No abstract

No abstract

This paper investigates the applicability of Bayesian inference to oil spill related situation assessment in order to facilitate the Net Environmental Benefit Analysis (NEBA) based decisions in evaluating the threat or probable overall environmental impact of the spill. Bayesian networks are believed to be useful in integrating the NEBA related information imported from 1) oil spill scene surveillance, 2) simulation results on an oil spill incident with human response, and 3) ecological sensitivity maps. This paper exemplifies the use of Bayesian Belief Networks in answering the questions: can the oil spill be combated at sea, and if it cannot then is the oil threatening a sensitive environment?

This paper explores the problem of allocation of potential places of refuge for a ship in distress along the Estonian coast of the Gulf of Finland balancing the advantage for the affected ship and for the environment resulting from bringing the ship into a place of refuge. The integrated oil accident response simulation environment (PISCES II, ArcGis spatial modeling tools and Bayesian Belief Networks) proved to be instrumental for operational decision support in the case of a hypothetical oil accident. The knowledge of the ecological sensitivity of the Estonian coastal sea in the Gulf of Finland is used to estimate the probability of expected ecological damage associated with different towing directions of the vessel in distress. It is shown that the choice of the most favorable towing direction of the vessel in distress depends on the season, the position of the accident, and the ecological sensitivity level of the coastal sea area concerned.