This dataset was used to identify the effect of winds and waves on albatross flight performance, specifically a reliance on wing flaps. It consists of hourly GPS positions for 370 foraging albatrosses nesting in Bird Island, South Georgia (Southern Ocean) or Midway Atoll (North Pacific) during breeding seasons between 2018 and 2023. The dataset includes identifying characteristics of each albatross, including species, nesting location, field season, trip type (incubation or brood-guard), and a unique ID. Study species include black-browed, grey-headed, and wandering albatrosses at Bird Island, and black-footed and Laysan albatrosses at Midway Atoll. Windspeed (km/h) and swell height (m) experienced by foraging birds were extracted from ERA5 hourly data on single levels (hourly temporal resolution, 0.25x0.25 degree spatial resolution). Wing flaps per hour were calculated by using LULU filters to identify peaks in the z-axis of accelerometer data (tags mounted on the central dorsal contour feathers of the albatrosses. Times where birds were resting on water were removed from this dataset using Hidden Markov Models on fine-scale GPS data to ensure that analyses were performed on albatrosses in flight. Generalized Additive Models were used to identify the effect of wind speed and wave height predictors on flap rate.

Please refer to Maywar et al. (doi.org/10.1186/s40462-025-00614-w) for more details regarding the Methods used to create this dataset and to run subsequent analyses.