
High-Resolution Seafloor Grain Size Classification and Mapping Using Multibeam Backscatter Data and Geostatistical Interpolation Methods
Giladi Asaf (1), Kanari Mor (1), Katz Timor (1) & Tibor Gideon (1)
(1) Israel Oceanographic & Limnological Research Ltd., Tel-Shikmona, P.O.Box 8030, Haifa 31080, Israel
Grain size distribution along continental shelves is essential for understanding benthic habitats, supporting environmental management, and guiding coastal infrastructure planning. This study presents the results from the first phase of a project analyzing grain size distribution on a continental shelf using integrated acoustic and geostatistical methods. Backscatter data were collected during annual monitoring programs (2017–2023) using multibeam sonar aboard the R/V Bat-Galim. The data underwent angular response corrections (ARA) and were validated against field-collected sediment samples to ensure accuracy. Normalization and validation processing focused on flat, homogeneous seafloors. Optimal correction values, achieved with muddy sand (Phi 3.3), produced highly accurate classifications for grain sizes ranging from medium silt (Phi 5.5) to medium sand (Phi 1.5), with an error margin of less than 1 Phi value across water depths from 10 m to 100 m. Seafloor classification revealed clear transitions in grain size, including shifts from fine to very fine sand and from coarse to medium silt at specific depth ranges. Hard substrates, such as platforms and ridges, were identified using both bathymetry and backscatter data, with strong acoustic returns offering distinct signatures. Areas with hard substrates surrounded by softer sediments were excluded from interpolation due to their artifact effect on surrounding values. In regions where hard substrates were widely distributed, their values were included in the interpolation process. The resolution and accuracy of the interpolated maps were closely linked to data density and spatial distribution. Kriging, leveraging spatial autocorrelation, produced the most reliable results, balancing data density and resolution at 250 m grid scale maps. The results highlight the value of combining normalized backscatter with geostatistical tools for cost-effective, precise seabed classification, offering valuable insights into sediment dynamics and supporting broader applications in marine spatial planning and environmental monitoring.