waveforms from 20 OBSs and 7 island stations deployed in 2012-2013 are correlated to reveal the empirical green functions. Due to the distinct structure difference across the trench, we adopt two different lithosphere models when calculating the phase velocity predictions. For each station pair, water depth estimated by averaging values along the great circle path connecting two stations is also introduced to predict the phase velocity. For each frequency, only high quality station pairs (SNR > 5) with distance larger than twice the wavelength are included. The dispersion curves are then interpolated to obtain the 2-D phase velocity map for period from 8 s to 32 s. Phase velocity maps from short periods (< 15 s) are strongly affected by the water depth. For periods between 15 s - 18 s, extremely low phase velocities near the trench result from water depth up to 9 km. Longer period results show fast phase velocity anomalies in the incoming plate and slow anomalies beneath the forearc, consistent with faster mantle velocities expected for the incoming plate. We will also show the results of inverting this phase velocity dataset for 3-D shear velocity structure. We are also investigating the use of higher mode Rayleigh waves, because in deep water regions the fundamental mode results fail to constrain the velocity structure of the shallowest part of the mantle.