A Bayesian-Loss Function Model for Assessing  Marine Liability Regime for Ship-Source Spills

Abstract

It is notoriously hard to model risk of ship-source oil spills in the Arctic waters for insurance purposes due to many unknowns and the lack of data. Despite such challenges, marine traffic and activities in the Arctic continue to expand, indicating the urgent needs for innovative ways to estimate losses from potential ship-source oil spill in this region. Thus, in this study, we present a hybrid-Bayesian-Loss function model to assess the marine liability for ship-source oil spills and apply it to the Baffin Island area, Nunavut, Canada. The model is a comprehensive template for assessing loss and subsequently the insurance for activities in the Arctic and sub-Arctic regions. Governmental and non-government organisations alike will benefit from the tool by using it as a loss estimation mechanism for liability for ship-source oil spills.

Keywords: Arctic, Bayesian theory, oil spills, Insurance, Shipping