The effects of fine-scale surface roughness and grain size on 300 kHz multibeam backscatter intensity in sandy marine sedimentary environments

High-frequency acoustic backscatter intensity that is returned from sediment is affected by a number of seabed characteristics including sediment grain size distribution, the nature and magnitude of seabed surface roughness, and volume scattering by subsurface scatterers. In order to develop quantitative remote sediment classification techniques, it is important to determine the relative and combined effects of the dominant variables to which the acoustic signal is sensitive. Data used in this study were collected at two sandy sites off the south shore of Long Island, New York, and include 300 kHz multibeam (Simrad EM 3000) backscatter intensity data as well as grab samples and stereo bottom photographs. Multivariate statistical analyses were used to investigate the relationships between grain size properties, seabed roughness, and backscatter intensity at a number of sample locations at each study site encompassing a moderate range of backscatter intensities from − 32.5 to − 24 dB. The results of this study demonstrate that the relative importance of variables that affect backscatter intensity varies at different sites, suggesting that backscatter intensity alone cannot be used to quantitatively predict seabed characteristics. However, our study indicates that in sandy sediments, the median grain size, the standardized magnitude of seabed roughness, and the sorting of sediment can together be used to predict 300 kHz backscatter intensity with high significance, and that the relative importance of these variables varies from site to site. We present an empirical model derived from a combined dataset that includes two sites of similar sedimentary and hydrodynamic conditions to predict the magnitude of seabed roughness based on backscatter intensity and sediment grain size characteristics. The spatially coherent patterns in predicted seabed surface roughness revealed through the use of this kind of model greatly improve our understanding of variations in backscatter intensity in sandy marine sedimentary environments. As a result, the multivariate statistical approach that we employ improves our overall ability to remotely assess the details of the sedimentary environment that are significant with respect to benthic habitats and physical processes, and are applicable to other geophysical datasets including sidescan sonar data.