Optimal Sand Nourishment Decisions

Abstract

To maintain the Dutch coastline, every year millions of cubic meters of sand must be supplied at locations subject to permanent erosion. A decision model has been developed to obtain optimal sand nourishment decisions whose expected costs are minimal on the basis of the only information that is available: the probability distribution of the limiting average rate of permanent erosion. This probability distribution is derived on the basis of actual erosion data using Bayes' theorem. In order that the stochastic erosion process be based on this uncertain limiting average, the writers consider it as a generalized gamma process. There are three cost-based criteria for comparing sand nourishment decisions: the average costs per unit time, the discounted costs over an unbounded horizon, and the equivalent average costs per unit time. From these three criteria, only the last two are appropriate to obtain optimal sand nourishment decisions. In a case study, the decision model has been tested to sand nourishment at...