R. A. Stephen
S. A. Swift
Woods Hole Oceanogr. Inst., Woods Hole, MA 02543
A major problem in understanding sound propagation in the seafloor is to distinguish between the loss of energy due to intrinsic attenuation (anelasticity) and the loss of energy due to scattering from intermediate scale heterogeneities and bottom roughness. Energy lost to intrinsic attenuation (heat) disappears entirely from the system. Energy lost to scattering is conserved in the system and can effect observations as incoherent noise (time spread, angle spread) and/or mode converted waves. Dougherty and Stephen (1988, 1991) showed that the finite difference synthetic seismogram method can be applied to the seafloor scattering problem. Recently, this two-dimensional finite difference code has been extended to include intrinsic attenuation using the formulation of Day and Minster (1984). The formulation assumes a stress--strain relation for which Q is independent of frequency over a specified bandwidth. For each node on a two-dimensional grid arbitrary values of compressional and shear Q can be specified in addition to the usual values of compressional and shear velocity and density. Fluids are represented by setting the shear modulus to zero. Q's computed from the output time series are in good agreement with the input values for homogeneous media. Now by forward modeling the trade-off between scattering and intrinsic attenuation in seafloor models with both surface roughness and volume heterogeneities can be studied. Results can be compared with field observations from vertical seismic profiles (VSP's) in the seafloor and marine refractions experiments.