Persistent and non-negligible aerodynamic entrainment during dust emission: new evidence from field observations and wind tunnel experiments

Abstract. Dust emission is a key process in wind erosion that influences climate and air quality. Consequently, the physical mechanisms of dust emission have attracted widespread attention. However, whether aerodynamic entrainment can persist during natural dust emission events remains debated, and this mechanism is often neglected or treated as secondary in dust emission parameterizations. To address this dispute, we combine field observations and wind tunnel experiments to investigate the persistence and contribution of aerodynamic entrainment. Field observations show that, in the observed plots where saltation rarely occurred, aerodynamic dust entrainment persisted throughout the emission process, whereas in most of the observed plots, the threshold friction velocity for dust entrainment (TFVa) was generally lower than that for saltation initiation. Even where the TFVa exceeded that for saltation, dust emission persisted beyond the cessation of saltation and its emission rate was enhanced by saltation disturbance. Wind tunnel experiments further demonstrated that aerodynamic entrainment persisted and contributed substantially to total dust emissions, accounting for 16–62 %. Moreover, the enhancement of dust emission by saltation resulted primarily from surface renewal and aggregate breakdown, which continuously replenished erodible fine particles. Furthermore, existing dust emission models based solely on aerodynamic entrainment or saltation bombardment fail to reproduce observed dust emission magnitude and temporal dynamics. Together, these findings indicate that aerodynamic entrainment is persistent and non-negligible, while saltation contributes not only through dust production but also by modifying surface erodibility. We therefore propose an “entrainment–saltation–renewal–entrainment” framework and suggest future models should consider coupled multi-mechanism processes.