The probability distribution function (PDF) of the liftoff velocities of aeolian (wind-blown) sand particles, an important parameter in sediment transport and dust emission models, remains poorly understood due to the challenges in tracking particle dynamics during the liftoff process. To reliably describe this distribution function, measurements of the individual particle trajectories near the sand-bed during sediment transport are required. In this study, we address this issue by employing an improved particle tracking velocimetry technique, enabling us to capture the particle-bed collision process of many liftoff grains in consecutive images and, consequently, to obtain multiple liftoff particle trajectories with different liftoff modes during sand transport in a boundary layer wind tunnel. Subsequently, using the wind tunnel, we measured the trajectory of splash particles, rebound particles that generated no splash particles, and rebound particles that generated only one splash particle near the sand bed. By analyzing the probability distribution of the liftoff velocities of these three liftoff modes, a probability distribution model for the liftoff velocities of particles in a steady-state wind-sand flow was constructed. The results indicate the following: (1) The liftoff angles and speeds of the three liftoff modes all follow a Lognormal probability distribution. (2) The PDF for the liftoff angle of the total liftoff particles is largely unaffected by the liftoff mode. (3) The PDF for the liftoff velocity of the total liftoff particles is mainly controlled by the number of splash particles to that of rebound particles. This indicates that the development process of wind-sand flow or turbulence fluctuations impacts the PDF of liftoff velocity. The different shapes of the probability distribution curve for liftoff velocities may only reflect the specific development of wind-sand flow. Our results offer new insights into the aeolian sand saltation process near the bed surface and are crucial for simulating particle-bed interactions and improving geomorphic dynamics models in aeolian sand transport. Our study on the probability distribution model of liftoff velocity opens a new avenue of investigation in the aeolian and soil sciences and aids the understanding of the dynamics of near-bed particle transport. This transport plays an essential role in aeolian research, as well as in the geomorphodynamics of Earth, Mars, and other celestial bodies within our solar system.