Development of a Finite Difference Time Domain (FDTD) Model for Propagation of Transient Sounds in Very Shallow Water

Abstract

We developed a finite difference time domain (FDTD) model for sound propagation using pressure and velocity grids with both three-dimensional Cartesian and two-dimensional cylindrical implementations. Propagation parameters, including water and sediment properties, can vary in each dimension. The cylindrical implementation uses less computation but requires axial symmetry. The three-dimensional implementation allows anisotropic geometries. We can model both steady-state and transient sounds from discrete and distributed sources such as the surface of a vibrating pile. We compare our calculations to those made using a split-step parabolic equation. Applications of this model include calculating the propagation of individual fish sounds, fish aggregation sounds, and distributed sources in very shallow water.