Environmental Engineering Reference
In-Depth Information
Predicting the Weather-Dependent Collision
Risk for Birds at Wind Farms
Henrik Skov and Stefan Heinänen
Abstract
Collision risk for birds remains a potential conservation issue and
environmental barrier to the development of wind farms on land as well as at sea.
Baseline and post-construction studies in Denmark carried out at coastal and marine
wind farms during 2010-2012 have aimed at developing prediction tools which
could pave the way for improved planning and siting of wind farms in relation to
movements of birds. Detection of flight trajectories by means of visual observations
is severely constrained, and thus field campaigns were undertaken using a combination
of visual observations and radar- and rangefinder-based tracking. The collection of
two- and three-dimensional track data was necessary to obtain useful information
on the responses of migrating bird species to the wind farms, and on flight altitudes of
the birds during different weather conditions and in relation to landscape components.
To be able to assess general patterns in the migration behaviour of birds, we developed
statistical models capable of explaining the differences in altitude based on relation-
ships with wind and weather conditions and distance to coast. As these relationships
in many cases were non-linear, the error structure of the data non-normally distributed,
and the track data spatially and temporally auto-correlated we chose to use a gener-
alized additive mixed modelling (GAMM) framework. The resulting models of the
migration altitude of raptors and other groups of landbirds made it possible to assess
the weather-dependent flight altitude at the wind farm sites. The studies provided
strong indications that wind speed and direction as well as humidity, air clarity and
air pressure are important predictors in general for all species in addition to distance
to land and wind farm, and the birds favour tail winds and decreasing wind speed.
Collision models display a variety of specific trends with rates of collisions of
landbirds increasing during periods of head winds and reduced visibility, while the
collision rates of seabirds typically increase during periods of tail winds and
increased visibility. Our studies have shown that birds across a wide range of species
show clear weather-dependent movements which can be predicted for specific spatial
settings using statistical models. These findings stress the potential for intensifying
the strategic planning processes related to wind farms.
