Modeling the conversion of hydroacoustic to seismic energy at island and continental margins: preliminary analysis of Ascension Island data

Seismic stations at islands and continental margins will be an essential component of the International Monitoring System (IMS) for event location and identification in support of Comprehensive Nuclear-Test-Ban Treaty (CTBT) monitoring. Particularly important will be the detection and analysis of hydroacoustic-to-seismic converted waves (T-phases) at island or continental margins. Acoustic waves generated by sources in or near the ocean propagate for long distances very efficiently due to the ocean sound speed channel (SOFAR) and low attenuation. When ocean propagating acoustic waves strike an island or continental margin they are converted to seismic (elastic) waves. We are using a finite difference code to model the conversion of hydroacoustic T-waves at an island or continental margin. Although ray-based methods are far more efficient for modeling long-range (> 1000 km) high-frequency hydroacoustic propagation, the finite difference method has the advantage of being able to model both acoustic and elastic wave propagation for a broad range of frequencies. The method allows us to perform simulations of T-phases to relatively high frequencies ({>=}10 Hz). Of particular interest is to identify factors that affect the efficiency of T-phase conversion, such as the topographic slope and roughness at the conversion point and elastic velocity structure within the island or continent. Previous studies have shown that efficient T-phase conversion occurs when the topographic slope at the conversion point is steep (Cansi and Bethoux, 1985; Talandier and Okal, 1998). Another factor impacting T-phase conversion may be the near-shore structure of the sound channel. It is well known that the depth to the sound channel axis decreases in shallow waters. This can weaken the channeled hydroacoustic wave. Elastic velocity structure within the island or continent will impact how the converted seismic wave is refracted to recording stations at the surface and thus impact the T-phase amplitudes. For this paper we will focus on validating the finite difference method for modeling T-phases in the ocean and land environments and on modeling T-phases observed by the May 1999 Ascension Island Experiment. A network of broadband seismometers on Ascension Island recorded a large number of offshore airgun shots. The shots occurred at all azimuths around the island and at ranges from 1-45 km. Measurements of signal amplitude and duration will be made to understand the variability of T-phase behavior on Ascension Island. The sensitivity to topographic slope and island structure will also be investigated