This paper provides results from a multi-model experiment of ozone radiative forcing (using three metrics: instantaneous, stratospheric adjusted and effective) for 2050 relative to 2015 under the SSP3-7.0 scenario. This addresses an important gap in the literature that was not available during the IPCC Sixth Assessment on the likely future evolution of ozone radiative forcing, and will be a valuable resource to the community that will be referenced for years to come.

The title and abstract don’t make it clear that only a single scenario is considered: we trust that the future will follow SSP3-7.0. Either call out SSP3-7.0 specifically or frame in more general terms: “Climate forcing due to future ozone changes in a high emissions scenario…” for example. Why was this scenario selected – presumably because it had the biggest signal? In the abstract, please state the 2050 relative to 2015 timeframe.

Line 56: I think that tropospheric ERF is AR6 for 1750-2019 was +0.41 W m-2 and not +0.45 W m-2. Not that we mentioned it in the AR6 text, but I did do the trop/strat split. Does the Skeie et al. number include climate effects on ozone forcing? More generally, I am supposing that the experimental design here, using 2015 SSTs, does not include the climate effects on ozone either. SSP3-7.0 is a degree or more warmer in 2050 than in 2015 in many models so the effects are probably not insignificant.

The section starting 1.1 on radiative forcing should possibly be promoted to a level-1 section (section 2). Possibly also section 1.2, but could sit under the radiative forcing header. I also thought that section 1.1 was a bit textbook and may not be required for this paper, but the coordinator of TOAR-II likes it, so it’s probably a matter of taste and my familiarity with the topic.

Line 270: “present day” in GISS means 2015? In all cases where “present day" is mentioned, please be specific on the year(s) (e.g. line 337).

Line 337: where do these climatologies come from? Are they model-specific or centrally provided?

Lines 348-349: “the models’ respective radiation and cloud schemes…”: will changes in aerosols (that are not fixed) affect clouds, which will affect the ERF?

Line 473: 298.3 ± 8.3 DU – any observations to compare this to? Figure 6 has data from NIWA, perhaps this could be compared?

Lines 545-548: I’d also say the historical trend of CMIP6 models compared to the obs is quite good, even if biased high.

Figure 8: the 150 ppb ozone tropopause forcings agree quite well between models. Is this expected and/or worth a comment?

Very minor, editorial things

Line 51: reference after full stop

Line 79: write out equation on a new line

Line 116: comma after full stop

Line 310: Walters citation as author (year).

Line 545: Bodeker Scientific doesn’t show up in the references.

Line 576: superscript -2

General Comments

The manuscript compares across different metrics and methods to estimate ozone radiative forcing by first presenting a synthesis of prior work and then calculating the radiative forcing from 2015 to 2050 using the current generation of Earth system models.  The manuscript documents inconsistencies in approaches in prior work as well as unique configurations in some models that complicate a straightforward inter-model comparison. This detailed documentation is invaluable to the modeling community and is a critical piece for interpreting some simulations and possibly the next round of multi-model studies.  It does, however, lead to a lengthy manuscript in which major conclusions may be missed. At the same time, providing some additional context to the abstract and conclusions may help a reader understand the importance of the work.  Two general suggestions:

1. Articulate more clearly the key messages in the abstract/conclusions with some short synthesis statements that provide slightly broader context. For example, the authors may wish to consider the following questions: How much confidence is there in estimates of ozone forcing and the stratosphere versus troposphere contributions as reported here and in prior work?  Given the emphasis on this prior work in the introduction, I expected the authors to conclude by comparing their results to that earlier work. Do the conclusions drawn here have implications for the interpretation of historical radiative forcing estimates for ozone? How important is ozone relative to other greenhouse gases?  How do the emission trends driving the ozone forcing in this scenario compare with other future scenarios?
2. Justify the use of SSP3-7.0. For example, is this the scenario that regional emissions have followed most closely in the last decade or simply the one that all the models ran? What are NOx, NMVOC, and CO emission trends as well as ODS and methane in this scenario? Consider showing a plot with zonal mean 2050-2015 emission changes (and ODS) to highlight regional differences.

Specific Comments

1. The abstract should include the time frame and scenario over which the forcings are calculated.
2. Please clarify what appears to be conflicting findings: Checa-Garcia et al. (2018) report that tropospheric ozone cools the lower stratosphere (line 117) but Figure 9g shows warming in the lower stratosphere in response to increasing tropospheric ozone (fODS).
3. Pages 2-3: Clarify focus here is not only on tropospheric ozone.
4. Line 42: Include example of a non-cloud adjustment the first time this term is mentioned.
5. Figure 1: Clarify if this is for tropospheric ozone only?
6. Section 2.1: Different information is provided for different models (number of chemical species, reactions given for some but not others).  Consider a summary Table with consistent information provided for all models. This table could be main text and detailed text describing unique model aspects could move to supplement to shorten main text.
7. Lines 341-342: Was stabilization achieved in all the runs?
8. Section 2.4: mention here that IRF is also calculated (line 797 points to IRF kernels but only SARF kernels are introduced)
9. Figure 6: Is the tropopause used for the observations consistent with the definition in the models? Clarify that the TOAR value is from this study.
10. Sections 3.3/3.4/3.5 and 4.3/4.4: Add a final summary sentence or short paragraph to convey the key message(s) emerging from the findings reported in each section?
11. Lines 589-591: Given the different cloud fields, is this finding meaningful?
12. Line 653. GEOS-Chem reads in the cloud fields, so rather than a zero response it’s the same clouds being used as input.
13. Lines 720-722. So tropospheric ozone precursors and ODS (&N2O?) contribute equally to ERF in 2050 in this scenario?  Figure 12 might better illustrate this by using hatching as in Figure 8. Consider adding this point to the abstract.

Technical Corrections

1. On Figure 5, is the topography used to determine where to plot as a function of pressure the same in the different panels? Around 40S and 90N there are some differences.
2. Bodeker Scientific, 2024 is not in the bibliography