This paper presents a soil-derived dust emission scheme designed to accurately represent desert dust sources for atmospheric transport models simulating the desert dust cycle. The scheme considers two key factors: the size distribution of erodible particles, which controls erosion threshold and emission strength, and surface roughness, which affects wind friction velocity. These parameters are incorporated into a formulation of the threshold wind friction velocity, adapted from Iversen and White (1982), and a drag partition scheme derived from Arya (1975). The threshold friction velocity is included in an horizontal flux equation proposed by White (1979), allowing for specific production rates for each soil size range. The dust flux is then expressed as a fraction of the total horizontal flux, with the ratio determined by soil clay content. The scheme's steps have been validated independently against experimental data, showing satisfying agreement and reducing uncertainties in dust flux simulations compared to previous studies. The physical approach is detailed, including the parameterization of the threshold friction velocity and the drag partition, and the computation of horizontal and vertical fluxes. The validation section discusses the theoretical limitations and application conditions of the scheme, emphasizing the importance of accurate soil size distributions and roughness lengths.This paper presents a soil-derived dust emission scheme designed to accurately represent desert dust sources for atmospheric transport models simulating the desert dust cycle. The scheme considers two key factors: the size distribution of erodible particles, which controls erosion threshold and emission strength, and surface roughness, which affects wind friction velocity. These parameters are incorporated into a formulation of the threshold wind friction velocity, adapted from Iversen and White (1982), and a drag partition scheme derived from Arya (1975). The threshold friction velocity is included in an horizontal flux equation proposed by White (1979), allowing for specific production rates for each soil size range. The dust flux is then expressed as a fraction of the total horizontal flux, with the ratio determined by soil clay content. The scheme's steps have been validated independently against experimental data, showing satisfying agreement and reducing uncertainties in dust flux simulations compared to previous studies. The physical approach is detailed, including the parameterization of the threshold friction velocity and the drag partition, and the computation of horizontal and vertical fluxes. The validation section discusses the theoretical limitations and application conditions of the scheme, emphasizing the importance of accurate soil size distributions and roughness lengths.