Fine-scale flight behaviour of golden eagles reveals the use of different uplift sources in a topographically complex region

Soaring birds are highly dependent on rising air currents (uplifts) to fly long distances at very low energetic costs. Previous studies, based on observations and low-resolution atmospheric data, have shown that soaring birds tend to use mainly two types of uplifts, orographic and thermal uplifts, to which they respond using two typical flight modes: linear and circular soaring. In this project, I combined a high-resolution behavioural dataset with a high-resolution weather model to investigate the detailed behavioural response of a highly efficient soaring flyer, the golden eagle (Aquila chrysaetos), to fine-scale atmospheric dynamics in the topographically complex regions of the Alps and Apennines. Following a recent study in collaboration with meteorologists at ETH Zurich, I developed an automated uplift classification algorithm that categorizes the uplift source of 13,421 soaring events into orographic, thermal uplift, and gravity waves, for the years 2020 and 2023. After a detailed classification of the atmospheric context in which the eagles flew, I used a random forest algorithm to combine their behavioural response to the three uplift sources based on several flight metrics. My results suggest that eagles fly very flexibly in their aerial environment and can use a similar range of behavioural patterns in all the uplift sources explored, challenging the common view that directly associates binary flight modes with uplift sources. This analysis also highlighted the limitations of high-resolution weather models in resolving fine-scale uplift events such as thermals. As a final step, I therefore explored how the climbing rate of soaring birds could be used to inform the vertical wind field of weather models, and suggest this as a first future step for integrating animal movement into meteorological models.