Variance-based global sensitivity analysis of a stochastic Collision Risk Model: how wind turbine parameters impact seabirds

To achieve the United Nation's goal of net-zero greenhouse gas emissions by 2050, we must create a world powered by renewable energy. This will rely on large-scale infrastructure projects, which must be developed with consideration for the natural environment. Offshore wind farms (OWFs), in particular, must balance the collision risk that turbines pose to seabirds with green energy targets. Within the UK and most of Europe, collision risk modelling (CRM) is a required component of Environmental Impact Assessments (EIAs) and integral for developers and regulators to gauge the risk an OWF development poses to seabird populations. Despite this, how different combinations of OWF parameters may impact collision risk estimates is poorly quantified. We assess the sensitivity of the stochastic CRM to seven key OWF parameters. Using variance-based global sensitivity analysis methodologies, we quantify results for six seabird species deemed at high risk of turbine collision. We find that the estimate of annual bird-turbine collisions is most sensitive to variation in the number of turbines and the distance between blade tip height and highest astronomical tide (air gap), whereby increasing turbine number and decreasing air gap contributed to increased collision rates. A smaller number of large turbines with greater energy output (larger rotor radius) reduced collision risk per unit energy output. These results quantify the impact of OWF parameter variation on seabird collision risk, which could be used to inform design envelopes, focusing the number of turbine scenarios that developers consider within EIAs and maximising cost-effectiveness whilst minimising seabird collision risk.