This paper is a very welcome addition to the ongoing discussion on the adoption of improved ozone cross sections for stratospheric ozone measurements. The paper also provides a clearer understanding of the effects of stratospheric temperature and instrument parameters on the values retrieved by Dobson and Brewer instruments, and how the well known seasonal differences between the two are virtually eliminated when these effects are included in the retrieval algorithms. The methodology is sound and allows the authors to make a set of recommendations for the adoption of new temperature corrected cross sections which are both scientifically sound and realistic. I.e. if the recommendations are to be followed then it must be possible to implement them within the bounds of the current resources available. However, the information is there should more be required for specific research projects requiring more in depth calculation.

Although the paper is well written and needs no major corrections, there could be some confusion in the use of the terms weighting coefficient, slit weighting function and slit function. It may be better if the weighting coefficients are described more fully as the coefficients of the linear combination, and the difference, if any, between slit function and slit weighting function is explained. Otherwise if none, use one or the other. In tables 1 and 2 the weighting coefficients,  w, are included . Again this could be confusing as the discussion in that section is about the slit functions. It could be helpful to remind again here that these are then used to calculate the differential alpha. 

Technical:

line 11 reveals should be reveal

line 33 "Much of these biases has"  to  "Many of these biases have"

line 99 " last columns of Tables 1 and Table 2." to "last columns of Tables 1 and 2"

line 242 "in Fig. 3" not needed

line 267 "TEMIS" is not defined until two lines later

Review of Voglmeier et al.,  The transition to new ozone absorption cross-sections for Dobson and Brewer total ozone measurements

General comments

The submitted manuscript deals with the question of how best to update the processing of total column ozone data from the global Dobson and Brewer networks. Particularly for Dobsons, a wish has existed in the community for many years now to make use of improved measurements of the ozone absorption coefficients (also frequently known as ozone 'cross-sections') and to incorporate the temperature dependence of these coefficients into the standard processing algorithm, in order to improve the agreement between Dobsons and Brewer and between ground-based and satellite-based measurements. However, until now there has been disagreement within the community about how best to introduce the desired improvements, noting that there is a wide variance of expertise and resources across the many different countries and stations of the global network, and a poor implementation of such a major change in the algorithm could potentially introduce a multitude of problems for users of the data stretching well into the future. The Dobson network has existed for many decades and, for largely historic reasons, each site is responsible for processing the raw data and only the derived total ozone values are provided to the data centres.  

The results of this manuscript will be extremely useful in moving this long-needed implementation forward and thereby significantly improving the quality of total ozone data from ground-based instruments, particularly Dobsons.

The authors maintain a very specific and pragmatic focus, and carefully assess various options from the perspective of what will make the biggest improvement to the quality of the data while being realistically feasible to implement.

At a number of points they differentiate between what is recommended for the network as a whole, and what is better considered the subject of further investigations by well-resourced operators able to carry out such research. Further improvements in the algorithm are certainly still attainable, but would be less practical to implement operationally.

The authors are to be commended on their very clear and thorough approach. I found all sections of the manuscript were examined carefully and logical conclusions discussed, which were very easy for the reader to follow.

No great scientific breakthroughs have been presented, of course, but I still consider the manuscript to be work of high quality and undoubtedly very relevant and worthy of publication in AMT.

I have little doubt that the relevant authorities will choose to implement all the recommendations presented.

I have only a handful of comments of any substance.

(1) In figures 5 and 7 I think it would be very valuable to add at least one station from the southern hemisphere mid-latitudes to show the global applicability of your results. If there is space, it would also be helpful to show a site closer to the equator than Izana with an annual cycle of a different shape.

(2) You don't discuss trends. An objection to using the climatology is that, in theory, a long-term systematic trend in either stratospheric temperature, or the relative vertical distribution of ozone, would introduce a spurious trend in total ozone values. Presumably you have looked at this and found that the effect would not be significant, however I think it would be worth adding a sentence or two on this subject.

(3) There is a comment I found odd in tables 3 and 4 in which it seemed the various co-authors obtained different answers for the same calculation, which should be clarified.

(4) The use of multiple wavelengths to largely eliminate the effect of aerosols is presented only very briefly (lines 96 and 97) and would not be able to followed by someone new to the field. I would prefer you add just a little bit more detail and ideally, show why the aerosol terms do not algebraically interfere with the later formulations, for example, equation 12 on line 315.

Specific comments

Lines 89-90 To make the meaning clearer it would be worth adding that this is in contrast to some other studies, eg Gröbner et al. who used a non-standard (updated) Rayleigh formulation. I realize you come back to this point later on.

Line 124 -  slant path not pass

Lines 184-194 (Section 2.3.2) How well does the dispersion test really work for Brewers? The comment about Redondas et al. is a little bit sketchy compared to the Dobsons. I know you touch on this point again later. The broader question here is whether the general applicability of the results to the global network is compromised by large variations in the slit functions of operational instruments? Again I realize that you return to this point later on.

Line 200 I would prefer something like "four sets of cross-sections" rather than "four cross sections"

Line 209 I find the word "broadband" misleading here because we are still talking about high spectral resolution

Lines 249-251 This formula takes into account the temperature, pressure, humidity and CO2 concentration, I believe – how do you determine these inputs?

Line 273 Change "mandatory" to "necessary"  ("mandatory" has  tones of needing to follow orders, to me). This is an important point though which I think has been missed in some previous work.

Line 290 – Figure 4 caption - "OS-LIDAR minus TEMIS" please clarify whether this is "daily" or "climate" for both.

Line 296 – Would it be possible to also include a southern hemisphere site at a similar latitude to MOHp? (This shouldn't clutter up the plot too much because it will be out of phase) The  day-to-day variability would be lower in the SH than the NH I expect. If you have space it would be good to also include an equatorial site. Izana is at a relatively low latitude but clearly still has a NH mid latitude type seasonal cycle, albeit of smaller amplitude.

Line 331 – I don't understand what you mean by "independent evaluation by co-author" – are you saying the different authors calculated the same quantity but came up with different numbers? That would be strange, and something I have never seen in a published paper before! I suggest trying to resolve the discrepancy and only including one value in the table. If the method was slightly different than that should be explained to the reader. Or, do you mean these are the values previously given in Gröbner et al. (2021)? Then it would not be surprising if there were slight differences.

Line 545 Personally, I think the expression "way to go" used here is too informal for a scientific paper.  

Line 580 Another very minor point – the author contributions should be written from the perspective of all the authors as a team. I don't have any doubt WS "offered valuable insights" but this only makes sense from the perspective of the lead author. For WS to claim he offered "valuable insights" would be very boastful!

The paper by Voglmeier et al. is a very well-written paper that has important implications in long-term stratospheric ozone monitoring work, related satellite validation, and ozone recovery trend studies. Dobson instruments provided the most trustworthy total ozone column (TCO) measurements on global scales since the 1920s. The authors explored four new ozone cross-sections and concluded that the “SG16” will decrease Dobson measurements uncertainty from 3-4% to better than 2%. Implementation of the correction routine suggested will significantly improve the Dobson network’s TOC data quality. In general, the approach suggested by the authors is logical and practical (i.e., centralized correction via WOUDC, instead of corrections by individual sites). However, I can see such work will need to be coordinated between WOUDC and the Dobson community and validation and verification work would be needed in future. Nevertheless, this work is a good match with AMT and has a significant impact. I highly recommend publishing this work after addressing the following minor issues.

Line 17: define TEMIS

Line 17: Some general numbers for this decreased seasonality difference should be included in the abstract. E.g., decreased from xxx +/ xxx% to …

Line 25: reference or link of WOUDC could be provided

Line 33: the number of fixed effective temperatures for B&P should be included

Line 43: here “significantly” is hard to define, could use some general numbers. Note that, a 3-4% bias might be large for TOC, but could be trivia for some other GAW measurements.

Line 69: Should include a proper reference for Brewer SO₂. E.g., 10.1029/98GL51305

Line 70: define DU

Line 97: The definition of AD or CD pairs was not introduced

Line 140: Figure 1 shows the Teff, but without information on where this came from. Is this based on reanalysis data or measurements?

Line 154: changed “averaged” to “convolved”?

Figure 2: SG16 is first shown in the paper, please define it.

Line 267: A bit more information on the TEMIS dataset should be provided. Which ECMWF reanalysis has been used? ERA-5, ERA-interim, or something else? Which pressure levels and/or time stamps (e.g., daily mean or sunlit period only?) have been used in the Teff integration? The link provided did not show any further details. Or, at least, some references should be provided.

Line 411: Could you please provide some reasonings why the BW dataset had the largest changes? Theoretically, good ozone xs will improve the agreement between Brewer and Dobson, so I fully agree with the authors that BW’s performance is not ideal. But, some reasoning and/or explanations might be needed.

Figure 5, 7 & Table 6: Using Amundsen-Scott as an example, it is obvious that daily Teff from TEMIS could have larger differences than its climatology results. I.e., for Oct. 2019, the difference between daily Teff (blue line) and climatology (dashed blue line) could be as much as 20 degrees. For Alert, the difference could also be as large as 10 degrees. In Fig 7, the Rel RIC difference for the Amundsen-Scott example is more than -2%, beyond the 2 sigma level. It looks like a similar issue for Alert in 2019 January.

Also, if we are using 1990-2020 climate to correct observations from Dobson records from 1960-70, will that introduce some more errors? The value of long-term Dobson records is for ozone recovery trend studies. Any suggestions?

Figure 7: I would not use instrument numbers as the title for the figure. This puzzled me as to why D104 and B010 were at those four locations. Anyway, could be less misleading. E.g., change to “Using D104’s slit functions”.

Line 549-550: Please add “at the Hohenpeissenberg site”. I would expect the difference at high-latitude sites could be larger than 1.2 degrees C.

Line 555-556: For Dobson data on WOUDC, the daily ozone values can be traced back to 1926, which is sparse anyway. However, observations in the 1960s and 70s already jumped to 15000 total observations each year, which is similar to the 1980s-2000s levels.  Do you think using Teff climatology that only covers 1990-2020 will have any impact on corrections from the 1920s to the 1980s? Such impact on long-term ozone trend analysis should be at least mentioned and maybe discussed.

Line 560: To be practical, I fully agree with the authors such work should be carried out by WOUDC. Also, do you think there should be two versions of Dobson data published on WOUDC for the transition period? Anyway, the WOUDC team should be included in future discussions with the Dobson ozone cross-section committee.

Line 571-572: please include some quantitative numbers in such conclusion part.