A Probabilistic Approach for Spatial Scour Burial Prediction

Scour burial prediction of objects located or placed on the seabed (e.g., sensors, unexploded ordnances) is affected by the inherent variability of hydrodynamics and by limited data availability of sediment parameters, which reduces the reliability of deterministic methods. This particularly applies if the analysis is carried out over a larger area or at unexploded ordnance sites. This research aims to explore a probabilistic analysis framework for scour burial prediction of cylindrical objects with case studies in two distinct coastal environments: Sequim Bay, a coastal embayment in the Pacific Northwest, Washington, and the nearshore zone in Duck, along the Atlantic Outer Banks, North Carolina, both in the United States coastal waters. Field data, including significant wave height, peak wave period, depth-averaged current velocity, median grain size, water depth, and the angle between waves and currents, were collected from existing literature and opportunistic data collection efforts. A scour model, developed for the risk assessment of unexploded ordnances in subaquatic environments, was employed in combination with Monte Carlo simulations to generate normalized equilibrium scour burial depths for a cylindrical object. Scour burial statistics were then mapped across spatially interpolated sediment data. The inlet region of Sequim Bay and the shoaling region of the nearshore zone in Duck, were found to exhibit a high risk of scour, while reduced risk was observed in the central part of the bay and farther offshore in the nearshore zone of Duck. Sensitivity analyses indicated that the most critical parameters to the mean scour burial prediction in Sequim Bay are depth-averaged current velocity and median grain size. At Sequim inlet, higher scour burial risk is observed when current velocities exceed 0.9 m/s and particle sizes are within the sand range but smaller than 0.4 mm. In Duck, significant wave height was found to be the most influential parameter, followed by water depth and depth-averaged current velocity, with scour burial risk exceeding 50% when wave heights exceeded 2 m, water depths were less than 2 m, and current velocities were more than 0.9 m/s. The approach introduced here offers a pathway to address data limitations associated with sporadic point measurements of sediment properties and quantify uncertainty arising from the inherently variable nature of environmental conditions for scour burial prediction of cylindrical objects. It contributes by transforming an established deterministic, single-location scour prediction model into a probabilistic, area-covering approach.