Predicting Seafloor Facies from Multibeam Bathymetry and Backscatter Data

An empirical technique has been developed that is used to predict seafloor facies from multibeam bathymetry and acoustic backscatter data collected in central Santa Monica Bay, California. A supervised classification used backscatter and sediment data to classify the area into zones of rock, gravelly-muddy sand, muddy sand, and mud. The derivative facies map was used to develop rules on a more sophisticated hierarchical decision-tree classification. The classification used four images, the acoustic-backscatter image, together with three variance images derived from the bathymetry and backscatter data. The classification predicted the distribution of seafloor facies of rock, gravelly-muddy sand, muddy sand, and mud. An accuracy assessment based on sediment samples shows the predicted seafloor facies map is 72 percent accurate.