The authors have addressed all my previous comments. I recommend publication in its current form.

This is a review of the manuscript titled “Vertical Profiles of Aerosol Chemical Species Concentration retrieved through Synergy of Spaceborne Lidar and Polarimeter Observations” submitted to AMT by Merdji et al.

The paper describes an analysis of the performance of the AEROCHEMPro algorithm to retrieve aerosol properties from combined lidar and polarimeter data, based on simulated observations. The subject is very relevant as the combinatiomn of lidar and polarimeter is very promising to advance aerosol observations. The paper is generally well written and figures are clear. However, there are many limitations and assumptions that need to be explained and highlighted more before the paper can be accepted. Hence major revisions are needed as detailed below.

Limitations leading to optimistic results:

There are many limitations and simplications in the work and analysis that need to be explained and highlighted more. I understand many assumptions are due to practical limitations of the analysis and some is left to future work. However, they need to be clearly stated in the manuscript and in the abstract and conclusions to inform the reader of the scope of the paper. Especially the implications of these simplifications need to be made clear, as they may mostly lead to overly optimistic results. Limitations that I identified are listed below.

• If I understand correctly, the forward model and assumptions for the simulated observations and the retrievals are entirely consistent. Hence they make the same simplifications. I I amd correct, please make this explictely clear. This will result in a overly optimistic result, which then also need to be discussed.
• The surface parameters are not fitted according to section 2.1.6. Partitioning between aerosol and surface signals is one of the main challenges in aerosol remote sensing, so this is a very important limitation again leading to overly optimistic results.
• From the very low differences between simulated and retrieved U/I in table X, I suspect that the geometries used for the observations and retrievals are all assuming the principal plane, in which U is essentially zero. Polarimetric retrievals are sensitive to geometry and a principal plane geometry may be consisdered a best-case scenario. Please make clear which geometry is used and discuss the implications of the simplifications in this assumption (if applicable).
• The first guess used in the inversion is very important, as a first guess too far away from the truth may result in a non-covergence, while using the truth as first guess may lead to overly optimistic results. The first guess is not discussed in the manuscript, neither is a definition of (non-)convergence. Please discuss this and also the implication if a overly optimistic first guess is used (if applicable).
• In line 732 it is mentioned that only cases with AOD>0.1 are consisdered. Most of the global AOD values are below 0.1, so this is is also a limitation that need to be highlighted more.

Polarimeter noise assumptions:

Generally multi-angle polarimeters aim for an absolute accuracy of the degree of linear polarization (DoLP) better than 0.005, as also indicated for the AOS-MAP in line 456. In their error model, the authors add 3% noise to I and 0.5% noise to Q and U. Depending on the specific values of I, Q, U and DoLP, this can lead to noise greater or smaller than 0.005 on DoLP. Please discuss in the paper the relation between the choises made for the applied noise and the target of DoLP absolute uncertainties being below 0.005.

Other specific comments:

Table 3 : The number of significant digits varies throughout the table. Please make this consistent.

Table 4: There are additional entries for Nadir resolution and Swath width in the table. I assume these are threshold values. Please make this clear.

Line 500: Please do not use math mode for the text in the equations

Section 2.3.1: Here the surface parameters are described as part of the retrievals set up. However, they are not retrieved according to section 2.1.6. It is unclear how the surface parameters are kept fixed. Is that through the gamma assumptions? Please explain clearly in the text that the surface parametrs are fixed and how this is achieved.

Line 610 and further: Instead of “10-5” use E or superscript. 

Figure 7:

• How do these differences relate to the added noise levels? Explain in the manuscript
• As discussed in the main comments I suspect that the geometries are assumed to be principal plane, in which U is zero. It then is also meaningless to report on the errors on U, so then I suggest to leave this out and explain in the text.

Line 665: These claims on AOD per specie nessecitate the reader to look at the supplement figure. I suggest adding it in the main text or limit this text to general statements. 

Figure 9: LR and SSA are vertically varying parameters, but only one value is plotted here. What is the definition of reported LR and SSA here (and in section 3.2)? Please explain in the manuscript

711: Please change "surface-2km" to “between surface and 2km” or something similar