Statistics of seafloor backscatter measured with multibeam sonar systems

Abstract

A number of theoretical models for seafloor backscatter statistics developed for the recent years show a good agreement with experimental measurements made with sonar systems. However, methods of data collection used in multibeam systems are commonly not taken into consideration when analysing backscatter statistics. Using data collected with a Reson Seabat 8125 system and based on theoretical considerations, it is shown that the seafloor backscatter strength derived from the peak intensity measured as a single value for each beam leads to considerable backscatter overestimation at oblique angles of incidence when the beam footprint is much larger than the insonification area. This occurs because variations of the peak intensity are extreme value distributed, which can be well approximated by the Gumbel distribution. The location parameter of the Gumbel distribution depends on the ratio of the footprint and insonification areas, which results in distorted angular dependence of backscatter strength estimates. On the other hand, the average backscatter strength derived from the integral intensity, i.e. backscatter energy, is a consistent estimate of the actual seafloor backscatter strength. The Gamma distribution is demonstrated to be a good approximation for statistics of the integral intensity, even when the scattering regime is expected to be non-Rayleigh.