Use of high-resolution roughness power spectra in predicting bottom backscattering

Abstract

Seafloor relief has been measured at very high (<1-cm) resolution with underwater stereo photogrammetry in the course of environmental characterization for high-frequency acoustic experiments. Information concerning bottom microrelief for the purpose of validating backscatter models has been heretofore expressed as rms height roughness. The validity of using Gaussian statistical methods to describe bottom relief is examined with the Hinich bispectrum test [Brockett et al., J. Acoust Soc. Am. 82, 1386–1394 (1987)] for normality in continuous data. Because bottom backscatter models such as the composite roughness model require the slope of the two-dimensional roughness power spectrum as an input parameter, generation of the two-dimensional relief spectrum is preferable to utilization of the rms relief or estimation from the one-dimensional spectra. Two-dimensional roughness spectra generated from photographs of rippled and smooth bottoms are compared to one-dimensional roughness spectra produced from the same photographs. Model and data comparisons are made for bottom backscattering using both roughness spectra.