Stochastic parameterization of dust emission and application to convective atmospheric conditions

Abstract. We develop a parameterization scheme of convective dust emission for regional and global atmospheric models. Convective dust emission occurs in the absence of saltation as large eddies intermittently produce strong shear stresses on the surface and entrain dust particles into the air. This dust emission mechanism has not been included in the traditional dust models. The scheme presented in this study is a new approach which takes account of the stochastic nature of convective dust emission. It consists of the statistical representations of soil particle size, inter-particle cohesion, and instantaneous surface shear stress. A method of determining the probability density function of the latter quantity is proposed. Dust emission is then estimated from the overlap of the probability density functions of the aerodynamic lifting and inter-particle cohesive forces. The new scheme is implemented into the WRF/Chem model and applied to dust modeling in the Taklimakan Desert. A comparison with lidar data shows that the model can reproduce the main features of the dust patterns and their diurnal variations. For the case studied, convective dust emission is typically several μg m⁻² s⁻¹ and at times up to 50 μg m⁻² s⁻¹.