Backscatter vs. grazing angle, which can be extracted from multibeam backscatter data, depend on characteristics of the multibeam system and the angular responses of backscatter that are characteristic of different seafloor properties, such as sediment hardness and roughness. Changes in backscatter vs. grazing angle that are contributed by the multibeam system normally remain fixed over both space and time. Therefore, they can readily be determined and removed from backscatter data. The variation of backscatter vs. grazing angle due to the properties of sediments will vary from location to location, as sediment type changes. The sediment component of variability can be inferred using the redundant observations from different grazing angles in several small pieces of seafloor where the sediment property is uniform in any given piece of seafloor yet vary from one piece of the seafloor to another. Thanks to the multibeam survey (Roger Flood, State University of New York) at SAX 99 Project sponsored by Office of Naval Research (ONR), which had 800\% coverage in most of the survey area; there is a data set, which is suitable for investigating seafloor characterization. The investigation analyzed the spatial variation of the backscatter vs. grazing angle and compared that with ground truth sediment data. In this research, the 6.9 gigabytes raw multibeam data were cleaned using an automated outlier detection algorithm (Tianhang Hou, Lloyd Huff and Larry Mayer. 2001). Then, the surveyed area was equally divided into 52X78 rectangle working cells (4056), the side of each cell was about 20 meters. The backscatter vs. grazing angle of backscatter data for each cell is computed by averaging backscatter data by the corresponding beam numbers using all data with the same beam number from different survey lines. Systematic effects on the backscatter vs. grazing angle, caused by multibeam system hardware or software as well as system installation, were corrected in order to remove the asymmetric and skew effects. In order to easily evaluate the spatial variation of the backscatter vs. grazing angle, a graphic interface was developed. With a mouse click, the images based on different subsets of the data can be compared throughout the survey area. The subsets were created using specific beam numbers. These images for different beams show significant variations between nadir and off-nadir beams. These variations allow an interesting interpretation to be made of the images in light of seafloor characteristics, which were derived from ground truth data, such as sediment grain size, density and velocity.

Publication Date


Journal or Conference Title

EOS, Transactions American Geophysical Union, Fall Meeting, Supplement


82, Issue 47

Conference Date

Dec 10 - Dec 14, 2001

Publisher Place

San Francisco, CA, USA


American Geophysical Union Publications

Document Type