The environmental impacts of oil shipping and offshore wind power at the eastern Gulf of Finland : A Bayesian approach to marine spatial planning

The seas and oceans are the scene of multiple human actions, all of which cause pressures on the marine environment. Marine spatial planning (MSP) systematizes the evaluations of the spatial impacts of the human actions and take into consideration the cumulative impacts of the actions. A probabilistic model is constructed to estimate the impacts of oil shipping and offshore wind power on 16 species. The quantitative indicators of impacts are the loss of breeding success of 5 birds, the loss of the early development stages of 3 fish species and the change in the probability of presence/absence of 3 benthic species and 5 algae. The thesis model works as an independent application, but can be merged as such into an MSP tool that works with a geoinformatic system (GIS) interface. The impacts of offshore wind power and oil shipping, and especially the possible oil spill, have been studied at other marine areas, but there are only few studies about their impacts in the brackish water conditions of the Baltic Sea. The study area of this thesis is the eastern Gulf of Finland (EGOF). The model is constructed using Bayesian networks (BNs) which are graphical probabilistic models. The most important human pressures caused by the actions are identified based on literature and placed in the model accordingly. The pressures caused by operational offshore wind power are the disturbance to birds and underwater noise. The pressures caused by oil shipping are underwater noise and the oil exposure of species after a possible oil spill. The attenuation of the pressures as a function of increasing distance from the source of pressure is calculated mathematically, where possible. Expert elicitation is conducted to fill in the gaps in existing data over the subject. Altogether 6 experts were interviewed and another two were consulted informally. The different types of data are integrated in the BN, which allows quantified comparisons between different management options and alternative scenarios. The model predicts that both human actions have negative impacts on the marine environment of the EGOF. The impacts of an offshore wind mill will realize without uncertainty but they will be negligible. An oil spill, on the other hand, is unlikely to happen, but if it does, the losses will be extensive. The disturbance of the wind mill on birds extends to some hundreds of metres from the mill, depending on the bird species. The losses of the early development stages of fish caused by the underwater noise of a wind mill are nearly certainly below 20% at all distances from the mill for all studied species. With the most likely sound pressure levels of tankers, the losses to the early development stages of the fish also remain below 20% with a high level of certainty at all distances. At these tanker noise levels, the harmless noise class of <90 dB re 1µPa will be reached at some kilometres of the fairway, depending on the original noise level from a tanker. Three alternative oil shipping scenarios for 2020 were compared. The differences among the scenarios are negligible both when it comes to the impacts of underwater noise on fish and to the probability of a species to get exposed to oil. The model successfully describes the impacts of the human pressures that are known to take place, such as the impacts of offshore wind power, but requires a GIS environment and drift models to be able to predict the probabilities of an oil exposure. The applicability of the model can be increased by taking into consideration additional human actions and a wider selection of human pressures. The thesis model is a part of a MSP tool produced in TOPCONS (Transboundary tools for the spatial planning and conservation of the Gulf of Finland) project, which is a prototype of a tool that can be later applied at marine areas worldwide.