Sonar echo statistics as a remote-sensing tool: Volume and seafloor

Sonar echoes from unresolved features tend to interfere with each other. When a sonar is moved through an area where features of interest are not resolved, echoes will fluctuate from ping to ping due to the interferences. These fluctuations are the subject of this paper. We classify both fish and the rough seafloor by analyzing the nature of echo fluctuations. Similar statistical techniques apply to all of the problems. We use the techniques to determine, from echo fluctuations, parameters that describe the physical situation. Fish: Two conditions are low density and high density of fish (density is defined as number of fish per unit volume), a) In the low-density case, fish are resolved and their individual echoes are distinguishable from each other. From single fish there are interferences between echoes from the anatomical features causing fluctuations. Echoes from resolved fish also fluctuate because of beampattern effects. After deconvolution (or "removal") of beam-pattern effects, echo amplitudes from resolved fish fit a Rician probability density function (pdf). b) In the high-density case, the fish are not resolved and echoes from individual fish overlap. These echoes interfere and cause the echoes to fluctuate. The Rayleigh pdf describes the instantaneous echo amplitude from a cloud or school of fish. We process echoes by picking the peak amplitude in a finite-time gate and obtain an extremal pdf. We examine both regions of density (a) and (b) and present methods we have developed to determine, from the nature of the fluctuations, properties such as fish size frequency distribution and possibly gross anatomical features, fish density, and to count occasional large fish swimming in a cloud of plankton or near the sea surface. Seafloor: Interference of echoes from individual protuberances such as rocks, nodules, and ripples causes fluctuations. The Rice pdf describes echo amplitudes from the seafloor. We combine Eckart acoustic-scattering theory for a downward-looking sonar and Rice statistics. As a result we describe echo fluctuations in terms of rms roughness of the seafloor and correlation area (product of
x
and
y
correlation distances) of the seafloor. We use boomer data to estimate seafloor microrelief as well as to predict echo fluctuations for a variety of other sonars.