We thank the referee for comments and questions. They help us to explain better several aspects of the paper. Here we are indicating our answers in boxed frames after each point raised by the reviewer and our changes/actions in the manuscript within a green colour box. C1 ACPD Interactive comment Printer-friendly version Discussion paper INFORMATION: The Table 6 has been double-checked by the different modelling groups. CNRM reported that, instead of our previous estimate, their diagnostics of

more »

... ry deposition are not including sedimentation which means different values of total and dry deposition (without sedimentation). With this revision the CNRM-6DU model has a larger bias due to an unclosed budget but the CNRM-3DU decreases the previous bias by a factor 2. In this situation we have removed the model CNRM-6DU from the multi-model mean, but we kept the CNRM-3DU. Given scale of the differences between models and observations, the comparison of total deposition draws the same conclusions and the results are very similar. Because the dust emission scheme is not affected by the bias, we kept their results in the analysis. All the Tables and Figures has been revised, and several of them improved according to the new information. General comments This manuscript presents the results of five Earth System Models simulations of the global dust cycle, emissions, dry and wet deposition, optical depths, and surface concentrations comparatively to satellites and in situ observations. The authors explore global and regional variability between models in three different simulated experiments: PD (calculated winds), PDN (reanalysed winds), and PI (prescribed chemistry and aerosols). Overall, the manuscript is well written and provides ample content. Having said that, this manuscript is quite extensive and important information is left for the reader to find in the supplement. The content of this manuscript could be divided in two different publications. In the first one, you could explore the differences between the five models, and then, in the second, you could explore more deeply the differences between the three simulated experiment scenarios. C2 ACPD Interactive comment Printer-friendly version Discussion paper