Technical note: Northern midlatitude baseline ozone – long-term changes and the COVID-19 impact

Parrish, David D.; Derwent, Richard G.; Faloona, Ian C.; Mims, Charles A.

doi:https://doi.org/10.5194/acp-22-13423-2022

Articles | Volume 22, issue 20

https://doi.org/10.5194/acp-22-13423-2022

© Author(s) 2022. This work is distributed under
the Creative Commons Attribution 4.0 License.

https://doi.org/10.5194/acp-22-13423-2022

© Author(s) 2022. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 22, issue 20

Technical note

|

18 Oct 2022

Technical note |

| 18 Oct 2022

Technical note: Northern midlatitude baseline ozone – long-term changes and the COVID-19 impact

David D. Parrish, Richard G. Derwent, Ian C. Faloona, and Charles A. Mims

Download

Final revised paper (published on 18 Oct 2022)
Supplement to the final revised paper
Preprint (discussion started on 16 Jun 2022)
Supplement to the preprint

Interactive discussion

Status: closed

RC1:
'Misleading paper with major flaws. Needs fundamental revisions and more years with data.', Anonymous Referee #1, 21 Jun 2022
The manuscript tries to address how unusual the free tropospheric ozone anomalies observed in 2020 during and after the COVID related emission reductions were in a context of longer term ozone trends. The underlying big question is which "normal" ozone would have been expected for 2020, without COVID related emission reductions. Essentially, the manuscript claims that their parabolic trend used to describe "normal background" ozone in a previous publication (and based on data from 1979 to 2018 only, Parrish et al., 2020), would have continued in the same form throughout 2020 and would have resulted in ozone levels similar to the observed low ozone of 2020. If this claim were true, there would have been no COVID related ozone reductions in 2020 - in stark contrast to a number of observational and modelling studies (Steinbrecht et al., 2021; Christofanelli et al., 2021; Weber et al., 2020; Bouarar et al., 2021; Miyazaki et al., 2021).

I think the manuscript has major flaws, needs very fundamental revisions, and especially additional data, before it might become acceptable as an ACP paper.

The "conventional wisdom", that tropospheric background ozone showed a large increase from the 1960s until around 2000, but has been consistently decreasing since sometime after 2000 is held only by the authors themselves. In particular, the claim that their reported ozone decrease by about -4 ppbV per decade since about 2005 (Parrish et al. 2020) is significant and representatative, is in clear contrast to many other current studies, which generally indicate small and often non-significant mixed positive and negative trends with small magnitudes (typically +-1 ppbV per decade or less, e.g. Cooper at al., 2021; Chang et al., 2022).

The authors' parabolic trend is the only estimate that results in very low expected "background" ozone in 2020. Almost all other authors / studies have used a constant climatology, or a linear trend to estimate "background" ozone in 2020. These more conservative estimates provide substantially higher "background" ozone for 2020, and they all point to unusually low tropospheric ozone in 2020 (with the explanations provided by e.g. Weber et al., 2020; Bouarar et al., 2021; Miyazaki et al., 2021).

The authors' parabolic trend fit has no degree of freedom that would allow a different behaviour of long-term ozone changes before the maximum around 2005 and after the maximum, since 2005. Essentially the authors are assuming that since about 2005 ozone MUST be going down in the same way, as it has been going up before 2005. Clearly this is a very strong assumption, and completely ignores the very different economic and societal circumstances that have been driving the observed very large ozone increases from the 1960s to about 2000, and are now driving small possible ozone changes since 2005 (with regional differences and many more complications, e.g. Cooper at al., 2021; Chang et al., 2022).

The authors use no data after 2018. There is no constraint for "background" ozone just before 2020, and also no constraint for "background" ozone after the 2020 anomaly. Without data from these important additional years, the authors' claim that the 2020 ozone anomaly was not an anomaly but instead was normal, has no physical basis at all!!

I summarize my critique by providing two alternative versions of Fig. 1 of the manuscript (having digitized the data points).

My Fig. 1 is essentially the same as Fig. 1 of the Parrish et al. manuscript. It shows the 2 year average background ozone data (blue circles and line), along with three fits:

mean after 2000, cyan line;

linear fit using data after 1994, magenta line and confidence interval;

parabolic fit, black line with grey confidence interval (same as used in manuscript)

The 2020 anomaly observed by Steinbrecht et al., 2021 (brown square), and the parabolic "background" extrapolation to 2020 by Parrish et al. (green circle) are shown as well. As in Parrish et al., the fitted parabola here gives the same extrapolated green circle for 2020, which is close to the observed 2020 anomaly (brown square) of Steinbrecht et al. (2021). However, the (grey) 95% confidence interval derived here (by Monte-Carlo bootstrap) is wider than the green uncertainty bar given by Parrish et al. The confidence interval is also not symmetric around the extrapolated 2020 value, and reaches closer to zero. As in Parrish et al., the 2020 observed value (brown square) lies far below the mean since 2000, and far below the linear trend since 1994. It also lies at the bottom of the confidence interval of the extrapolated parabola (green circle).

My Fig. 2 shows the same data as Fig. 1, but now an additional "hypothetical" data point is added for 2019. I chose zero anomaly for this data point - inline with e.g. Fig. 4 of Steinbrecht et al. 2021, which shows slightly higher tropospheric ozone in 2019 compared to previous years. The addition of this one data point changes both the parabolic fit (black line, grey confidence interval), and the linear fit (magenta line and confidence interval). Now, the parabola predicts higher "background ozone" in 2020 than Fig. 1, and the observed anomaly (brown square) lies outside of the 95% confidence interval. The linear fit has changed very little. It still predicts "background ozone" close to zero for 2020. In addition, Fig. 2 has a cubic fit added to the 2 year anomaly data (red dashed lines). Importantly, this fit has an additional degree of freedom, which allows for different trends before and after 2005. This cubic fit also predicts "background ozone" close to zero for 2020 (but with large uncertainties for values after 2015 and before 1985, reflected in the wide confidence interval). So all background estimations, with the notable exception of the Parrish et al. parabola, give close to zero "background ozone" in 2020, and indicate a large negative observed anomaly for 2020 - consistent with many other studies, as mentioned above.

Hopefully, my two Figures demonstrate clearly the very problematic use of the Parrish et al. parabolic fit for an extrapolation of "background ozone" to 2020. Given this and other important flaws, I feel that the manuscript is not acceptable as an ACP paper. In fact it is quite misleading, and should be definitely by rejected in anything resembling its current content. (This was the case for a previous version of the manuscript, which was rejected by Geophysical Research Letters). I suggest that the authors wait for a number of additional years of data, including 2022 and 2023 (as 2021 may still be affected by ongoing COVID related emission reductions, for example due to still reduced air traffic), and then redo their analysis. I also strongly suggest to use a trend estimator that allows different trends before and after the years around 2005, and to better consider the large uncertainties of trend estimators, e.g. for the year 2020.

Without new data and new information, the present manuscript is just a rehash of Parrish et al. (2020). It does not report "substantial new results and conclusions", and does not provide the "substantial advances and general implications for the scientific understanding", which are required for an ACP research article.

Figures:

Figure 1, same as Fig. 1 of the Parrish et al. manuscript: 2 year average baseline / background ozone, along with three fits to estimate background ozone after 2000: Mean since 2000 (here cyan line), linear trend since 1994 (magenta line and 95% confidence interval), parabola (black line and grey 95% confidence interval). The green circle gives the value of the parabola extrapolated to 2020. The brown square gives the observed 2020 anomaly from Steinbrecht et al. (2021).

Figure 2: Same as Fig. 1, but including an additional (hypothetical) data point in 2019 (black circle), and including also a cubic fit (dashed red lines and 95% confidence interval).

References:

Bouarar, I., et al. (2021). Ozone anomalies in the free troposphere during the COVID-19 pandemic. Geophys. Res. Lett., 48, e2021GL094204. https://doi.org/10.1029/2021GL094204

Chang, K.-L. et al. (2022). Impact of the COVID-19 Economic Downturn on Tropospheric Ozone Trends: An Uncertainty Weighted Data Synthesis for Quantifying Regional Anomalies Above Western North America and Europe. AGU Advances, 3, e2021AV000542. https://doi.org/10.1029/2021AV000542

Cooper, O.R., et al. (2020). Multi-decadal surface ozone trends at globally distributed remote locations. Elementa Science of the Anthropocene, 8, 23. https://doi.org/10.1525/elementa.420

Cristofanelli, P., et al. (2021). Negative ozone anomalies at a high mountain site in northern Italy during 2020: a possible role of COVID-19 lockdowns? Environ. Res. Lett., 16, 074029. https://doi.org/10.1088/1748-9326/ac0b6a

Miyazaki, K., et al. (2021). Global tropospheric ozone responses to reduced NOx emissions linked to the COVID-19 worldwide lockdowns. Science Advances, 7(24). https://doi.org/10.1126/sciadv.abf7460

Parrish, D.D., et al. (2020). Zonal similarity

of long-term changes and seasonal cycles of baseline ozone at northern midlatitudes. J. Geophys. Res., 125, e2019JD031908. https://doi.org/10.1029/2019JD031908

Steinbrecht, W., et al. (2021). COVID-19 crisis reduces free tropospheric ozone across the Northern Hemisphere. Geophys. Res. Lett., 48, e2020GL091987. https://doi.org/10.1029/2020GL091987

Weber, J., et al. (2020). Minimal climate impacts from short-lived climate forcers following emission reductions related to the COVID-19 pandemic. Geophys. Res. Lett., 47, e2020GL090326. https://doi.org/10.1029/2020GL090326
Citation: https://doi.org/10.5194/acp-2022-424-RC1
- AC1:
  'Reply to Comment of Anonymous Referee #1: acp-2022-424-RC1', David Parrish, 22 Jun 2022
  David D. Parrish, Richard G. Derwent, Ian C. Faloona, and Charles A. Mims
  We thank the referee for providing this perspective on our submitted paper. In order to encourage additional discussion, we briefly respond to specific comments in this review before developing a final, comprehensive response. The referee asserts that the paper has major flaws. However, since we consider only published data and analysis results, and that consideration is based on a very simple mathematical model, we do not believe that any major flaws are present. We believe that the crux of the issue is simply differing interpretations of those published results.
  Here we address the referee’s four assertions (reproduced below as bullets in plain text) of major flaws in our paper by briefly summarizing our reasoning (italic text) for believing that the interpretation presented in our paper is indeed correct:
  The "conventional wisdom", that tropospheric background ozone showed a large increase from the 1960s until around 2000, but has been consistently decreasing since sometime after 2000 is held only by the authors themselves. In particular, the claim that their reported ozone decrease by about -4 ppbV per decade since about 2005 (Parrish et al. 2020) is significant and representatative, is in clear contrast to many other current studies, which generally indicate small and often non-significant mixed positive and negative trends with small magnitudes (typically +-1 ppbV per decade or less, e.g. Cooper at al., 2021; Chang et al., 2022).
  
  The reviewer misinterprets the results reported in the cited references.
  First, Parrish et al. (2020) report an average trend of −0.9 ± 0.8 ppb decade⁻¹ from 2000-2018. Our quadratic fit over the 2005-2018 period corresponds (see Equation 2) to an average trend of −2.3 ppb decade⁻¹. These are significant, but smaller, than the negative trend mistakenly quoted by the referee.
  Second, Cooper et al. (2021) and Chang et al. (2022) are two of the references to which we refer as the “Linear Trend View”. As we discuss in our manuscript, these papers do not attempt to analyze baseline ozone trends “since about 2005”; they only quantify average trends beginning about a decade earlier, and thus include periods of increasing and decreasing baseline ozone, which therefore do give only “small and often non-significant mixed positive and negative trends”. In the manuscript we demonstrate our data and analysis results are consistent with those of Chang et al., (2022) (see Table S1 and Figure S1), Cooper et al. (2020) and Gaudel et al. (2021) (see Figure S2). As we also note on page 3 of the manuscript, “Parrish et al. (2021b) synthesized multiple published linear trend analyses of western U.S. baseline ozone, and showed that all results are consistent with an overall, non-linear change – a rapid increase (~5 ppb/decade) during the 1980s that slowed in the1990s, maximized in the mid-2000s, and was followed by a slow decrease (~1 ppb/decade) thereafter.” Many of the published linear trend analysis results considered were taken from Cooper et al. (2020) and Gaudel et al. (2021).
  The authors' parabolic trend is the only estimate that results in very low expected "background" ozone in 2020. Almost all other authors / studies have used a constant climatology, or a linear trend to estimate "background" ozone in 2020. These more conservative estimates provide substantially higher "background" ozone for 2020, and they all point to unusually low tropospheric ozone in 2020 (with the explanations provided by e.g. Weber et al., 2020; Bouarar et al., 2021; Miyazaki et al., 2021).
  
  The referee’s comment very nicely emphasizes the importance of our manuscript. The other, substantially higher "background" ozone estimates for 2020 were derived from analyses that neglect the non-linear character of long-term ozone changes at northern midlatitudes. Thus, they are not properly characterized as “more conservative”; they are better characterized as “misconceived”. We prepared and submitted our manuscript for the very reason of bringing this important disagreement to the attention of the Atmos. Chem. Phys. community.
  The authors' parabolic trend fit has no degree of freedom that would allow a different behaviour of long-term ozone changes before the maximum around 2005 and after the maximum, since 2005. Essentially the authors are assuming that since about 2005 ozone MUST be going down in the same way, as it has been going up before 2005. Clearly this is a very strong assumption, and completely ignores the very different economic and societal circumstances that have been driving the observed very large ozone increases from the 1960s to about 2000, and are now driving small possible ozone changes since 2005 (with regional differences and many more complications, e.g. Cooper at al., 2021; Chang et al., 2022).
  
  The referee’s assertion is incorrect due to misinterpretation of our analysis approach. We do not simply assume that a quadratic function is appropriate to describe long-term ozone changes. Rather, we perform a power series analysis of the observational data; that analysis indicates that only the terms through second order are statistically significant. If the post-2005 decrease were significantly different than the pre-2005 increase, then the third order (i.e., cubic) term of the power series would be significant. This power series analysis is quite flexible for fitting long-term changes of any functional form (see Parrish et al. (2019) for detailed discussion). For the data shown in Figure 1 of our manuscript, the coefficient of the cubic term is 0.5 ± 1.4 x 10^-4 ppb yr^-3 (95% confidence limit indicated). Thus, the quadratic fit does capture the statistically significant information regarding long-term ozone changes in the data plotted in Figure 1; inclusion of the cubic term in the fit makes only a negligible difference (+0.3 ppb) in our extrapolation of past ozone changes to 2020. Section S1 of the Supplement discusses this issue in greater detail, and Figure S2 shows that the post-2005 decrease is statistically consistent with the pre-2005 increase.
  Notably, at some future time, the cubic term must become significant, as the decreasing trend cannot continue indefinitely, since the quadratic fit would approach zero. If one were to interpret the cubic term given above as significantly positive, it would indicate that the rate of acceleration of the decreasing trend has already begun slowing.
  The authors use no data after 2018. There is no constraint for "background" ozone just before 2020, and also no constraint for "background" ozone after the 2020 anomaly. Without data from these important additional years, the authors' claim that the 2020 ozone anomaly was not an anomaly but instead was normal, has no physical basis at all!!
  
  Responding to this comment nicely summarizes our entire thesis, and demonstrates the important advantage that our analysis has over the other approaches that gave substantially higher background ozone estimates for 2020. We agree with the referee that it would be desirable if our analysis included observations from 2019 and after 2020. However, our non-linear long-term change analysis provides a strong constraint on background ozone in the years immediately preceding 2019; that constraint is much stronger than provided by the other analyses that rely only on either a long-term average climatology or a long-term linear trend derived over preceding decades without consideration of the non-linear character of the long-term change, which became increasingly pronounced in the years immediately preceding 2020. We also agree that we have no constraint for background ozone after the 2020 anomaly, but of course, the other studies suffer from that same lack of constraint.
  In summary, the analysis results we discuss have been thoroughly reviewed and published. They provide firm support for our conclusion, as quoted from the abstract:
  “… reported COVID-19 related ozone changes in the background troposphere based on the linear analysis are significantly larger than those derived considering recent long-term decreases in background ozone, which the linear trend analyses do not quantify. We further point out that the extensive loss of lower stratospheric ozone in the unprecedented 2020 springtime Arctic stratospheric ozone depletion event likely reduced the natural source to the troposphere rendering the background anomalously low that year. Consideration of these two issues indicates that the COVID-19 restrictions had a much smaller impact on background tropospheric ozone in 2020 than previously reported.”
  Additional Reference:
  Parrish, D.D., R.G. Derwent, S. O'Doherty, and P.G. Simmonds (2019), Flexible approach for quantifying average long-term changes and seasonal cycles of tropospheric trace species, Atmos. Meas. Tech., 12, 3383–3394, https://doi.org/10.5194/amt-12-3383-2019.
  
  Citation: https://doi.org/10.5194/acp-2022-424-AC1
  - RC2: 'Reply on AC1', Anonymous Referee #1, 24 Jun 2022
    
    I thank the author for his reply to my review. Unfortunately, I still disagree with the claim that the manuscript provides significant new scientific information, and advances our knowledge.
    Correctly, the author states "we consider only published data and analysis results". So there are no new data and there is no significant new information over what is already published in Parrish et al. (2020).
    Using old data and information, the authors come to the conclusion "that the COVID-19 restrictions had a much smaller impact on background tropospheric ozone in 2020 than previously reported". Based on old data and information only, this conclusion is in clear contradiction to a large number of more recent scientific publications, which show that
    the COVID-19 related lockdowns resulted in very significant emission reductions worldwide
    
    these emission reductions resulted in significant reductions of ozone in the free troposphere, as evidenced by studies based on observations, and by studies based on model simulations.
    
    A contribution from the 2020 Arctic ozone hole is also not new - this is mentioned already, e.g. in Steinbrecht et al. (2021), or Bouarar et al. (2021).
    
    Given this, I can only repeat my previous opinion that the manuscript "does not report substantial new results and conclusions, and does not provide the substantial advances and general implications for the scientific understanding", which would be required for an ACP research article.
    
    Citation: https://doi.org/10.5194/acp-2022-424-RC2
    
    AC2: 'Reply to 2nd Comment of Anonymous Referee #1: acp-2022-424-RC2', David Parrish, 25 Jun 2022
    
    David D. Parrish, Richard G. Derwent, Ian C. Faloona, and Charles A. Mims
    We thank the referee for continuing the discussion of our submitted paper. This 2^nd comment helps to clarify the difference between the interpretations presented in earlier publications and that in our manuscript.
    Here we briefly respond (in italics) to the initial general statement and the three bulleted points (copied in plain text) that the referee makes in that 2^nd comment:
    “So there are no new data and there is no significant new information over what is already published in Parrish et al. (2020).
    
    Using old data and information, the authors come to the conclusion "that the COVID-19 restrictions had a much smaller impact on background tropospheric ozone in 2020 than previously reported". Based on old data and information only, this conclusion is in clear contradiction to a large number of more recent scientific publications, …”
    
    Science is not merely the accumulation of new data. Indeed, science progresses through the continued reinterpretation of old and new data, and concerning the question at hand, namely the impacts of the COVID-19 lockdowns on tropospheric ozone, our submission represents a novel interpretation of the available data. We disagree that there is no "new information" in our submission as posited by the referee. Our considered analysis of the published data shows that while many have already published what the referee asserts is settled science, there is a significant body of evidence that indicates that these accepted conclusions may have been overstated because of two important points that we discuss in our submission. We urge the editors and reviewers to focus on the scientific validity of our arguments, and whether they correctly open the possibility of reinterpreting the conclusions of previously published studies. Logical argumentation based on precedent in the literature is not scientific; this is why Richard Feynman famously declared that "science is the belief in the ignorance of experts."
    The COVID-19 related lockdowns resulted in very significant emission reductions worldwide.
    
    We agree qualitatively with this point.
    These emission reductions resulted in significant reductions of ozone in the free troposphere, as evidenced by studies based on observations, and by studies based on model simulations.
    
    We agree only that observations do show that 2020 ozone in the free troposphere was lower than the 2000 to 2020 climatological mean, which Steinbrecht et al. (2021) chose as their reference. However, we demonstrate that Steinbrecht et al. (2021) overestimated the magnitude of the reduction due to the COVID-19 related emission reduction, because their chosen reference neglects the non-linear aspects of the long-term ozone changes that are ongoing throughout the northern midlatitude troposphere.
    A contribution from the 2020 Arctic ozone hole is also not new - this is mentioned already, e.g. in Steinbrecht et al. (2021), or Bouarar et al. (2021).
    
    A reduced contribution of STE to tropospheric ozone due to the record large 2020 Arctic ozone depletion is potentially a second reason that Steinbrecht et al. (2021) overestimate the magnitude of the ozone reduction. Although they do mention this issue, they do not include its quantitative impact even though their Figure 3 shows that large negative ozone anomalies were present in the lower stratosphere in 2020. Further, computer modeling very likely inadequately quantifies the influence of the Arctic ozone depletion. A review of global model sources of O₃ from the stratosphere to the troposphere by Young et al. (2018) found substantial spread among model estimates, and concludes quite stringently that "model results should be approached critically", which is exactly what we are calling on our peers reviewing this submission to do.
    In summary, there are substantial differences in the interpretation of the long-term changes in tropospheric ozone at northern mid-latitudes. Parrish et al. (2020) quantify significant non-linear behavior, with a substantial decrease since the mid-2000s. Neglect of this decrease led Steinbrecht et al. (2021) to overestimate the magnitude of the COVID-19 related impact.
    Notably, this is not simply an academic disagreement. As we discuss in our paper, the ongoing decrease in tropospheric ozone has substantial air quality implications, since baseline ozone has a major impact on surface urban and rural ozone concentrations. Thus, accurate characterization of these changes is important; our paper does indeed “provide the substantial advances and general implications for the scientific understanding, required for an ACP research article.”
    Finally, we emphasize a rather unique aspect of this discussion by quoting the final sentence from our abstract: “Analysis of baseline ozone measurements over several years following the COVID-19 impact is expected to provide a firm basis for resolving the inconsistencies between the two views of long-term northern midlatitude ozone changes and better quantifying the COVID-19 impact.” Thus, a resolution of the disagreement is in the offing.
    Additional Reference:
    Young, PJ, et al. 2018 Tropospheric Ozone Assessment Report: Assessment of global-scale model performance for global and regional ozone distributions, variability, and trends. Elem Sci Anth, 6: 10. DOI: https://doi.org/10.1525/elementa.265
    
    Citation: https://doi.org/10.5194/acp-2022-424-AC2
RC3:
'Comment on acp-2022-424', Anonymous Referee #2, 19 Aug 2022
General comments

This paper discusses the issues in detecting the long-term trends of northern midlatitude ozone and the related consequences in quantifying the COVID-19 impact on ozone levels in 2020. Due to long-term variations in emissions of ozone precursors, tropospheric ozone has been changing substantially since the 1950s. Temporal changes of ozone (in particular surface ozone) are found highly dependent on locations because of the differences in local/regional emissions as well as meteorological/geographical conditions. Therefore, it is impossible to obtain a consistent picture of long-term change of ozone for all sites and regions. Nevertheless, there have been efforts to establish a relatively consistent spatiotemporal variation of baseline ozone (meaning free of recent continental influences). Baseline ozone levels were found to have a high degree of zonal similarity at northern midlatitudes, and increased nonlinearly by a factor of 2 during 1950-2000 and began to decrease around the mid-2000s (Junge, 1962; Logan et al., 2012; Parrish et al., 2012; 2014). This understanding of baseline ozone is referred by the authors as the “Conventional Wisdom”. Least squares regression is a common way to quantify long-term trends of ozone concentrations. In view of the Conventional Wisdom, some studies (e.g., Logan et al., 2012; Parrish et al., 2012; 2014; 2017; 2020; 2021a; 2021b; Derwent et al., 2018; Derwent and Parrish, 2022) used quadratic functions in the regression, which well addressed the nonlinear long-term change of ozone. However, some recent studies (Gaudel et al., 2018; 2020; Tarasick et al., 2019; Cooper et al., 2020; Chang et al., 2022) disregarded the nonlinearity and estimated ozone trends using linear fits, obtaining much smaller positive or negative trends for varying periods. These recent studies are referred by the authors as the “Linear Trend View”. The inconsistent ozone trends between the Conventional Wisdom and the Linear Trend View are caused by different treatments of historic ozone measurements. It is controversial which trend detection approach is superior. However, some recent publications take the climatological means of ozone as references, report larger negative anomalies of ozone in 2020 at a high mountain site (Cristofanelli et al., 2021) and in the northern hemisphere free troposphere (Steinbrecht et al., 2021; Clark et al., 2021; Chang et al., 2022) and attribute the negative anomalies to the COVID-19 impact. This is the critical issue raised in this paper. The authors review the results from the related publications, evaluate the reported 2020 ozone anomalies in the context of linear and nonlinear ozone trends, and show that the 2020 anomalies are well within the uncertainty range of the estimated 2020 baseline ozone level (extrapolation of their quadratic fit). They argue that even without the COVID-19 impact, the expected level of baseline ozone in 2020 would be 3.2±1.3 ppb lower than the reference value in 2000 and conclude that the COVID-19 impact on baseline ozone in 2020 was only -1.2±1.3 ppb estimated from the Conventional Wisdom instead of -4 ppb (Steinbrecht et al., 2021) or -3.7 ppb (Chang et al., 2022) from the Linear Trend View. The authors claim that the Conventional Wisdom estimate combined with the influence of the reported 2020 Arctic ozone depletion is sufficient to explain all of the 2020 ozone decrease without any impact from COVID-19 emission reductions. They also point out that a clear resolution of the inconsistencies between the Conventional Wisdom and the Linear Trend View is important for designing air quality improvement strategies in earlier developing economies and they emphasize the importance of cooperative, international emission control efforts in further ozone reductions.

Overall, I think this paper addresses some important issues in current researches of tropospheric ozone. The methods applied in this paper are acceptable. Although I cannot judge at this time to what extent the authors of this paper are right, I do think the Linear Trend View may have exaggerated the COVID-19 impact on baseline ozone. Further studies and discussions are definitely needed to come to a consensus. This paper could be a starting point for these. The paper is mostly well written. I have only a few minor points and recommend publication of this paper in ACP after revisions.

Specific comments:

To be more robust and convincing, the 2-year average point with error bar for 2018 and the related monthly mean ozone values should be included in Figure 1. Some of the 2-year averages do show quite large deviations from the quadratic fit curve. In case a large deviation occurred in 2018, the extrapolation could be substantially impacted. In addition, it is not known why 2-year means are used in the regression. Can we obtain a significantly different fit using 1-year means? Will the conclusion also be different?

The quadratic fit is obtained from selected datasets that are believed to represent baseline ozone, while the Linear Trend View does not pay much attention to the datasets selection. Therefore, it should be made clear that compared trends are based on same or similar (baseline ozone) datasets.

Figure S2 is not cited in the main text and some related descriptions in Supplement seem to be unclear. For example, I do not understand why “t2=17” (line 65).

Line 189: Clark et al. (2022) should be Clark et al. (2021).

References

Chang, K.-L., et al. (2022). Impact of the COVID-19 economic downturn on tropospheric ozone trends: An uncertainty weighted data synthesis for quantifying regional anomalies above western North America and Europe. AGU Advances, 3, e2021AV000542. https://doi.org/10.1029/2021AV000542.

Clark, H., et al. (2021). The effects of the COVID-19 lockdowns on the composition of the troposphere as seen by In-service Aircraft for a Global Observing System (IAGOS) at Frankfurt, Atmos. Chem. Phys., 21, 16237–16256.

Cooper, O. R., et al. (2020). Multi-decadal surface ozone trends at globally distributed remote locations, Elem. Sci. Anth., 8,23. doi.org/10.1525/elementa.420

Cristofanelli, P., et al. (2021). Negative ozone anomalies at a high mountain site in northern Italy during 2020: a possible role of COVID-19 lockdowns? Environ. Res. Lett., 16 074029

Derwent, R. G., et al. (2018). Long-term trends in ozone in baseline and European regionally-polluted air at Mace Head, Ireland over a 30-year period. Atmos. Environ., 179, 279–287. https://doi.org/10.1016/j.atmosenv.2018.02.024

Derwent, R.G., and Parrish, D.D. (2022). Analysis and assessment of the observed long-term changes over three decades in ground-level ozone across north-west Europe from 1989 – 2018. Atmos. Environ., doi: https://doi.org/10.1016/j.atmosenv.2022.119222.

Gaudel, A., et al. (2018). Tropospheric ozone assessment report: Presentâday distribution and trends of tropospheric ozone relevant to climate and global atmospheric chemistry model evaluation. Elem. Sci. Anth., 6(1), 39. https://doi.org/10.1525/elementa.291

Gaudel, A., et al. (2020). Aircraft observations since the 1990s reveal increases of tropospheric ozone at multiple locations across the Northern Hemisphere. Sci. Adv., 6, eaba8272.

Junge, C. E., (1962). Global ozone budget and exchange between stratosphere and troposphere, Tellus, XIV, 363-377.

Logan, J. A., et al. (2012). Changes in ozone over Europe: Analysis of ozone measurements from sondes, regular aircraft (MOZAIC) and alpine surface sites. J. Geophys. Res., 117, D09301. doi:10.1029/2011JD016952

Parrish, D. D., et al. (2012). Longâterm changes in lower tropospheric baseline ozone concentrations at northern midâlatitudes. Atmos. Chem. Phys., 12, 11,485–11,504.

Parrish, D. D., et al. (2014). Longâterm changes in lower tropospheric baseline ozone concentrations: Comparing chemistryâclimate models and observations at northern midlatitudes. J. Geophys. Res., 119, 5719–5736.

Parrish, D. D., et al. (2017). Reversal of longâterm trend in baseline ozone concentrations at the North American west coast. Geophysical Research Letters, 44, 10,675–10,681.

Parrish, D. D., et al. (2020). Zonal similarity of longâterm changes and seasonal cycles of baseline ozone at northern midlatitudes. J. Geophys. Res., 125, e2019JD031908. https://doi.org/10.1029/2019JD031908

Parrish, D.D., et al. (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., et al. (2021b). Long-term baseline ozone changes in the Western US: A synthesis of analyses. J. Air & Waste Manag. Ass., DOI: 10.1080/10962247.2021.1945706

Parrish, D.D., et al. (2022), Observational-based Assessment of Contributions to Maximum Ozone Concentrations in the western United States, J. Air & Waste Manag. Ass., 72:5, 434–454.

Steinbrecht, W., et al. (2021). COVID-19 crisis reduces free tropospheric ozone across the Northern Hemisphere. Geophy. Res. Lett., 48, e2020GL091987. https://doi.org/10.1029/2020GL091987

Tarasick, D., et al. (2019). Tropospheric ozone from 1877 to 2016, observed levels, trends and uncertainties. Elem. Sci. Anth., 7,39. https://doi.org/10.1525/elementa.376.
Citation: https://doi.org/10.5194/acp-2022-424-RC3
RC4: 'Comment on acp-2022-424', Anonymous Referee #3, 22 Aug 2022

This is another in a series of papers from David Parrish, which all have as their main purpose to maintain that Parrish et al. (2014), was correct and subsequent work is all flawed. It is tiresome to keep refuting them. Parrish et al. (2014) was indeed a valuable contribution, in 2014, and its finding that tropospheric ozone had increased by a factor of 2-3 was challenging to models. However more recent work, particularly Tarasick, Galbally et al. (2019), which examined biases in historical measurements in great depth, has found smaller increases in surface ozone, of the order of 50%, which are in general agreement with model predictions. The analysis of ice-core data by Yeung et al. (2019), and the independent analysis of aircraft and balloon data by Tarasick, Galbally et al. (2019), also both support a smaller increase of surface ozone, of the order of 50%. Dr. Parrish’s papers invariably fail to cite these corroborating analyses.

The main issue appears to be Dr. Parrish’s insistence that his few selected sites, primarily in Europe, are more representative of “background ozone” than averages from the much more extensive TOAR set of rural ozone measurement records. There seems to be no justification for this other than Dr. Parrish’s insistence. See Cooper et al. (2021), for a more extensive discussion.

The data presented all seem to be from previous publications. The sole novelty is the projection of Dr. Parrish’s peculiar quadratic fit to 2020, using data up to 2018, and comparing it with other, more conventional linear fits. Since he is attempting to publish this in 2022, surely it is reasonable to insist that he extend his dataset to see which projection is closer to the observations? The current Figure 1 has all the interest of a weather forecast for 2020, made in 2018.

Cooper, O.R, D.W. Tarasick, I.E. Galbally and M.G. Schultz, Comment on acp-2020-1198, community comment on "Investigations on the anthropogenic reversal of the natural ozone gradient between northern and southern midlatitudes" by David D. Parrish et al., Atmos. Chem. Phys. Discuss., https://doi.org/10.5194/acp-2020-1198-CC1, 2021

Tarasick, D.W., I. Galbally, et al., (2019), TOAR- Observations: Tropospheric ozone from 1877 to 2016, observed levels, trends and uncertainties, Elem Sci Anth, 7(1), p.39. DOI: http://doi.org/10.1525/elementa.376.

Yeung, L.Y., L.T. Murray, P. Martinerie, E. Witrant, H. Hu, A. Banerjee, A. Orsi and J. Chappellaz (2019), Isotopic constraint on the twentieth-century increase in tropospheric ozone, Nature, 570, 224-227, https://doi.org/10.1038/s41586-019-1277-1.

Citation: https://doi.org/10.5194/acp-2022-424-RC4
AC3: 'Final response to referee comments on acp-2022-424', David Parrish, 07 Sep 2022

Here we respond to all 4 referee comments. Some of the material from our two previous responses (AC1 and AC2) are reproduced here.
Please see uploaded .pdf file.

Citation: https://doi.org/10.5194/acp-2022-424-AC3

Peer review completion

AR: Author's response | RR: Referee report | ED: Editor decision | EF: Editorial file upload

AR by David Parrish on behalf of the Authors (07 Sep 2022) Author's response Author's tracked changes Manuscript

ED: Referee Nomination & Report Request started (09 Sep 2022) by Jianzhong Ma

RR by Anonymous Referee #2 (17 Sep 2022)

RR by Anonymous Referee #1 (17 Sep 2022)

Suggestions for revision or reasons for rejection

I have read the author's extensive response to the reviews. I have also read the change-tracked revised manuscript with its very minor changes.

Unfortunately, I maintain my opinion that this manuscript does not present any important new information over what is already published in Parrish et al. 2020. The extrapolation of data up to 2018 only (all already presented in Parrish et al. 2020, with basically an identical Figure), to provide a background value for 2020, is highly questionable. The similarity of the extrapolated 2020 anomaly to the observed 2020 anomaly of Steinbrecht et al. 2021 is purely co-incidental.

The author's claim that there is little or no effect of the COVID-19 related large emission reduction in the 2020 anomaly. This contradicts chemistry-transport model simulations (e.g. by Gaubert et al. 2021 or Miyazaki et al. 2021). These studies very cleary show effects of COVID-19 emission reductions on free tropospheric ozone, which are fully consistent with the 2020 anomalies reported by Steinbrecht et al. 2021. In these chemistry transport model studies, simulations with full emissions are compared to simulations with reduced emissions. The difference gives the COVID-19 effect, without any need for estimating a 2020 background from previous years (different from what is done in the manuscript).

So a major message of the paper (that the COVID-19 effect is much smaller than estimated e.g. in Steinbrecht et al. 2021) is wrong or misleading. There are also no new data that support it. Figure 2 does show a substantial COVID-19 effect for Mace Head (about -5 ppb or -10%), when compared to the 1997 to 2006 and 2007 to 2016 baselines. This effect disappears only, if the 1987 to 1996 data are used as the baseline (for 2020!). Almost all other statements of the manuscript present nothing new over what ist already presented in Parrish et al. 2020.

I fully agree with the author that scientific discussion and critique of existing literature is very important, and that new ideas should be introduced even if their statistical significance is not at the 95% or 99% level. But this should be done with substantial new data or ideas, and it should not be misleading. My conclusion remains: This present manuscript has not enough to warrant an ACP paper. In a much shorter form, it might be appropriate as a comment to one of the existing papers below.

References:
Bouarar, I., et al. (2021). Ozone anomalies in the free troposphere during the COVID-19 pandemic. Geophys. Res. Lett., 48, e2021GL094204. https://doi.org/10.1029/2021GL094204

Miyazaki, K., et al. (2021). Global tropospheric ozone responses to reduced NOx emissions linked to the COVID-19 worldwide lockdowns. Science Advances, 7(24). https://doi.org/10.1126/sciadv.abf7460

Parrish, D.D., et al. (2020). Zonal similarity
of long-term changes and seasonal cycles of baseline ozone at northern midlatitudes. J. Geophys. Res., 125, e2019JD031908. https://doi.org/10.1029/2019JD031908

Steinbrecht, W., et al. (2021). COVID-19 crisis reduces free tropospheric ozone across the Northern Hemisphere. Geophys. Res. Lett., 48, e2020GL091987. https://doi.org/10.1029/2020GL091987

Referee Report: PDF

Hide

ED: Publish subject to minor revisions (review by editor) (20 Sep 2022) by Jianzhong Ma

AR by David Parrish on behalf of the Authors (23 Sep 2022) Author's response Author's tracked changes Manuscript

ED: Publish as is (25 Sep 2022) by Jianzhong Ma

AR by David Parrish on behalf of the Authors (01 Oct 2022)

Download

Article (1926 KB)
Full-text XML

Short summary

Accounting for the continuing long-term decrease of pollution ozone and the large 2020 Arctic stratospheric ozone depletion event improves estimates of background ozone changes caused by COVID-19-related emission reductions; they are smaller than reported earlier. Cooperative, international emission control efforts aimed at maximizing the ongoing decrease in hemisphere-wide background ozone may be the most effective approach to improving ozone pollution in northern midlatitude countries.