346 IEEE JOURNAL OF OCEANIC ENGINEERING, VOL. 27, NO. 3, JULY 2002 High-Frequency Subcritical Acoustic Penetration Into a Sandy Sediment Darrell R. Jackson, Kevin L. Williams, Eric I. Thorsos, and Steven G. Kargl Abstract—During the sediment acoustics experiment, SAX99, a hydrophone array was deployed in sandy sediment near Fort Walton Beach, Florida, in a water depth of 18 m. Acoustic methods were used to determine array element positions with an accuracy of about 0.5 cm, permitting coherent beamforming at frequencies in the range 11–50 kHz. Comparing data and simulations, it has been concluded that the primary cause of subcritical acoustic pen- etration was diffraction by sand ripples that were dominant at this site. These ripples had a wavelength of approximately 50 cm and rms relief of about 1 cm. The level and angular dependence of the sound field in the sediment agree within experimental uncertainties with predictions made using small-roughness perturbation theory. Index Terms—Acoustic scattering, buried object detection, mine countermeasures, sonar detection, sonar scattering. I. INTRODUCTION A S A PART of the sediment acoustics experiment in 1999 (SAX99), a hydrophone array was deployed in sandy sed- iment near Fort Walton Beach, Florida, in a water depth of 18 m. A moveable source tower transmitted short pulses from the water column to the buried array at center frequencies in the range 11–50 kHz. The primary objective was to study sound penetration into the sediment for incident grazing angles less than the critical angle (approximately 30 ), so called “subcrit- ical” penetration. Subcritical penetration is of practical importance in sonar detection and classification of buried targets such as mines, pipelines and archeological artifacts. Experiments by Chotiros and collaborators [1] and Lopes [2] have shown subcritical sound levels in the sediment that far exceed the levels that would be predicted assuming that the sediment is simply a flat, homogeneous, fluid half space. While it has been suggested that subcritical penetration may be due to refraction of a Biot slow wave [3], this mechanism is only possible if the Biot parameters of the medium depart considerably from those employed by most investigators [4]–[6]. Observations by Simpson and Houston [7] and Maguer et al. [8] of subcritical penetration have been interpreted as evidence for interface scattering as the primary cause. Using rough-interface fluid sediment models, Thorsos et al. [9] and Pouliquen et al. [10] have shown that published experimental data are consistent with this interpretation. Finally, Lim et al. [11] have conducted a laboratory experiment showing subcritical penetration with Manuscript received November 8, 2001. This work was supported by the U.S. Office of Naval Research under Code 321OA. The authors are with the Applied Physics Laboratory of the University of Washington, Seattle, WA 98195 USA (e-mail: drj@apl.washington.edu). Publisher Item Identifier S 0364-9059(02)06363-X. an artificial “rough” interface consisting of polystyrene beads floating on the boundary between two immiscible fluids. One of the primary goals of SAX99 was to conduct acoustic penetration measurements in conjunction with extensive envi- ronmental measurements in order to allow quantitative tests of hypothesized subcritical penetration mechanisms. Penetration measurements were made by groups from the Applied Physics Laboratory, University of Washington (APL-UW) and the Ap- plied Research Laboratories, University of Texas (ARL-UT). The present paper focuses on the APL-UW measurements and the interface scattering mechanism and compares measurements with scattering models based upon first-order roughness scat- tering theory. Another paper in this issue covers the ARL-UT measurements [12]. II. ENVIRONMENTAL DESCRIPTION A summary of SAX99 environmental measurements has been given by Richardson et al. [13] and this issue contains several papers giving details of these measurements [14]–[19]. For the work reported here, the environmental parameter of greatest interest is the ratio of sediment sound speed to water sound speed immediately above the seafloor, as this ratio determines the critical angle. Model predictions are also sensitive to parameters defining the water-sediment interface roughness. Of lesser importance are sediment mass density, sediment sound attenuation and water sound speed, as model predictions are more weakly dependent upon these parameters. Conductivity-temperature-depth measurements gave water sound speed values at the seafloor ranging from 1528 m/s to 1536 m/s. The representative value 1530 m/s is used in model calculations. Using core data [14], the sediment-water density ratio, , is taken to be 2.0. The dimensionless loss parameter, , is 0.01 [15]. From Richardson’s data [15], the sound speed ratio, , is taken to be 1.16, with a corresponding critical grazing angle of 30.5 . The sound speed ratio has a slight frequency dependence (1%) over the 11–50-kHz frequency band considered here [15], but this dependence is too weak to affect the conclusions of the present work. Changes in water temperature over the course of the experiment caused slight changes in sediment sound speed, but measurements of sound speed versus depth and time showed that such changes would have negligible effect on high-frequency acoustic properties [20]. Likewise, the spatial dependence in the sound speed and density ratios is about 0.7% over the SAX99 experiment site [14] and is small enough to be of no concern in this application. Interface roughness at the SAX99 site can be decomposed into isotropic roughness and an anisotropic ripple field. Both 0364-9059/02$17.00 © 2002 IEEE