Wind-blown sand trajectory database

To support quantitative analyses of aeolian transport processes and sand transport flux under multiple wind-velocity and grain-size conditions based on complete trajectories, we developed a comprehensive trajectory database of wind-blown sands (758,295 trajectories in total) that contains complete saltation trajectories constructed by the Deep Learning-based Saltation Trajectory Construction (DSTC) method. The database consists of three subsets: (1) a basic database with 647 constructed complete trajectories of medium (M) sands under 8 m s⁻¹; (2) an extended database with 409,766 trajectories covering a wide range of wind velocities (8–20 m s⁻¹) and grain-size conditions with fine (F), medium (M), medium-coarse (MC), and coarse sands (C); and (3) a complex database with 347,882 trajectories generated for extreme environments, including the Tibetan Plateau (5,000 m a.s.l. at the Tanggula Mountains) and Mars (gravity = 3.71 m s⁻²; air density = 0.02 kg m⁻³).