Strong winds can cause various disasters, such as vortex vibration, flutter, gallop, and buffeting, which can significantly decrease the service life of bridge structures. Therefore, it is crucial to accurately characterize the wind field characteristics at bridges. This study presents a comprehensive framework integrating data-driven methods and physical modeling to characterize wind field randomness and predict bridge tower deformation. Long-term monitoring data from the Jiashao Bridge in 2022 are utilized to analyze spatiotemporal variations in wind speed, direction, and attack angle. The paper also establishes a probability model that considers the uncertainty of wind shear index. Furthermore, this paper discloses deformation data by five GPS receivers, which can be used to characterize the spatial deformation of main beams and pylons. The spatial deformation data of the bridge tower top correspond to the wind field data of the tower top. Therefore, a data-driven method is developed for predicting wind-induced spatial deformation of the bridge tower top. In other words, with wind field data obtained through monitoring or future wind field predictions, it is possible to predict the spatial deformation of bridges. The proposed methodology bridges the gap between wind field characterization and structural response prediction, offering a practical reference for bridge safety assessment and operational warnings.
Ran et al. (2026) studied this question.