Recent concern about the adverse effects of collision mortality of avian migrants at wind farms has highlighted the need to understand bird-wind turbine interactions. Here, a stochastic collision model, based on data of seabird behaviour collected on-site, is presented, as a flexible and easy to take tool to assess the collisions probabilities of off-shore wind farms in a per-constructive phase.
The collision prediction model considering the wind farm area as a risk window has been constructed as a stochastic model for avian migrants, based on Monte Carlo simulation. The model calculates the probable number of birds collided per time unit. Migration volume, wind farm dimensions, vertical and horizontal distribution of the migratory passage, flight direction and avoidance rates, between other variables, are taken into account in different steps of the model as the input variables. In order to assess the weighted importance of these factors on collision probability predictions, collision probabilities obtained from the set of scenario resulting from the different combinations of the input variables were modelled by using Generalised Additive Models.
The application of this model to a hypothetical project for erecting a wind farm at the Strait of Gibraltar showed that collision probability, and consequently mortality rates, strongly depend on the values of the avoidance rates taken into account, and the distribution of birds into the different altitude layers. These parameters should be considered as priorities to be addressed in post-construction studies.