The manuscript titled "Retrieval of aerosol properties from zenith sky radiance measurements" by Herrero-Anta et al., presents the procedure to extract aerosol properties from zenith sky radiances measured by a newly developed instrument. The inversion algorithm is performed with GRASP and the results are compared with AERONET products. The methodology is well explained, providing details of the calibration, inversion strategy and sensitivity analysis. The manuscript is well written and it presents a substantial contribution to scientific progress within the scope of Atmospheric Measurement Techniques. There are several issues that require a better explanation, and also some minor issues and typos that should be corrected, highlighted in the attached file.

Firstly, the main difference of this measurement strategy respect to AERONET is the use of vertical data instead of that derived by multiple-angles. The AERONET instruments perform principal plane and almucantar measurements to obtain angle information, while the ZEN instrument only measure vertically. Please comment on these different strategies and their effect on the inversions. For instance, would CIMEL vertical measurements (like "Cloud mode" but on clear-sky days) be as reliable as PPL and almucantar? The authors conclude that the GRASP-ZEN measurements produce reliable volume concentration estimations, but not radii estimations. Is it related with the limitation on the measurements (only vertical) or the inversion procedure (limited to five size distributions)?

Secondly, the normalization by extraterrestrial spectra is not clear. The authors conclude (lines 578 & 579) that “proposed methodology incorporates the advantage that it includes the normalization used by GRASP and therefore there is not any need to use extraterrestrial spectra to normalize the data when they are used as input in GRASP”. However, section 2.2 “GRASP methodology” seems to imply that normalization is required. Please clarify this issue

Finally, some technical issues:

• The procedure described in Section 4.2 AERONET scenarios is not clearly explained. Does the AERONET inversions determine the five aerosol types concentrations? Or the AERONET information is input into GRASP directly?
• Details in Figure 1 are difficult to see. Maybe a logarithm Y-axis would help. Also, the details of the fitting may be added, and a residuals graph, to check randomness

Please find attached the review of the paper egusphere-2023-1040 “Retrieval of aerosol properties from zenith sky radiance measurements”

Please refer to the attached file

This manuscript describes the retrieval of aerosol properties with the synergetic use of zenith sky radiance measurements at 4 wavelengths and GRASP.

The instrument used ZEN-R52   radiometer,  is a simple radiometer with uncertainties arising mainly from the temperature dependencies.  The authors have done a laboratory characterization of the instrument, introduced corrections and accessed their uncertainties,  which were further used in the inversion algorithm.

This paper provides detailed information for the normalization,  validation and inversion strategies followed. 

This work provides new insights in the possibility of retrieving aerosol properties  (AOD, VCT, VCC and VCF)  using  a simple measurement geometry and skipping the laboratory radiance calibration(or using it for stability monitoring purposes).  The results are very promising, and the authors provide information to the community of possible improvements through their validation against AERONET products and uncertainty budget.  

The manuscript is clear, but it would improve,  if a more concisely writing style was used.

I propose that this article is accepted for publication, after minor revisions.