Bruckner et al use the RAQMS-Aura chemical reanalysis to explore the relationship between tropospheric O₃ and ENSO. Using observations from the TRMM satellite, they demonstrate that their reanalysis captures the variability in tropical convection. Then, using composite analysis, EOFs, and a multiple linear regression they show that ENSO is the dominant driver of O₃ variability in the tropics. The MLR analysis highlights that ENSO-related dynamical changes are the dominant driver of this ENSO-related variability, although biomass burning also has a meaningful contribution. This paper is suitable for publication in ACP after the following minor issues are addressed.

Line 95: What AIRS CO products are you using? The total column, which is not advised, or the layers product? If it is the layers product, please describe which layers are included.

Section 3.1: While you perform a thorough evaluation of the RAQMS-Aura precipitation, a discussion on the evaluation of the accuracy of the O₃ and CO using independent data is also warranted. While the anomaly plots in Section 3.2 do suggest that the reanalysis captures the ENSO related variability of the column amounts of these species, a more thorough analysis akin to the TRMM analysis would be beneficial. If this has already been done in a separate paper, a simple reference would suffice. Also, how accurate is the vertical O₃ in the reanalysis? This could be evaluated with something like SHADOZ ozonesondes.

Figures 4 -6: Why not combine these into a single figure?

Line 332: Show how you are defining ozone production. What reactions/species are included here?

Line 410: In your MLR analysis, you include both the PC1_CO and PC1_precip terms, which are somewhat related (r = -.435). How does this relationship affect the interpretability of the results of your MLR? How can you be sure that the impacts of one variable aren’t being convolved with the impacts of the other?

Section 3.3.2: Oman et al (2013) demonstrated an impact of the QBO on upper tropospheric O₃. Can you separate the QBO and ENSO impacts on your results, since the Nino3.4 and QBO EOF1 indices are relatively strongly anti-correlated? How does omission of this term from your MLR affect the results? Some discussion of potential QBO impacts on your results is warranted.

Oman, L. D., Douglass, A. R., Ziemke, J. R., Rodriguez, J. M., Waugh, D. W., & Nielsen, J. E. (2013). The ozone response to ENSO in Aura satellite measurements and a chemistry-climate simulation. Journal of Geophysical Research-Atmospheres, 118(2), 965-976. https://doi.org/10.1029/2012jd018546

This study takes a closer look at ENSO related variability in tropical tropospheric ozone from the RAQMS-Aura chemical reanalysis. Using multiple linear regression analysis and compositing techniques to show the dominant drivers causing this ozone change depending on region. While the calculated response is similar to what has been shown in some past studies, the use of RAQMS-Aura appears novel including ozone production and loss terms, convective mass flux, and diabatic heating and should be of interest of ACP readership. I would recommend publication pending the authors consideration of a few mostly minor suggestions below.

I think it would be particularly helpful if the authors consider especially for figures 8-14, and 20 to show some indication of significance in the response plots to put the changes in context and help draw the readers eyes to the key areas.

Some of the ozone response to ENSO in the subtropics is related to stratosphere-troposphere exchange, are there any concerns about representing that properly with a model having 35 model levels. Does the use of assimilated data help with any model deficiencies related to resolving those processes?

I don’t believe you cite Oman et al. 2013, please have a look and note some of the discussion related to Quasi-biennial Oscillation (QBO) and its influence that can extend into the troposphere and consider checking and possibly adding terms to your regression analysis to account for its influence.

Oman, L. D., A. R. Douglass, J. R. Ziemke, J. M. Rodriguez, D. W. Waugh, and J. E. Nielsen (2013), The ozone response to ENSO in Aura satellite measurements and a chemistry-climate simulation, J. Geophys. Res., 118, 965–976, doi:10.1029/2012JD018546.

Any concerns about the shift in bias noted in precipitation fields between 2009-2011, compared to observations, impacting ozone response, are there corresponding shifts in any of the ozone response terms.

Lines 404-405 and comment in general: Shifts in precipitation also impact clouds, would this impact be reflected in your net ozone production related to changes in photolysis rates. Or stated more generally on the net ozone productions term can you differentiate that caused by biomass burning emission changes with clouds and photolysis changes.

Lines 51-53 while La Niña responds in a generally opposite manner it is not quite symmetrical 

Line 130 al missing period al. 

Line 339 need to subscript 3 in O3