Thank you to the reviewers for their detailed comments and the authors for their detailed, thorough responses. I have a few more points which are mostly clarifications which I would like the authors to act on – these are mostly minor and should be easily addressed.

Abstract – It would be appropriate to mention the choice and impacts of horizontal alignment in the non-sphericity calculations, since this is a fairly novel and important application on the scale they authors employ in this study.
P3 l1 – ‘output’ – please specify what specifically e.g. dust mass, size, location, altitude etc.
Section 2.2, l18-23 – are the simulations nudged/what boundary conditions are used? ERA-interim? Are dust-radiative feedbacks on atmospheric circulation included?
P10 l 16-28 – Concentrations in bins 1 and 2 are also fairly high in the model compared to observations, which is another common challenge for models (e.g. Kok et al., 2017, Nature Geosciences). It would be interesting if the authors could also speculate/quantify the impacts on pinkness of scaling down the concentrations in bin 1 and 2, in the same way they illustrate this for scaling up bin 5.
P10/11 l29-3, aspect ratios. Please can the authors clarify which orientation their AR is specified in for oblate spheroids. They should also note that few studies have retrieved the 3-D AR such as done by Chou et al., 2008, who found a height-to-maximum-axis ratio of 0.4, besides the AR value of 1.7. Is a 3-D AR considered here, or are two of the main ellipsoid axes the same length?
P11 l15-28 – Fig 7 does not appear to show results for spheres or vertically aligned particles. If this is the case, please state ‘not shown’ in the text to clarify.
P11 l15-28 – What is the reasoning behind only taking forward the case of horizontally aligned particles, given that vertically aligned also provide a lower bound, and both appear equally likely?
P12 l4-10 – It is possible to cite ACPD discussion reviews – please do so here since the additional plots in the author response are valuable to the interpretation of the article (or put the plots in a supplement).
P12 l30 – please add Ouarzazate and Tamanrasset locations to the map on Figure 1
P14 l31 paragraph – What is the reasoning behind returning to a spherical assumption here in Fig 10?
P14 l35 – ‘Saharan domain’ – please give latitude/longitudes here. Is the Sahel included?
P15 l18 – ‘left set’ – i.e. hatched bars?
P16 l31-34 – Similar to my comment above, the additional plots of the BT differences for the extra coarse case are useful to the interpretation of the paper. Please either cite the author response, or add the plots to a supplement.
p 20 l16 – this may be particularly pertinent since dust and moisture are often correlated at dust emission during Saharan summer (e.g. Marsham et al., 2013, 2016).
Data availability – It would be useful to provide the SO99 refractive index dataset in some way, since they do not appear available via the Helmert 2007 publication.
Table 3 – please expand ‘SEV’ and ‘AER’ for clarity in the caption.
Fig 11 – please add to caption what the left/right columns represent.

Thank you for making the requested changes, and congratulations on an excellent paper.