In the tracks of a whale: inferring size class, orientation, and swimming speed from thermal flukeprints

Despite decades of innovation, monitoring whales remains constrained by their brief surfacings, wide-ranging movements, and limitations of conventional survey methods. Unmanned aerial vehicles (UAVs) and thermal infrared (TIR) sensors offer non-invasive alternatives, yet most applications rely on short-lived direct detections. Here, we evaluate a novel approach using UAV-based TIR imaging to extract biological and behavioural information from thermal flukeprints, surface disturbances generated by whale tailbeats. Field surveys of free-ranging humpback whales (Megaptera novaeangliae) off Reunion Island demonstrated that flukeprint width robustly distinguished calves from adults. Flukeprint spacing provided swimming-speed estimates consistent with RGB-based measurements (mean error 0.42 m s⁻¹; median relative error 16%), while flukeprint-derived orientations aligned closely with visual headings (mean offset -3.1°). These results indicate thermal flukeprints capture age-class, movement, and directional information even when animals are partially visible or submerged. By extending observation beyond direct sightings, TIR-detected flukeprints offer a low-impact, complementary tool for cetacean monitoring.