High-Frequency Sound Interaction in Ocean Sediments
2 July 1999

Laboratory and Field Experiments for a Baseline Model
High Frequency Sediment Acoustics

Robert D. Stoll
Lamont-Doherty Earth Observatory of Columbia University
Palisades, New York 10964
(914) 365-8392, -8156 (fax), stoll@ldeo.columbia.edu

Return to Contents Page


The goal of our experiments is to provide ancillary information about the dispersion that is to be expected during seismic and acoustic propagation in sand sediments. We are performing both laboratory and field experiments designed to measure velocity and attenuation of p- and s-waves at frequencies well below those being used in the main higher frequency studies in order to make direct comparisons to see if this dispersion can be documented. Moreover, the experiments are designed to replicate environmental conditions in the first few meters immediately beneath the seafloor where bottom-penetrating, high-frequency sonar is particularly important. One of our main objectives is to establish a baseline model, valid over many decades of frequency, that may be used for studying the perturbing effects of various inhomogeneities such as gas bubbles, surface roughness, and various kinds of buried objects. The lower frequency data are necessary to constrain this kind of baseline model as was recently pointed out by Stoll and Bautista (1998).

Laboratory Experiments

Preliminary laboratory experiments are described by Stoll (1998). These made use of cylindrical, water-saturated specimens which were subjected to vertical, steady state, fluid flow in order to reduce the vertical effective stress to negligible values before acoustic measurements were made. High frequency measurements were made across the diameter of the specimen, and lower frequency measurements were made in the axial direction using a resonance technique wherein the bottom end of the specimen was driven by a vibrating piston and the motion of the upper surface was detected using a sensitive, non-contacting displacement sensor. Results of these experiments were encouraging and they suggested that there was a measurable decrease in p-wave velocity in going from the high to low frequency regime as predicted by the Biot theory. However because of radial boundary conditions imposed by the tube containing the specimen and the resulting tube waves that were excited, it was difficult to interpret the resulting motion of the specimen with complete certainty. For this reason we have fabricated a new cylindrical shell with a thick wall to confine the specimen and eliminated all appendages that tend to cause spurious resonances during the low frequency experiments. In addition the driver piston has been modified to decouple it from the shell and to allow direct measurement of the movement of both the top and bottom of the specimen. Hence the phase difference between top and bottom is known at all frequencies. This greatly facilitates the interpretation of our ongoing experiments. Tests on several different sand gradations are planned.

Field Experiments

The field experiments that we plan for the SAX99 program involve the deployment of a linear array of sensors (gimballed, vertical geophones) and an impulsive source located on the seafloor. Because of the increase of p-wave speed with depth caused by increasing overburden pressure, shallow "diving" waves produce the first motion observed at the geo-phones, and the travel-time vs. distance curve may be inverted using the classical HBW integral equation to obtain p-wave velocity versus depth in the sediment very near the bottom. In these experiments it is important to choose the proper geophone spacing and a source that will excite a narrow-band pulse with a frequency content in the range of interest. Over the past 13 years, we have developed several bottom-deployed sources that use a blank cartridge to produce a gas bubble focused downward into the sea floor by a helmet-shaped chamber resting in the bottom (Stoll, 1991; Stoll et al., 1994). This type of source generates strong diving and interface waves in the appropriate frequency range. In our prior work with the equipment described above, we have concentrated on the generation of interface waves (Scholte waves) in order to invert for shear-wave velocity. For the SAX99 experiments we have changed the sampling rate and geophone positions to obtain an accurate p-wave velocity from the travel-time curves of the diving waves that are generated.

We have examined data from our 1993 experiments in the Gulf of Mexico during the CBBL experiments and have decided to use our 11 shot source that employs 22 caliber blank cartridges based on the excellent records that were obtained with this source. We were able to obtain reasonable velocity measurements from the 1993 data even though the sampling rate was too low and phone spacing was not ideal. For the SAX99 experiments, a geophone spacing of 5 m is planned, with the first phone located approximately 1 m from the source. A sampling rate of 5 kHz will be used in order to obtain a spatial resolution of a little over 30 cm assuming a typical velocity of 1600 m/s. The source and shot phone will be mounted on a self-righting sled and the geophone cable with 12 gimballed phones attached so that dragging a short distance produces a straight line (see Stoll et al., 1994). Divers are needed only to check for any tangled phones and for correct alignment. The source and array have been modified in a number of ways and a preliminary field trial with the altered gear was carried out successfully on March 26 in New York harbor in both sand and sandy mud areas.

In the SAX99 experiments, we would like to deploy our array at the same time and in the same location as the ISSAMS deployment that will be made by Mike Richardson in order to get a direct comparison with his higher frequency p- and s-wave measurements. We estimate each deployment of our array to require two to three hours of ship time, so that our work can be completed in a day's time assuming two deployments and no interruptions.

Other Work in Progress

In addition to working on our laboratory and field experiments, we are using the modified version of the OASES computer code, which now contains the Biot model as an option, to study the effects of vertical inhomogeneity on shallow grazing angle response. We also plan to use this code to generate synthetic seismograms to be compared with the signals recorded in our field experiments.


Stoll, R. D. (1991) "Using seafloor arrays to measure sediment geoacoustic and geotechnical properties," Proc. IEEE Conf. "Oceans 91," pp. 97–101.

Stoll, R. D. (1998) "Measuring parameters that control acoustic propagation in granular sediments near the seafloor," Proc. 135th meeting Acoust. Soc. Am., Seattle.

Stoll, R. D. and Bautista, E. O.(1998) "Using the Biot theory to establish a baseline geoacoustic model for seafloor sediments," Continental Shelf Research, 18, pp 1839–1857.

Stoll, R. D., Bautista, E. O. and Flood, R. (1994) "New tools for studying seafloor geotechnical and geoacoustic properties," J. Acoust. Soc. Am., 96, pp. 2937–2944.

| Top of Page | Return to Contents Page |