Western US model estimates rising golden-eagle deaths at wind turbines with wide uncertainty

Raptor deaths are not counted with a pile of eagle carcasses beneath turbines.

It is the middle of a probability distribution.

Researchers modeling wind-turbine collisions across the western United States estimated a median of 110 golden-eagle deaths in 2013.

The Bayesian collision-risk model combined that turbine volume with eagle exposure and the probability that an eagle entering hazardous space would actually collide.

By 2024, that median had risen to 270.

But the model’s 80% credible interval for 2024 stretched from 72 to 877 deaths.

The increase is clear inside the model.

The exact toll is not.

The model starts with where eagles fly

The researchers covered the contiguous United States west of the 100th meridian, the core of the golden eagle’s American range.

They first used eBird relative-abundance estimates to divide that enormous landscape into lower- and higher-risk zones.

The logic is physical.

More eagles flying through an area means more opportunities for one to enter the three-dimensional space occupied by turbine rotors.

Pre-construction eagle surveys from 44 wind developments then supplied exposure data: eight projects in lower-risk areas and 36 in higher-risk areas.

Measured exposure in the higher-risk zone averaged more than 11 times the lower-risk value.

Turbine count misses the bigger change

There were 25,704 operational onshore turbines in the study region in 2013.

By November 2024 there were 32,579, an increase of 26%.

The rotors changed much faster.

Average reported rotor diameter grew from 0.041 miles (about 217 feet), to nearly 0.059 miles (311 feet), while the study’s estimated total turbine “hazardous volume” rose from 4.17 to 11.28 cubic miles.

That is a 171% increase.

This is why simply counting towers misses part of the collision problem.

A larger rotor sweeps a larger volume of the sky.

In the higher-risk zone, turbine numbers actually fell 14% between 2013 and 2024, yet hazardous volume increased 119% as newer and repowered machines grew larger. 

 

The median more than doubled

The Bayesian collision-risk model combined that turbine volume with eagle exposure and the probability that an eagle entering hazardous space would actually collide.

For 2013, it produced a median estimate of 110 deaths, with an 80% credible interval from 28 to 374.

For 2024, the median reached 270.

The corresponding interval ran from 72 to 877.

That range is not a footnote.

A median of 270 does not mean researchers established that exactly 270 golden eagles died in 2024.

It means half of the model’s posterior probability lies on either side of that value.

The plausible range remains hundreds of birds wide because the underlying exposure estimates themselves vary substantially among wind facilities.

Location changes the risk dramatically

The model estimated 2024 median mortality of 44 eagles in the lower-risk zone and 136 in the higher-risk zone, although those components should not simply be added to reconstruct the overall median because they come from separate posterior distributions.

The difference comes from exposure.

Eagles were estimated to spend far more time within turbine-height airspace in the higher-risk landscapes. Tethys

That makes turbine siting as important as turbine quantity.

The authors note that a relatively small number of turbines placed where eagle exposure is high could produce mortality comparable to a much larger buildout in lower-risk country.

Bigger turbines amplify that calculation because each rotor enlarges the volume through which an eagle can fly. Tethys

The uncertainty is part of the result

Golden eagles make this question unusually sensitive because they are long-lived and reproduce slowly.

The study cites previous demographic work indicating that adult survival has particularly strong leverage over population growth, while human-caused mortality already accounts for a large share of deaths among older birds. Tethys

But this paper did not measure a population decline caused by wind turbines.

It modeled collision mortality and explicitly carried uncertainty through the calculation.

The authors expect those intervals to narrow as more exposure and mortality data become available.

For now, the strongest number is therefore not 270 by itself.

It is 270 surrounded by 72 and 877.

The model says the collision toll has probably grown substantially as western wind infrastructure expanded and rotors grew larger.

It also says scientists still do not know the annual body count closely enough to express it as one clean number.

The full study can be read here: Gedir, J. V., Gould, M. J., Millsap, B. A., Howell, P. E., Zimmerman, G. S., Bjerre, E. R., & White, H. M. (2025). Estimated golden eagle mortality from wind turbines in the western United States. Biological Conservation, 302, 110961.