Proxy-based characterization of physical properties in fine-grained shelf sediments using X-ray fluorescence

Seafloor sediment characterization is fundamental to understanding marine environmental processes, including sediment transport, depositional dynamics, and acoustic behavior. However, direct measurements of sediment physical properties such as bulk density and P-wave velocity are often limited by high costs, logistical constraints, and sparse spatial coverage. As a result, large-scale seafloor models frequently rely on simplified or incomplete representations of sediment properties. This study examines the application of sedimentary proxies, specifically elemental concentrations determined by X-ray fluorescence (XRF), to enhance predictions of key physical and acoustic sediment properties.

Sediment cores from the fine-grained New England Mud Patch (NEMP) provide a test case for evaluating proxy-based approaches in a depositional environment dominated by cohesive sediments. Multivariate regression models were developed using both wet and dry XRF measurements to assess how sample preparation influences predictive performance. Results indicate that dry XRF consistently produces stronger elemental signals and higher predictive accuracy than wet XRF, with coefficients of determination (R²) exceeding 0.7 to 0.9 for bulk density and P-wave velocity. These differences reflect the attenuation effects of pore water on elemental detection and highlight the importance of XRF methodology when integrating geochemical data into sediment property models.

Beyond their application to geoacoustic prediction, XRF-derived elemental proxies capture environmentally meaningful variations in sediment composition related to sediment source, grain size distribution, and depositional processes. Incorporating these proxies into statistical frameworks offers a rapid, non-destructive, and cost-effective means of enhancing spatially continuous sediment characterization. This approach supports improved parameterization of seafloor sedimentation models and provides a pathway for reducing uncertainty in marine geophysical analyses and seabed infrastructure assessment and design.