A Monte-Carlo based 3-D ballistics model for guiding bat carcass surveys using environmental and turbine operational data

Shivendra Prakash; Corey D. Markfort

doi:10.1016/j.ecolmodel.2022.110029

•Integrates 3-D ballistics model with Monte-Carlo technique.•Derives parameter distributions from SCADA, bat biometrics data, and literature.•Demonstrates the model capacity in guiding individual survey visits.•Displays model utility in planning carcass surveys and seasonal fate prediction.•Identifies the need to investigate the rotor strike distribution. Wind turbines cause direct mortality of bats through collision with turbine blades. However, accurate bat fatality estimates are unavailable for many facilities and regions due to lack of physically consistent guidance for surveying bat carcasses. We develop a Monte-Carlo based 3-D ballistics model to address this important gap to guide carcass surveys, help improve fatality estimates by combining it with statistical models and offer post-construction impact assessment guidelines. Monte-Carlo simulations account for variability of environmentally relevant parameters and subsequently the variability of the generated carcass fall zone distributions. The input parameters are bat characteristics, including mass, size, and drag coefficient, wind speed, turbine RPM, yaw, bat flight speed, bat strike angle, and bat strike locations on the rotor. Using turbine SCADA (Supervisory Control and Data Acquisition), bat biometrics data, and carcass drag coefficient range, probability density functions of the input parameters were computed. With the help of carcass survey and SCADA details, bat fatalities were selected, and Monte-Carlo based ballistics model simulations were performed using corresponding values of the input parameters of the ballistics model. The surveyed carcass location for a bat found on the same day was in good agreement with the modelled fall zone distribution, providing a measure of model validation. Next, simulations were conducted for a bat fatality for which the carcass was found four days after collision. It was found that the observed location was within the modeled fall zone distribution showing the model capacity for planning surveys. Further, Monte-Carlo based 3-D ballistics model simulations were performed for the full migration season and the resulting 1-D (radial) and 2-D (surface) histograms were compared with the survey 1-D and 2-D histograms, respectively. The modelled 2-D fall zone distribution hot spot overlaps with surveyed carcass positions indicating the model robustness over a wide range of meteorological conditions and time scales. However, by comparison, the modeled 1-D fall zone distribution was found to be different from the surveyed 1-D fall zone distribution. We hypothesize that uncertainty in bat radial strike location (distribution of bat collisions on turbine blades in radial direction) distribution is primary source of difference between the surveyed and modelled 1-D fall zone distribution. There are no measurements available for this key input parameter. As a next step, the sensitivity of carcass fall zone distributions with respect to bat radial strike location distributions were tested. It was found that the probability density function for bat radial strike location affects the modelled fall zone distributions significantly. There is a need to measure bat radial and angular strike locations which can be used as input in the ballistics modeling framework to improve prediction of carcass fall zone distributions. The proposed model is useful for carcass survey planning, improving survey estimates, designing turbine curtailment studies, and collision risk modeling. [Display omitted]

A Monte-Carlo based 3-D ballistics model for guiding bat carcass surveys using environmental and turbine operational data

View Online

Abstract

Details

Metrics