Characterization of seafloor by inversion of topographic roughness spectrum

This paper presents a method to quantify the percentage composition of seafloor sediment for classification using the power spectrum of the topographic roughness. Roughness profiles of the seafloor are computed from the echo signal of a echo sounder. It is shown here that the power spectrum shapes for different seabed material can be classified into three major groups (e.g. sand, silt and clay). It is shown that the basic roughness spectrum for these three categories can be computed easily from some available ground truth data on sediment type. Assuming the power spectrum for an unknown bathymetric profile is a linear function of these three basic spectra, the percentage composition (e.g., sand, silt and clay) at any position can be estimated. The estimated percentage composition of sand, silt and clay is compared with that of data obtained from the laboratory analysis of surface sediment analysis. The technique is capable to produce results with a fairly good accuracy.