This work present and iterative method to obtain lidar ratio at 292 nm, which is later applied, to Langley Mobile Ozone Lidar (LMOL) backscattered signal. Once lidar ratio at 292 is estimated, authors use the classical Klett method to obtain independent aerosol extinction and backscattering. It is well-known in lidar literature that Klett method can not provide accurate estimates of extinction profiles because of possible variations of lidar ratio with height.

Nevertheless, the authors try to address an important challenge and provide an estimation of lidar ratio at 292 nm. Typically, backscattered lidars use co-located measurements of sun-photometry AOD for estimating lidar ratios (see for example MPLNET or EARLINET/ACTRIS retrievals). Currently, there are not many radiometric measurements that provide aerosol AOD at 292.

However, I do not rely in the approach presented by the authors. It might need further explanations. But as I understand they propose iterative variations of lidar ratio in LMOL system and provide different aerosol extinction profiles. The range of variation of lidar ratios is not enough because absorbing aerosol can present lidar ratios larger than 90, and OMI satellite retrievals demonstrated the importance that aerosol absorption might have extinction. On the other hand, I understand that they vary Angström exponent iteratively in HALO system to obtain equivalent aerosol extinction at 292 nm. They are ignoring possible effects of variations of Angström exponent with altitude. If so I would rely more on Angström exponent measurements using sun photometry. Finally, for the evaluation they are using the same data that for the computation in the iterative method, which is not appropriate.

With all these points I propose to evaluate the method with CCNY lidar for 355 nm and make intercomparisons with extinction coefficient at that wavelength computed by Raman methodology.

Section 4.2 does not provide any relevant scientific results. It only shows coherence in the vertical structures of aerosols, and for that it is not necessary to retrieve extinction coefficients. Moreover the study-case selected to demonstrated the novel methodology must be different than that used for the validation.

Finally, section 5 only shows that lidar ratio is the most important parameter in the retrieval of backscattering and extinction, which is widely known in lidar community. What is necessary in section 5 is sensitivity test of the new methodology proposed by the authors, which can be done using synthetic data.

Review of  “Retrieval of UVB aerosol extinction profiles from the ground-based Langley Mobile Ozone Lidar (LMOL) system” by Lei et al.,

This paper describes an algorithm for the aerosol extinction retrieval out of the Langley Mobile Ozone Lidar (LMOL) as compared to 20 coincident flights with the NASA Langley High Altitude Lidar Observatory (HALO) 532 nm aerosol extinction product. This work also accomplishes the first known 292nm aerosol product inter-comparison between HALO and Tropospheric Ozone Lidar Network (TOLNet) ozone lidar.

In general, this paper would benefit from an additional proofreading.

Major Comments:

In general this is a very technically developed manuscript. Many of the equations are first principles and well known in the lidar community. In general, substitution of these for graphical elements (flow charts, or  signal  processing chains) would improve the readability. This also allows the author to highlight sections that are new to this original research.

 My major questions

Is this approach actually novel? The authors describe this as working from between 0.5 and 3.5km – does this indicate it may only work properly or is biased for aloft/transported aerosol layers? Please re-emphasize the importance of this work.

Can this method be extended to cases outside of when there were HALO overpasses? Otherwise this does not have as much appeal to the general audiences. A missed opportunity is using the ceilometer to compare with – this is a 24/7 measurement that is made very widely over the country. Then use the HALO data to act as a reference for the quality of the results in this specialized case – and then improve confidence in the ceilometer derived method.

The author described Canadian wildfire smoke, but I cannot tell clearly from any of the images 1) where the smoke resides, 2) how improved the retrieval is in these areas of smoke, or 3) what effects the data set has made to improving the remote sensing of the optical properties of the aerosols.

Minor Comments:

L55 – remove ‘a lot’

L60 -  Langley

L64 “In this paper, the impact of the aerosols was low enough that an aerosol correction to the O3 density was not necessary; otherwise, an interative process would have been necessary (Browell et al., 1985)” – Suggest rephrasing this statement. This sounds like it basically voids the need for this work. Although the ozone correction to the signals may not reduce accuracy, the authors state the uncertainty in ozone is 10-20% - that must impact the uncertainty of the aerosol correction.

L66 – New paragraph for LISTOS

L105  - rather than ‘raw’ data, what is being analyzed? Range corrected-Elastic Backscatter Profiles at 292?

L132 – LISTOS

2.4 The Ceilometer located nearby LMOL – consider moving this up to 2.2 since it was co-located with  2.1.

L160 – Lidar ratio is introduced here but is frequently used in the text up until this point

Section 3 could benefit from some sort of “flow chart” graphic. Or illustration of the changes in the corrected signals after certain steps.

Figure 1 caption -  (a) August 28fternoon needs to be fixed. Are these derived vertically or for a column?

Table 1- 521nm?

L280 – “These results show that by using the selected S1 and AE in Table 1, LMOL has the capability to retrieve aerosol extinction in 292nm with reasonable accuracy.”, What is the estimated uncertainty of this retrieved aerosol component?

Figure 6 – Are you  able to convert the ceilometer to 292nm? Please label in the plot title. Is the PBL height detection “in-house” for the ceilometer or the Vaisala standard product?

Figure 7  - total uncertainty (blue) – should be black. Why is the analog Det Nois decreasing with altitude? Wouldn’t you expect as the signal to be much higher compared to background noise values, that the uncertainty would increase? Is there a need to show both analog and photon counting here?

Figure 7  In some cases in Photon Counting it looks like the uncertainty in using 60sr is less than using the technique  applied in this paper. Is that the case?

Conclusions – rather than say ‘good’, please use specifics from the retrieval results. . Consequently, further research is needed to characterize S1 AE at UVB wavelengths – what exactly is needed?