Preliminary hard and soft seafloor substrate map derived from an unsupervised classification of multibeam backscatter and bathymety derivatives at Tutuila Island, American Samoa, South Pacific. The dataset was derived using a combination of Simrad em3002d and Reson 8101 backscatter, bathymetric variance and bathymetric rugosity. The sonar frequencies are 300 kHz and 240 kHz for the em3002d and the 8101 data respectively and all data were resampled to 5 m grid cell size prior to the classification. Initial supervised classifications of the backscatter data into hard and soft seafloor substrates, using seafloor photographs for groundtruthing and to define regions of interest, were used to define unsupervised class types and to visually evaluate the accuracy of the unsupervised classification seafloor substrate map.