We thank Martin Schultz for initiating the discussion of our recently posted Opinion. In acknowledging helpful discussions, we did not mean to imply that all were supportive; indeed constructively critical comments are often the most helpful. Notwithstanding his curt opinion regarding “simple models” being “complete nonsense”, we would like to point out that scientists in other fields have expressed much more supportive views of the important roles played by models of varying complexity. In fact, in the field of geophysical fluid mechanics it is quite common for researchers to rely on a variety of numerical models of differing complexity and representation to investigate the manifold, non-linear features of atmospheric motion.  Examples from the literature include:  

Held (2005) quite clearly states the need for model hierarchies in reference to climate modeling: “On the one hand, we try to simulate by capturing as much of the dynamics as we can in comprehensive numerical models. On the other hand, we try to understand by simplifying and capturing the essence of a phenomenon in idealized models, or even with qualitative pictures.”

Held (2014) again emphasizes this point: “The models used to simulate the climate are themselves complex, chaotic dynamical systems. To work with them effectively requires not only the careful examination of alternative formulations of these comprehensive models but also the construction of a hierarchy of models in which elements of complexity are added sequentially.”

In a similar vein, Emanuel (2020) argues: “As it becomes easier to undertake complex computer simulations of climate and weather, and as large volumes of satellite data become more available, it is tempting to use computers to simulate, rather than understand, nature. … simulation without understanding imperils scientific progress and, paradoxically, may impede the development of better models.”

We believe great advantage lies in emulating these other fields when it comes to studying the complex physical and chemical behavior of ozone in Earth’s atmosphere. 

References

Emanuel, K. The Relevance of Theory for Contemporary Research in Atmospheres, Oceans, and Climate. AGU Advances, 1, e2019AV000129, doi: https://doi.org/10.1029/2019AV000129 (2020).

Held, I. M. The gap between simulation and understanding in climate modeling. Bull. Am. Meteorol. Soc., 86, 1609-1614, doi:10.1175/Bams-86-11-1609 (2005).

Held, I. Simplicity Amid Complexity. Science, 343, 1206-1207 (2014).

We thank Referee #1 for carefully reading our Opinion and for their positive comments and recommendation (https://doi.org/10.5194/egusphere-2023-426-RC1).

“Development of a “model” of the underpinning science for tropospheric ozone” likely will indeed be challenging, as evidenced by the comment posted by the TOAR Steering Committee (https://doi.org/10.5194/egusphere-2023-426-CC4). In our response to that comment (https://doi.org/10.5194/egusphere-2023-426-AC5) we more fully describe the “model” that we envision; it will necessarily comprise a hierarchy of models of varying complexity. This issue will be more fully described in our revised manuscript.

The suggested correction of “International Panel on Climate Change” to “Intergovernmental Panel on Climate Change” has been made in our revised manuscript.

We thank Referee #2 for carefully reading our Opinion and for posting their thoughtful comments. Below the Referee’s original comments are reproduced in bold with our responses following in plain text. 

Major issues:

1) Do we not already have a “model” of tropospheric ozone?

A key premiss of this paper is that as a community we lack a parsimonious model that can describe the processes that control tropospheric ozone. Although the level of simplicity is arguable, at least in my mind we have such a parsimonious model. Indeed, Figure 3 in the paper outlines such a model and this model has been the de facto model used within the community since at least the mid 2000s. In which case, what new insight is this Opinion piece adding?

At its simplest we can say that the model of climate change is a question of forcing and feedback:

            ∆N = ∆F – α∆T                                                                                               Equation 1

The model of tropospheric ozone can also be written very simply (where ∇ is used to represent transport, P(O3) the production and L the first-order loss rate of ozone):

            𝑃(O₃) = (𝐿 + ∇)[O₃]                                                                                       Equation 2

However, these simple models are not practically useful. Complex problems require complex models. There is a good point to be made that the level of complexity of our model (Figure 3) is not fit for purpose but it’s not clear how we as a community go about determining this. It seems to me, at least, that the model we have for tropospheric ozone (Figure 3) is fine. The main problem is the problem of who owns the challenge of tropospheric ozone (the air quality community or the climate community) and so who are we simplifying the model (Figure 3) for; this is an issue that is intimately linked with the choice of metric.

Thank you for this perspective. We agree that “Complex problems require complex models.”; however we strongly believe that complex problems can only be understood through simplified models. As we argue in our response (https://doi.org/10.5194/egusphere-2023-426-AC1) to the comment of Martin Schultz (https://doi.org/10.5194/egusphere-2023-426-CC1) a model hierarchy is essential – complex models are required to simulate the temporal and spatial distribution of tropospheric ozone, but we must aim “to understand the drivers of that distribution by simplifying and capturing the essence of (the tropospheric ozone) phenomenon in idealized models” (Held, 2005).

In our revised manuscript we have expanded and clarified the description of the model hierarchy that we believe is required. The single equations presented above for climate change and tropospheric ozone can indeed be considered “models”, but they are not elaborate enough to provide intuitive, quantitative descriptions (i.e., Held’s goal of “understanding”) of the respective scientific issues. For climate change, perhaps the simplest such “model” that can provide important understanding is that developed by Arrhenius (e.g., see accessible discussion by Rodhe et al., 1997), which intuitively described the basic elements of the greenhouse effect and gave the quantitative estimate that a doubling of the atmospheric concentration of CO₂ would increase the global mean temperature by 5.7°C. Since that estimate was presented in 1896, this relatively simple model has been greatly refined and expanded into a hierarchy of models that illuminate the climate change issue from a wide range of perspectives. Over the decades, the climate change community has effectively determined the range of complexity of the model hierarchy needed to simulate and understand that issue.

The central point of our Opinion is that neither the air quality community nor the climate community “owns the challenge of tropospheric ozone”. Both of these communities have simulated aspects of the tropospheric ozone distribution, but neither community has developed a comprehensive yet clear and intuitive understanding of the issue. We are arguing that a separate tropospheric ozone community is required (which will undoubtedly include members of both the air quality and climate communities).

Working Group 1 of IPCC provides a useful template that could be followed in developing a conceptual ‘model’ of tropospheric ozone. For the AR6 Synthesis Report, they provide 1) a 13-chapter comprehensive Report, which gives a complete and clear description of the science comprising our current understanding of the climate system, 2) a Technical Summary that synthesizes the key findings of the Report, and 3) a Summary for Policy Makers. The Technical Summary serves as a bridge between the chapters of the full Report and the Summary for Policy Makers. These three formal reports are supplemented with additional, more accessible informative materials: a collection of Frequently Asked Questions, a series of entry point Fact Sheets; and an Interactive Atlas. In essence, it is the Technical Summary, combined with the supplementary materials that constitutes a conceptual ‘model’ for the climate system; we believe that a similar combination of materials is required, and can be developed for tropospheric ozone.

2) Do we have a process for deciding which metrics for tropospheric ozone are policy relevant?

Section 7, I think, is a key section for this Opinion piece. The authors outline some of the metrics used in the climate science and stratospheric ozone communities (GWP and ODP) and some of those used in the tropospheric chemistry community (OFP and POCP) but the authors don’t go on to highlight the problems with the GWP and ODP metrics. A discussion on the problems with these metrics would be helpful as that would help underscore the need for a process to develop the optimal policy relevant metric(s) for tropospheric ozone. See for example, Lynch et al. (2020) and Pyle et al. (2022).

And, thank you for this perspective. We agree that a robust process will be needed to develop the optimal policy relevant metric(s) for tropospheric ozone. We have suggested that emission inventories can serve this purpose, but other choices could possibly be developed as part of our proposed science-into-policy assessment process. And, that process may indeed confront problems. We have added discussion of these issues to our revised manuscript, and cite the two references suggested by the reviewer.

The discussion about the UN FCCC is important (not necessarily interesting) but the UN FCCC deals with emitted species only, as these emissions can be regulated. Should the UN FCCC also consider OH as one of the gases it “controls”? Tropospheric ozone cannot be part of emission based policy metrics because it is not an emitted species. The UN FCCC does include methane and a significant fraction of the methane GWP comes from the impacts that methane has on tropospheric ozone. If tropospheric ozone were to come under the remit of UN FCCC then the fraction of GWP that is attributable to tropospheric ozone formation from methane would have to be removed. This would create a huge issue in terms of recent work that targets methane mitigation as a priority as the GWP-100 of methane would drop by about 1⁄4. Again, a discussion of the impacts of the choice of policy metric would really help the community rally around a process to identify the right one(s).

These are interesting points. In answer to the first (rhetorical?) question, we do not believe that “the UN FCCC also consider OH as one of the gases it “controls”. However, we do think that emission based policy metrics for tropospheric ozone are not only possible (because the portion of tropospheric ozone that is under anthropogenic control directly depends upon emitted species), but also necessary since emission controls are the only practical policy tool available. We do not understand the issue that the reviewer suggests would arise from apportioning environmental improvement related to methane mitigation; there would certainly be co-benefits for both climate and air quality, but we see no need to apportion those co-benefits between the two issues.

Figure 2 highlights the alarming issue we have with metrics for tropospheric ozone. By my counting there are at least 4 different metrics being displayed. I think that an Opinion piece such as this should touch on this important aspect and draw on the literature which has discussed the choice of metrics at length. Through analysis of this literature it rapidly becomes evident that part of the problem with creating a “simple” model for tropospheric ozone is that the stakeholders for the impacts of tropospheric ozone are diverse and each want different things. A key and related aspect is which policy makers are the metrics being targeted at? Policy is a wide ranging world and many different tropospheric ozone metrics could be identified for different policy issues. This relates to my point about who owns the challenge of tropospheric ozone above.

We agree that the proliferation of metrics for tropospheric ozone is confusing; the TOAR community has addressed tropospheric ozone metrics in great detail (Lefohn et al., 2018).  As we note in our response to the first paragraph in this comment, we recognize the issue the reviewer raises, and have added some discussion of it in our revised manuscript, but we are not prepared to attempt further recommendations. 

Minor points:

Thank you for these editorial suggestions; all issues have been corrected.

L94: I suggest you delete the word “Interestingly” and let the reader make up their mind.

Suggested change made.

L115: The heading seems incomplete or at least it does to me. Delete “the” or add more words.

Suggested change made; “the” deleted.

L129: I’m sure there are others but with my UK-centric hat on I would suggest you add AQEG to this list who have done fantastic work on tropospheric ozone for decades.

Suggested change made; “AQEG” added to the list of organizations.

L183: See major comments above.

Our discussion of the required “model” has been improved, as discussed in the response to the major comment above.

Figure 4: Methane emissions should top out at about 500 Tg/yr. Please check panel (a). The use of NMVOC and AVOC is confusing. Can you be consistent and define what you mean here. Also, please check the units for panels (b)-(e). Should there not be an area dimension?

We have revised Figure 4 to correct the problems identified; thank you for the close attention. Panel (a) has been completely revised with all quantities reviewed and corrected where needed. Correct units with an area unit have been supplied for panels (b)-(e). The term AVOC has been removed and NMVOC is defined.

L240: Fragment. Re-word.

This sentence has been re-worded.

L242: Replace the comma with a semi-colon or re-phrase the sentence here.

The sentence has been re-phrased.

L255: Add “e.g.,” to the reference as this was not the first study to point this out.

This addition has been made.

L260&266: What do the authors mean by “ozone air quality” and “air quality for ozone”?

We have changed both terms to “ozone air quality”. In the introduction, we note that “tropospheric ozone is widely recognized as an important air pollutant”, so we believe the meaning of “ozone air quality” is now clear in the manuscript.

Additional References:

Lefohn, A.S., et al.: Tropospheric ozone assessment report: Global ozone metrics for climate change, human health, and crop/ecosystem research. Elem Sci Anth, 6: 28. DOI: https://doi.org/10.1525/elementa.279 (2018).

Rodhe, H., Charlson, R.,  and Crawford, E.: Svante Arrhenius and the Greenhouse Effect. Ambio , 26, 2-5 (1997).

We thank Worden et al. for their interest in our Opinion manuscript, and for providing the perspective of the TOAR Steering Committee (https://doi.org/10.5194/egusphere-2023-426-CC4). Their comments identify areas where our manuscript must be clarified and highlight important and productive disagreements within the tropospheric ozone research community. The following discussion of those disagreements further emphasizes the critical need for our proposed science-into-policy assessment.

Below the original comments of  Worden et al. are reproduced in bold text with our responses following in plain text. 

The following is a joint statement from current and former members of the Tropospheric Ozone Assessment Report (TOAR) Steering Committee.  The submitted manuscript proposes a science-into-policy process that would mis-interpret the findings from TOAR, and therefore we feel compelled to state our concerns regarding the scientific structure of the proposal.

Given the tremendous effort invested in TOAR, we concur that both our Opinion and the proposed science-into-policy process for tropospheric ozone assessment must correctly interpret the TOAR findings. Importantly, however, it is critical to not assume that our scientific understanding of tropospheric ozone is complete with the TOAR process – rather, that process gives renewed impetus to further improvements in our scientific understanding of tropospheric ozone. 

- The submitted manuscript makes no mention of IPCC’s well-known assessment of the co-benefits of greenhouse gas mitigation for air quality improvements, a concept that has been widely discussed by the atmospheric sciences community and by policy-makers for at least 10 years (e.g. see West et al., 2013; 391 citations according to Web of Science).  As summarized by the recent Synthesis Report of IPCC AR6 (https://report.ipcc.ch/ar6syr/pdf/IPCC_AR6_SYR_LongerReport.pdf), existing and new policies to reduce greenhouse gas emissions will have the co-benefit of reducing ozone at the surface and in the free troposphere, especially due to methane mitigation (see also IPCC AR6 WG-III).  This omission of IPCC findings is profound, and seriously undermines the suggested science to policy process. The authors also fail to discuss the inclusion of tropospheric ozone as a risk factor in recent Global Burden of Disease reports, which have brought tropospheric ozone into the public health community discourse (Murray et al. Lancet. 2020; 396: 1223-1249).



Thank you for this perspective. We completely agree that the assessment process for tropospheric ozone must be conducted so as to fully integrate the relevant findings of both IPCC global climate change and WMO on stratospheric ozone layer depletion. Both of these science-into-policy processes have elucidated atmospheric processes that are central to tropospheric ozone; the assessment we propose must build on these findings. We do mention the health effects of tropospheric ozone and point to the Global Burden of Disease through our reference to Cohen et al. (2017), but we do agree that this issue deserves more emphasis. In our revised manuscript, we will further emphasize the need for close integration with the IPCC and WMO assessments and provide added emphasis on the health impacts with appropriate references as you recommended.

- The submitted manuscript calls for the development of a single ozone policy metric, “with full buy-in from the atmospheric science community”.  TOAR is a grassroots organization sustained by the atmospheric science community, and TOAR’s great success is due to its popular and necessary use of multiple ozone metrics (for climate, health and vegetation impacts).  Tropospheric ozone chemistry is extremely complicated, concentrations of ozone vary in space and on hourly timescales, and no single ozone metric can adequately gauge its impacts on diverse biological systems, or climate.  The suggestion for a single ozone policy metric would not provide protection for the different receptors damaged by ozone which have very different exposure patterns, dose-response curves, and ozone damage thresholds.  Rather than a new metric which would necessitate the development of a new set of exposure-response curves, ozone policy could be guided by the more consistent use of existing response curves to convert exposure or dose to easily understood impacts such as years of life lost (YLLs), years of life lived with disability (YLDs), and disability-adjusted life-years (DALYs), crop production losses (CPL) and economic cost losses (ECL).

Thank you for identifying this point of confusion. We see how the term “metric” as applied to tropospheric ozone covers a rather broad range of quantitative measures of ozone impact. Worden et al. cite examples of metrics that quantify the impacts of ozone on biological systems. Our suggested metric is a policy-relevant one, which can serve to quantify the effects of ozone air quality policies on the ambient ozone distribution, much as GWP serves to quantify the impact of climate policies on the Earth’s climate. Just as GWP does not quantify the impacts of changing climate on biological systems, we cannot expect the ozone policy-relevant metric to quantify ozone impacts on biological systems. Further, the multiple ozone metrics used by TOAR will remain unaffected, as they are quantified by integrating appropriate dose-response curves over the observed or simulated ozone distribution; development of the proposed policy-relevant metric will not affect such integrations. In our revised manuscript, we will clarify this distinction between the proposed policy-relevant metric, and current metrics that quantify impacts on particular biological systems.

- A foundation of this proposal is the authors’ repeated claim that mid-latitude baseline ozone doubled from the 1950s to the early 2000s, but has since been steadily decreasing.  This claim runs contrary to the findings of IPCC AR6 and other recent assessments of tropospheric ozone trends, including the analyses from TOAR (collectively cited over 1300 times), which do not support a steady decrease in tropospheric ozone across the mid-latitudes in recent decades (further details are provided below).  This basic scientific error prevents us from having any confidence in the scientific structure of the proposed science-into-policy process.

We agree that there is disagreement within the community on important aspects of the temporal and spatial distribution of tropospheric ozone, including long-term changes in baseline ozone. However, the analyses showing “mid-latitude baseline ozone doubled from the 1950s to the early 2000s, but has since been steadily decreasing” remains firmly established, not having been refuted by any later analyses; a detailed discussion of this issue is given below in the Supporting Information. Furthermore, we suspect that the TOAR findings cited above are heavily influenced by “confirmation bias”; agreement with model simulations has been a criterion when choosing between observational analyses that give conflicting findings. Moreover, TOAR has the stated aim to provide a reliable historical record of background ozone levels and reliable guidance to other assessments such as IPCC AR6. Community reliance on TOAR analyses is based on this stated aim without further evaluation of the historical record, so that reliance does not support the accuracy of the TOAR analyses. One effort of the assessment that we propose is a rigorous, objective evaluation of observational analyses free of such biases. In any event, we do not believe that citation count can be used as logical argumentation in the debate of an open scientific question. 

- These authors call for the development of a simple, conceptual ‘model’ that would be used to understand the output of atmospheric chemistry models, guide research efforts and inform policy. They describe the attributes of the “model”, which exactly match the attributes of a conceptual model that these same authors have proposed in a recent paper (Mims et al. 2022). While the authors do not cite their own work, we briefly discuss the weaknesses of the Mims et al. model below.  In our expert opinion, output from modern atmospheric chemistry models can be effectively summarized for policy-makers, and there is no reasonable application for a simple, conceptual model that lacks basic atmospheric dynamics and is therefore unable to capture the temporal and spatial variability in column and ground level ozone, let alone allow for any attribution of ozone changes to driving forces.  There may be important roles for simple models, but new models must be vetted among the community of scientists and demonstrate their value before they are used in a science-to-policy process. 

Thank you for this comment. As we state in our manuscript: “Such an intuitive model would be an essential component of a modeling hierarchy (Held, 2005) by complementing the comprehensive numerical models that aim to simulate in full detail as much of the atmospheric chemistry and dynamics as possible.” That hierarchy would necessarily include the modern atmospheric chemistry models mentioned by Worden et al., as well as facilitating our understanding of the chemistry and dynamics “by simplifying and capturing the essence of a phenomenon in idealized models, or even with qualitative pictures”, as emphasized by Held (2005) in his distinction between ‘simulation’ and ‘understanding’. It is clear that in our revised manuscript we must more clearly describe the needed model hierarchy, which would consist of a wide spectrum of idealized models of varying complexity.

The specific criticisms by Worden et al. of the model presented by Mims et al. (2022) are discussed below in detail in the Supporting Information.

While we agree that science must inform policy, we have no confidence in this particular proposal for a science-into-policy process, which seems to oversimplify the science and relevant metrics, while misinterpreting the science.  TOAR follows the lead of other influential scientific processes like IPCC (which focuses on the science and summarizing that science for policymakers), to inform choices without prescribing policy.  TOAR does so in part by including studies of impacts on health, crops, vegetation, and climate.  TOAR will continue to work with IPCC, the Climate and Clean Air Coalition (CCAC, www.ccacoalition.org) and the Task Force on Hemispheric Transport of Air Pollution (TF HTAP) under the UNECE, as well as established regional organizations (for example, EMEP in Europe), to advise policy-makers to develop more effective approaches. 

We strongly support the TOAR efforts that are outlined above, and as we note in our Opinion, the assessment we propose could be built on the current TOAR and HTAP activities.

In summary, to our minds further progress in understanding the temporal and spatial distribution of tropospheric ozone must include a complete review of existing observational analyses as part of the science-into-policy assessment that we propose, a review guided by Richard Feynman’s sage advice on the imperative of doubting experts; viz. “Science is the belief in the ignorance of experts.”

---------------------------------------------------------------------------------

Supporting information:

Baseline ozone trends:

As assessed by IPCC AR6 WG-I (Chapters 2 and 6), the annual State of the Climate Reports, the Tropospheric Ozone Assessment Report (Tarasick and Galbally et al., 2019), CMIP6 and the UNEP Scientific Assessment of Ozone Depletion 2022 (Chapter 3.3) the tropospheric ozone burden has continued to increase since the 1990s including at mid-latitudes; these same assessments found no convincing evidence that mid-latitude baseline ozone doubled from the 1950s to the early 2000s.  These findings are corroborated by very recent studies published since the release of IPCC AR6 (Miyazaki et al., 2020; Christensen et al., 2022; Wang et al., 2022; Fiore et al., 2022; Chang et al., 2022).  Contrary to the evidence, the authors of the submitted manuscript have claimed (since at least 2017) that baseline ozone has been consistently decreasing across northern mid-latitudes over the past two decades. No independent study has been able to corroborate their claims, and these claims were not accepted by the assessment reports listed above.

The above described unfortunate situation originated from the Tarasick and Galbally et al., (2019) analysis of changes in surface ozone at northern temperate latitudes. That analysis suffered from multiple biases (Parrish et al., 2021a) that caused large underestimates in the mid-latitude baseline ozone increase from the 1950s to the early 2000s. Due to the stated aims of TOAR, the later IPCC, CMIP6 and UNEP assessments accepted the TOAR results without further observational analysis. The five later publications cited above do not present observational analysis before 1980, and thus do not address this issue. To our knowledge, the biases in the Tarasick and Galbally et al., (2019) analysis have not been addressed by TOAR, and the underestimates are still presented as accurate and repeatedly cited.

Also to our knowledge, all published observational-based analyses of baseline tropospheric ozone changes at northern midlatitudes are consistent with continuing decreases since the mid-2000s. This decrease is evident in the baseline-representative data sets (European alpine, Mace Head and U.S. National Park data) plotted in Figure 2 of our submitted manuscript under discussion here. Recent published observational analyses are generally based on decadal scale, linear trend analysis (e.g., Gaudel et al., 2020) that is insensitive to the important non-linear character of baseline ozone changes that have occurred at northern midlatitudes - an early rapid increase, a broad peak reached in the mid-2000s (approximately consistent with the turnover in midlatitude anthropogenic NO_xemissions), followed by a slow recent decrease. Nevertheless the linear changes quantified over periods beginning in the 1990s are quantitatively consistent with the derived overall, nonlinear change over the same period as verified in detail by Parrish et al. (2021b).

We do understand that the TOAR team disagrees with these analyses; however, the argumentation presented in Parrish et al. (2021a;b) has not been refuted. Continued deliberations on this important topic must objectively consider these issues; encouraging these deliberations is one goal of the submission of our Opinion manuscript. 

Simple conceptual ‘model’

As stated above, the submitted manuscript calls for the development of a simple, conceptual ‘model’ that would be used to understand the output of atmospheric chemistry models, guide research efforts and inform policy. They describe the attributes of the “model”, which exactly match the attributes of a conceptual model that these same authors have proposed in a recent paper (Mims et al. 2022). Even though the authors do not cite their own work, we briefly comment on this paper in order to point out the substantial shortcomings of a simple conceptual model.  This particular conceptual model is similar to a simple 1970s box model that scientists had to build in the days before adequate computing power was available to run more complex models (e.g. Oeschger et al., 1975; Thompson and Schneider, 1979).  It has no atmospheric dynamics and it assumes the mid-latitudes are isolated from the polar regions and the tropics; this is contrary to recent work, which shows that tropospheric ozone in the mid-latitudes is impacted by emissions and transport from the tropics, and this influence cannot be ignored (Zhang et al., 2016,2021; Gaudel et al., 2020).  In contrast, modern atmospheric chemistry models can handle global and regional atmospheric dynamics, in addition to emissions and photochemistry.  These models correctly reproduce the observed increase of the tropospheric ozone burden, and as shown by IPCC AR6 the output from these models can be effectively summarized to provide the answers to the questions from policy makers (see Chapters 6 and 7, and Box TS.7 in the Technical summary of AR6). 

A common aphorism in statistics is particularly apropos here: "All models are wrong, but some are useful", or put another way by oncologist Howard Skipper “A model is a lie that helps you see the truth.” Modern atmospheric chemistry models and the conceptual model presented by Mims et al. (2022) are both “wrong” in the sense that neither can faithfully simulate all aspects of the tropospheric ozone distribution, but both are “useful” because each can answer important questions regarding that distribution. It is only through a hierarchy of models that we can both simulate and understand tropospheric ozone. It is this type of interdependent model hierarchy that we are proposing.

Mims et al. (2022) used a minimal set of parameters with values taken from generally accepted and measured ozone behavior to describe the ozone sources, sinks and northern midlatitude zonal flow and mixing. This intentionally rudimentary model was designed to determine the drivers of 1) the vertical gradient in baseline ozone between the marine boundary layer and the free troposphere, and 2) the differing seasonal cycle of baseline ozone in these layers. The model quantitatively simulated these features, plus it reproduced the observed, nearly uniform free troposphere that behaves as an ozone reservoir, responding to the combined boundary layer and stratospheric inputs. The critical role of the marine boundary layer in the global ozone balance and the constraints that it places on the net continental production are clearly revealed. Sensitivity analysis identified which of the basic set of process parameters most require better understanding.

Minimalist models such as that of Mims et al. (2022) cannot stand alone, but when used in conjunction with detailed modern atmospheric chemistry model simulations, can provide the basis for a comprehensive understanding of tropospheric ozone.

From a more technical perspective, the omission of net meridional fluxes of tropospheric ozone in the model of Mims et al. (2022) is not indicative of the model having “no atmospheric dynamics” as suggested by Worden et al., but rather the simplified model employs the approximation that the influx from lower latitudes in one part of the globe (as discussed in Gaudel et al., 2020) is compensated by efflux elsewhere. This approximation is supported by results established using reanalysis data by Miyazaki et al. (2005). The model of Mims et al. (2022) does explicitly contain zonal advection throughout the domain, thus representing the most salient feature of atmospheric dynamics, not lacking basic atmospheric dynamics as charged by Worden et al.. 

Further, the transport of tropical ozone precursors into the midlatitudes cited in the comment (Zhang et al., 2016; 2020; Gaudel et al., 2020) is invoked to explain the authors’ conclusion that despite the recent reduction in midlatitude NO_x emissions, their analysis indicates a constant or increasing background ozone. This point is exactly relevant to the one on which we and the authors of the comment disagree. We suggest that the consideration of a simplified model could have been quite instructive in their assessment of this crucial point, investigating, for example, whether those tropical ozone precursors produce ozone that would be long-lived enough to be advected across the entirety of the midlatitudes (not just over N. America), and whether there may be a compensatory flow elsewhere into the midlatitudes that might countervail against this effect. In short, we are arguing that uncritically accepting the output of one complicated CTM simulation as the basis of an entire hypothesis is not an adequate analysis, and that the judicious use of simplified models in conjunction with “full-blown” CTM’s can more effectively advance our understanding than simply comparing complex model output with observations (Emanuel, 2020).

Additional References

Emanuel, Kerry. “The relevance of theory for contemporary research in atmospheres, oceans, and climate.” AGU Advances 1, no. 2 (2020): e2019AV000129.

Parrish, D.D., R.G. Derwent, and J. Staehelin (2021a), Long-term changes in northern mid-latitude tropospheric ozone concentrations: Synthesis of two recent analyses, Atmos. Environ., 248, https://doi.org/10.1016/j.atmosenv.2021.118227.

Parrish, D.D., R.G. Derwent & I.C. Faloona (2021b), Long-term baseline ozone changes in the Western US: A synthesis of analyses, Journal of the Air & Waste Management Association, 71:11, 1397-1406, DOI: 10.1080/10962247.2021.1945706.