Frequency Dependent Sound Speed and Attenuation Measurements in Seafloor Sands from 1 to 400 KHZ

Presented at Boundary Influences in High Frequency, Shallow Water Acoustics held in Bath, UK on 5-9 September 2005. Published in the Proceedings of Boundary Influences in High Frequency, Shallow Water Acoustics, p57-64, 2005. The original document contains color images.

As part of the SAX04 experiment, sound speed and attenuation were measured in the seabed off of Ft. Walton Beach, Florida, USA, at frequencies from 1 to 400 kHz. For the lowest frequencies, from 1 to 20 kHz, signals from two acoustic sources were recorded on an array of hydrophones emplaced at depths of up to 1 m below the seafloor within a 4 m by 4 m area. At frequencies from 15 to 200 kHz, sound speed and attenuation data was obtained using the In Situ Sediment Acoustic Measurement System (ISSAMS), a linear array of four piezoelectric probes that were inserted into the seabed to a depth of 0.3 m. Sound speed measurements were also made at 60, 100, 200, and 400 kHz on diver-collected cores using four separate pairs of ultrasonic transducers. Sound speed data from the core and ISSAMS measurements demonstrates an essentially uniform sound speed near 1780 m/sec. Preliminary analysis of the sound speeds from the low frequency array data demonstrate an anomalous increase in sound speed values with decreasing frequency when sources were emplaced below the seafloor, attributable to the influence of multi-path arrivals. Data from source locations at the seafloor demonstrate a decrease in sound speed with decreasing frequency, consistent with the trend predicted by the Biot model, but with large uncertainties. Attenuation values from seafloor source positions follow a square-root frequency dependence below 20 kHz, similar to the Biot predicted trend, while the core and ISSAMS data demonstrate a near linear trend at higher frequencies. A fit of the Buckingham model to the sound speed and attenuation data results in a good fit to the higher frequency attenuation data, but underpredicts the attenuation observed at frequencies below 20 kHz. Future work, including full waveform inversion of the low frequency data and detailed error analysis, should reduce the uncertainty in the sound speed analysis at low frequencies.