Segmentation of topographic profiles of the seafloor based on a self-affine model

Profiles of seafloor topography can be statistically described as a self-affine stochastic time series characterized by only a few parameters. Two characteristic parameters are used: a correlation parameter (the spectral exponent β, which is related to the fractal dimension D) and an amplitude parameter (an index of dispersion of the first differences S). This characterization provides a simple quantification of the intuitive notion of the roughness of topography. The quantification of roughness is important for the effective classification or areas of the seafloor where topography is dominated by different processes (e.g., sedimentation, volcanism, faulting). The procedure used can be subdivided into statistically homogeneous segments; statistical parameters are measured on each segment; and the segments are grouped into classes with similar statistical parameters