This paper presents preliminary results of a recent study whose overall objectives are to determine the mechanisms contributing significantly to subcritical acoustic penetration into ocean sediments, and to quantify the results for use in sonar performance prediction for the detection of buried objects. In situ acoustic measurements were performed on a sandy bottom whose geoacoustical and geomorphological properties were also measured. A parametric array mounted on a tower moving on a rail was used to insonify hydrophones located above and below the sediment interface. Data covering grazing angles both above and below the nominal critical angle and in the frequency range 2-15 kHz were acquired and processed. The results are compared to two models that account for scattering of sound at the rough water-sediment interface into the sediment. Although all possible mechanisms for subcritical penetration are not modeled, the levels predicted by both models are consistent with the levels observed in the experimental data. For the specific seafloor and experimental conditions examined, the analysis suggests that for frequencies below 5-7 kHz sound penetration into the sediment at subcritical insonification is dominated by the evanescent field, while scattering due to surface roughness is the dominant mechanism at higher frequencies.

No abstract

Executive Summary:The environment significantly affects minehunting system performance. This is especially true in difficult minehunting conditions to detect low contrast mines in cluttered environments. Inexact performance predictions result in poor evaluations of mine clearance which in turn lead to inaccurate assessments of remaining risk to follow-on forces in the channel or area. In this context, the present report provides a quantification and understanding of environmental phenomena critical to acoustic backscattering at minehunting frequency. The following study demonstrates the unambiguous importance of seabed interface roughness, internal bioturbation and upper sediment structure for object detection. In some cases (e.g. in the presence of strong seabed inhomogeneities), minehunting sonar detection performances are believed to be strongly reduced. In order to remedy to these environmental limitation, it is clear that minehunting operations need to quantitatively account for the environmental conditions and adapt to them in near real time. The development of minehunting systems having an adaptive behavior as a function of the environment (i.e. in term of frequency sweep, sensing geometry, multi aspect coverage, etc..) is currently the object of ongoing research at SACLANTCEN. Recent backscattering measurements made in the Gulf of La Spezia (Italy) using a sonar operating at 140 kHz combined with thorough seabed interface and volume ground truth illustrate the dominance of seabed volume scattering. Threedimensional fluctuation statistics of density variability and vertical density gradients, both of which relate directly to the level of bioturbation (e.g. sea shell fragments, burrows, pockets of water) have been quantified using X-Ray computed tomography. Two-dimensional interface roughness spectra have also been determined using a digital stereo photogrammetry system. The combined ground truth has allowed a backscattering model to be fully constrained. Measured backscattering strength versus angle is compared to a model that includes the effects of varying density and sound speed. Data-model comparisons show that scattering from the volume of strongly inhomogeneous sediments can often be a primary contributor to seafloor scattering away from normal incidence. Ill Resume:De recentes mesures effectuees dans le Golfe de La Spezia (Italie) par un sonar operant a une frequence de 140 kHz combin^es a une mesure approfondie des proprietes d'interface et de volume ont permis d'illustrer la dominance de la diffusion de volume. Des mesures tri-dimensionnelles des fluctuations statistiques de la variation de la densite ainsi que des mesures de gradients verticaux (ces phenomenes etant tons deux lies k la presence de fragments de coquilles, de terriers et de poches d'eau) ont permis de quantifier de nombreux parametres de volume grace a une technique de tomographie utilisant des rayons X. La rugosite bi-dimensionnelle d'interface a ete determinee grace a un systeme ster^o-photogramm^trique numerique. La combinaison...

The Helmholtz-Kirchhoff integral and the Kirchhoff approximation are applied to model the penetration of sound waves into rough sandy seafloors at grazing angles above and below the critical angle. As the seafloor of interest is anisotropic, emphasis is placed on simulating the response from a two-dimensional interface. The analytical development of the method is first presented, followed by numerical examples. Simulations and data acquired at sea are in very good agreement in the 2-15 kHz band [Maguer et al., J. Acoust. Soc. Am. 107, 1215-1225 (2000)]. The model predicts, in agreement with the 2-15 kHz acoustic data, the contributions due to roughness effects that dominate the evanescent wave component over most of this frequency band. Secondary effects such as coherent (Bragg) influence patterns and the loss of signal coherence with grazing angle or depth are correctly predicted. The model simulations strongly suggest that roughness of the sediment interface is most likely the cause of anomalous sound penetration into the seabed.

A time-evolution model of seafloor scatter is numerically implemented and experimentally evaluated. The model is based on analytically expressing the elementary time-backscattered response of every seafloor surface and every seafloor volume infinitesimal element. The implementation of the model is based on a statistical realization of the seabed interface and volume inhomogeneities, from which the time series are computed by coherent summation of the scatter from small elements over the insonified area and volume. The analytical expressions and the implementation are evaluated for the image solution case, for which an almost perfect agreement is found. Examples are shown of how the beam width and seabed roughness affect the time-series return from both the surface and from the volume. The results of the model are compared with data from two different bottom types recorded with a parametric sonar. Reasonable accordance is found between the model and the data.