Geoacoustic inversion for elastic bottoms: Matching model predictions and backscatter data from littoral limestone

The scattering strength of the ocean bottom as a function of angle and frequency is of fundamental importance in predicting the performance of active sonar systems, particularly in littoral waters. In this work, we apply an elastic-bottom scattering model [the small-slope model of Gragg et al., ‘‘Rough interface scattering,’’ J. Acoust. Soc. Am. 110 (2001)] to the problem of matching 2–3.5 kHz acoustic backscatter from rough limestone sea floors off the Carolina Coast and in the Straits of Sicily. We use a simplex/annealing algorithm to determine model parameters that optimally fit the data. Analysis of an ensemble of such inversion runs addresses the following questions. How well can the elastic theory match data measured at sea? How sensitive is the theoretical prediction to the values of its inputs-frequency and the (seven) parameters that characterize the geoacoustics of the bottom material and the roughness of the surface? Results to date indicate more sensitivity to roughness than to geoacoustic parameters. This supports the importance of estimating in situ bottom roughness and the feasibility of using a physics-based model for that purpose. [Work supported by ONR.]