High-frequency bistatic scattering on seabed and targets: Comparison of scaled and fullscale experiments with sea trials

Recent uses of bistatic sonars show the advantages of decoupling transmitter and receiver(s) to optimise the information from seabed and target scattering. However, high-frequency scattering needs to be better understood, especially in complex, multiple-target environments (e.g. dumpsites or highly cluttered seabeds). Sea trials are paramount in providing acoustic measurements to validate scattering models and show the different processes involved, but they are expensive, difficult to conduct, and fraught with difficulties. Laboratory experiments are complementary, because of the fully controlled environment and the repeatability of the measurements. The imaging frequencies (> 10 kHz) to be investigated, and the finite dimensions of the tanks usually employed (decametre-scale usually), mean that these experiments most often need to be scaled. Laboratory experiments are ideal to understand the role of each physical process in the overall scattering, and to optimise data collection strategies depending on the objectives. Experiments can be scaled, using higher frequencies, smaller tanks and smaller targets. But how does it influence bistatic scattering (and its interpretation)? How does the transition to full-scale experiments work out? This is particularly relevant as sea trials are expensive and difficult to conduct. We compare here the results from: (1) scaled experiments of bistatic scattering on bare seabed and targets, performed at Bath; (2) full-scale experiments in the GESMA submarine pens during the EC-SITAR project, with targets in a sand box and (3) sea trials from similar bistatic experiments performed in Elba (for a bare seabed) and Möja Söderfjärd (on a well documented dumpsite). Each series of experiments revealed particular experimental issues, or solved specific questions related to the conduct of the experiments and/or the physical scattering processes. The three approaches prove to be complementary, with advantages and drawbacks related to their distinct objectives. The comparison of these experiments with acoustic simulations shows agreement increasing with the sophistication of the models. Tank experiments, scaled or not, can be used for the design of future surveys and instruments, as well as analyses of past and future acoustic datasets. Comparing their analyses with those of sea experiments show future trials can now be devoted to more focused investigations, or more complex generic problems.