Aeolian erosion of granular materials is investigated by means of a mathematical emission model and experimental wind tunnel measurements. Main model input data are: friction velocity, relationship between cover rate (CR) and eroded height (H) and particle proper- ties (density, size distribution). It is proposed: (1) to evaluate the linearity of the relation between CR and H considering the presence of a multimodal distribution of particle sizes, (2) to validate the mathematical model with wind tunnel data, (3) to evaluate the protec- tive effect of non-erodible particles and (4) to qualitatively evaluate the final stage of ero- sion through experimental photographs of the oblong stockpile. The relationship between CR and H may still be considered linear for the tested mixture of particles. The modelled emission, when compared with experimental data, showed that the physical tendency of the aeolian erosion phenomenon was well predicted. The model showed to be useful in comparative analysis between scenarios but not in absolute values due to errors found. It is valid for the monitoring of air quality degradation due to aeolian erosion of open yards of storage piles. Detailed analysis of emitted mass explained that the smallest diameters among the non-erodible particles create a less effective protection effect leading to higher emissions. The qualitative analysis of high-quality photographs of the experiments showed that the non-erodible particle agglomeration on the stockpile surface can be well explained if one evaluates simultaneously, on the pile, the angle of velocity vectors (which influences the threshold friction velocity value) and shear stress.