Fractal analysis of muddy submarine channel slope instability from sub-bottom profile images

Abstract

Submarine channel slope stability is a key factor in marine traffic safety. Channel slope instability and collapse may change the geometry of a channel, block ports, and lead to marine traffic accidents. In this study, fractal and multifractal analyses are performed on sub-bottom profile images obtained before, during and after episodes of slope instability and collapse in sediments obtained from a muddy submarine channel. Results show that sub-bottom profiles exhibit distinctive fractal properties, and that changes in sub-bottom profiles are reflected by changes in fractal dimensions. For example, fractal dimensions are highest during slope collapse. During episodes of slope instability and collapse, fractal dimensions of temporal domain sub-bottom profile images exhibit distinct variations at different depths along the slope. Meanwhile, the fractal dimensions of vertical spatial domain sub-bottom profile images reveal that the slope experiences multiple fluctuating changes during collapse. Multifractal spectrum parameters highlight image non-uniformity. For example, spectrum width temporarily increases during slope instability and collapse. Results of this study suggest that the fractal dimensions and multifractal parameters of sub-bottom profiles are sensitive to channel slope instability and collapse, thereby providing useful indicators with high potential for use in slope collapse prediction and early warning systems.