Modeling of Conversion of Seismic to Acoustic Waves at the Seafloor Interface

Hydroacoustic waves are generated at the seafloor interface, by conversion of seismic waves and travel in the water column within the SOFAR channel with little attenuation. Recording T-waves with widespread arrays of autonomous hydrophones moored in the SOFAR channel allows to detect and localize many small-magnitude earthquakes in oceanic areas. However, hydroacoustic data cannot be used straightforwardly in seismic interpretations. In particular, because the physics of the seismic to acoustic conversion and the acoustic propagation is not completely understood, no direct information on the event magnitudes, focal mechanisms and focal depths can be directly derived from the hydroacoustic signals. In order to overcome some of these limitations, we have developed a mechanical model of the conversion from seismic to acoustic waves at the seafloor interface. The modelling is achieved through major adaptations of the 2D- finite element code "FLUSOL", which was originally developed to model fluid to solid energy conversion. Velocity displacement module within fluids and solids are derived from the stress and pressure computed for each grid element. We are able to model successfully, over a 10 x 10 km-grid, the seismic to acoustic conversion of waves generated by a source in the crust. Our model shows that a source with a high S-wave content appear to be more efficient in producing T-waves than a simple explosive source that only generates P-waves. Future work include the modelling of the conversion by more realistic seafloor topographies. Finally, we will use the output of SOLFLU as input to standard long-range acoustic propagation codes made available by the marine acoustics community. The modelled T-waves generated by various source mechanisms (tectonic or magmatic) will then be compared with real data to validate our conversion model.