A non-deterministic model of populations of iceberg scour depths

Abstract

In this paper a non-deterministic model of iceberg scour depth distributions is developed. The model employs data on an initial distribution of scour depths, annual iceberg scour rates and annual sedimentation rates to generate populations of ‘observable’ iceberg scour depths, such as can be obtained from present-day deep-tow seismic data. Relationships between initial and observed populations are established. These theoretical relationships provide a basis from which a variety of inferences regarding present day iceberg scour rates, depth exceedence probabilities, etc. can be inferred from an existing data base of iceberg scour depths. In particular, we show how these inferences can in turn be used to generate design criteria for pipelines and glory-holes in seabeds subject to scour from icebergs or sea ice. As a test of the predictive capabilities of the model, iceberg scour data from three sites on the Grand Banks are analysed and an annual scour rate determined. This rate is compared to an annual scour rate derived from iceberg arrival rates and size exceedence data for the region. Very good agreement is found.