Bayesian Inference For Predicting Potential Oil Spill Related Ecological Risk
Free (open access)
149 - 159
R. Aps, M. Fetissov, K. Herkül, J. Kotta, R. Leiger, Ü. Mander & Ü. Suursaar
The aim of this paper is to assess the potential oil spill related ecological risk for the southern Gulf of Finland coastal waters using the Bayesian Belief Network (BBN) methodology. The BBN prior probabilities were obtained from knowledge on spatial variability in the sensitivity of coastal ecosystem of the southern Gulf of Finland. The sensitivity data represented the three different ecosystem elements: the EU Habitat Directive Annex 1 habitats and associated habitat forming species, the EU Birds Directive Annex 1 birds and seals. Information on bird, seal and habitat layers were integrated into a single measure of ecosystem sensitivity. For this purpose the maximum value of different layers was calculated in each raster cell. The scenario modelling results showed that the western Gulf of Finland could be considered as an area of the highest ecological risk for the all seasons. Keywords: Bayesian inference, oil spill, ecological sensitivity, spatial, ecological risk assessment.
Bayesian inference, oil spill, ecological sensitivity, spatial, ecologicalrisk assessment