ECOS 38(6)

The data and models used to claim no effects from offshore wind farms on birds are unsuitable.

Credible data?

With the rise of evidence-based conservation management (, it is unsurprising that the conservation sector should be expecting a valid science base for the material provided in discussions of the potential impacts of development on large seabird populations.

The call for credible data to examine the pros and cons of wind turbine proposals seems reasonable. Recent work by Matthews et al (2016)[1] on the data used to assess the potential impacts of onshore turbines on bats showed that much of the evidence base used in support of developments was not actually very robust[2]

In Scotland, unlike England, the expansion of turbines has continued onshore and offshore. The latter has seen very large proposals off the east coast of Scotland for areas used by significant populations of seabirds. The RSPB queried the selection of these areas for wind turbine developments, and fought the case against them in the Scottish High Court, where it won, and then at Appeal, only to lose the case in June 2017 on a procedural technicality. 

The nub of the RSPB’s case was that the proposed turbine facilities would make significant impacts on the seabirds of the area, and that the quality of the material used in supporting windfarm applications was unsuitable. In a paper[3] and report[4] led by Rhys Green of Cambridge University, the spatial planning approach and data requirements used in offshore wind turbine spatial planning in Germany and Denmark and other countries were compared unfavourably with those in the UK.

Green et al looked at the methods and data used in planning cases offshore in the UK, and found them wanting. Green et al put it very simply: “sound science and its logical interpretation are lacking in EIAs of this large and expanding industry”.

Playing down the collision rates

Green et al started with the building blocks of the EIAs: the collision risk models (CRMs) and the data they use. The CRM model estimates the numbers of birds, potentially flying though the areas swept by turbines, that would be expected to collide with the turbine blades. For that reason, it is important that the data entered into the model are robust. If not, then there are major problems of reliability. Green et al showed that the estimates of bird heights and numbers in height bands are subject to a range of uncertainties and variabilities.[5] Other variables in the CRM model are also poorly checked for accuracy. In addition, simple data such as avoidance rates (where birds change direction to avoid the area) by birds of turbines at wind turbin developments are usually based on (error-prone) extrapolations from other locations, or rely on poorly explained expert judgement. After listing a further range of problems, such as the gaps in empirical data for rates of avoidance of turbines, Green et al concluded that collision rates were not tenable. As importantly, where offshore windfarms have been built, there are no data to show their actual effect: good or bad. The result was that there was no valid basis for understanding the industry’s claimed minimal effects of wind turbine developments on seabirds.

Flaws in threshold models

As all of the statements by developers about rates of kill by offshore fields of turbines being unimportant rely on several basic assumptions and models about population impacts, Green et al looked at the mathematics of the models for change and population declines. Like the CRM, once unpacked in some detail, they, too, were found wanting. Of particular concern was the use of threshold models: levels below which there would be minimal impact. Green found they were without a logical or empirical basis.

Rather than do a completely negative hatchet job, Green et al offered a model that was suitable to use the less than robust data available for most locations, but with caveats clearly stated. They recommended the use of a Leslie Matrix Model (LMM). The LMM lacks a built-in arbitrary threshold below which it might be argued (unlike the other rejected models) that impacts would be unimportant.

Green et al were, for Cambridge scientists, remarkably direct about the poor modelling approaches and the use of arbitrary unsupportable thresholds to claim that there were levels below which seabird sites would suffer no significant effect. As they put it, the emperor had no clothes: developers were using models not designed for the purpose, with inadequate data and scientifically unsupportable thresholds. This was quite a statement in a major peer-reviewed journal.

Robust data for a better model

What does all of this mean in practical terms? Almost all of the areas of dispute off the east coast of Scotland are locations used by internationally important seabird colonies, and cited as Natura 2000 Sites under the EU birds and habitats Directive. Populations in these sites are meant to be preserved at favourable conservation status[6]. Damaging populations will compromise that status. Whilst that might be argued for or against, Green et al conclude that poor science and data should not be used to claim that damage can be accepted.

Just like bats, proper data and models are needed to promote an informed debate as to the effects on wind farms on seabirds. What we have here is a viable model. Now we just need suitable data.


[1]. Mathews, F., Richardson, S., Lintott, P., & Hosken, D. (2016) Understanding the Risk to European Protected Species (bats) at Onshore Wind Turbine Sites to inform Risk Management. University of Exeter.

[2]. Reed, T.M. (2017) Mixed messages about bats and turbines British Wildlife 28: 217-18

[3]. Green, R.E., Langston, R. W., McCluskie, A., Sutherland, R. & Wilson, J.D. (2016) Lack of sound science in assessing wind farm impacts on seabirds. Journal of Applied Ecology 53, 1635–1641 doi: 10.1111/1365-2664.12731

[4]. Green, R. E (2016) Misleading use of science in the assessment of probable effects of offshore wind projects on populations of seabirds in Scotland.  Sandy, RSPB.

[5]. Johnston, A., Cook, A.S.C.P., Wright, L.J., Humphreys, E.M. & Burton, N.H.K. (2014). Modelling flight heights of marine birds to more accurately assess collision risk with offshore wind turbines. Journal of Applied Ecology 51, 31–41.

[6]. EU 2013. Court of Justice of the European Union. Decision in Peter Sweetman, Ireland, Attorney General, Minister for the Environment, Heritage and the Local Government v An Bord Pleanala [2014] PTSR 1092,[2013] EUECJ C-258/11.



Director of Tim Reed Ecological Consultants.

Contact the author

Leave a Reply