Source specific bias correction of US background ozone modeled in CMAQ

Skipper, T. Nash; Hogrefe, Christian; Henderson, Barron H.; Mathur, Rohit; Foley, Kristen M.; Russell, Armistead G.

doi:https://doi.org/10.5194/egusphere-2024-554

Preprints

https://doi.org/10.5194/egusphere-2024-554

Preprints

25 Mar 2024

| 25 Mar 2024

Source specific bias correction of US background ozone modeled in CMAQ

T. Nash Skipper, Christian Hogrefe, Barron H. Henderson, Rohit Mathur, Kristen M. Foley, and Armistead G. Russell

Abstract. United States (US) background ozone (O₃) is the counterfactual O₃ that would exist with zero US anthropogenic emissions. Estimates of US background O₃ typically come from chemical transport models (CTMs), but different models vary in their estimates of both background and total O₃. Here, a measurement-model data fusion approach is used to estimate CTM biases in US anthropogenic O₃ and multiple US background O₃ sources, including natural emissions, long-range international emissions, short-range international emissions from Canada and Mexico, and stratospheric O₃. Spatially and temporally varying bias correction factors adjust each simulated O₃ component so that the sum of the adjusted components evaluates better against observations compared to unadjusted estimates. The estimated correction factors suggest a seasonally consistent positive bias in US anthropogenic O₃ in the eastern US, with the bias becoming higher with coarser model resolution and with higher simulated total O₃ though the bias does not increase much with higher observed O₃. Correlation among different US background O₃ components can increase the uncertainty in the estimation of the source-specific adjustment factors. Despite this, results indicate that there may be a negative bias in modeled estimates of the impact of stratospheric O₃ at the surface. This type of data fusion approach can be extended to include data from multiple models to leverage the strengths of different data sources while reducing uncertainty in the US background ozone estimates.

Received: 24 Feb 2024 – Discussion started: 25 Mar 2024

Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims made in the text, published maps, institutional affiliations, or any other geographical representation in this preprint. The responsibility to include appropriate place names lies with the authors.

Download & links

Preprint (PDF, 3141 KB)

Notice on discussion status
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
Preprint (3141 KB)

Supplement (1612 KB)

Download & links

The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.

Journal article(s) based on this preprint

26 Nov 2024

Source-specific bias correction of US background and anthropogenic ozone modeled in CMAQ

T. Nash Skipper, Christian Hogrefe, Barron H. Henderson, Rohit Mathur, Kristen M. Foley, and Armistead G. Russell

Geosci. Model Dev., 17, 8373–8397, https://doi.org/10.5194/gmd-17-8373-2024,https://doi.org/10.5194/gmd-17-8373-2024, 2024

Short summary

T. Nash Skipper, Christian Hogrefe, Barron H. Henderson, Rohit Mathur, Kristen M. Foley, and Armistead G. Russell

Interactive discussion

Status: closed

RC1:
'Comment on egusphere-2024-554', Anonymous Referee #1, 16 Apr 2024
Review comments on the manuscript titled “Source specific bias correction of US background ozone modeled in CMAQ” by T. Nash Skipper et al.,
Background O₃ constitutes a significant portion of total surface O₃ and the contribution becomes higher when O₃ levels decline. This study used a measurement-model data fusion approach to assess CTM biases in USB O₃ and attribute these biases to different sources. Two sets of CMAQ simulations, PA and EQUATES, are conducted to analyze the contributions of different sources to USB O₃. While the study is of scientific significance in assessing model biases in background O₃ estimation, two major concerns need to be addressed by the authors.
(1) The two sets of simulations are for different years, the model configuration and inputs are different, different versions of emissions inventory are adopted, which will introduce substantial uncertainties. The PA simulations cover the year of 2016; while the EQUATES simulations span between 2002-2019. Additional simulations using the EQUATES modeling framework were conducted for 2016–2017 to estimate USB O₃ and USA O₃ using the zero-out method. CMAQ v5.2.1 was used for the PA simulations while CMAQ v5.3.2 was used for the EQUATES simulations. It seems this study is not outlined within a comprehensive framework but combine different modelling work to do the current study. The differences of biases caused by different model configurations and model setup between the two sets of scenarios need to be fully discussed. (2) While biases in the study are fully discussed, the reasons behind these biases remain ambiguous. Factors such as uncertainties in emissions inventory, meteorology simulations, and chemical mechanisms could contribute to biases. Providing insights into the main drivers of biases and offering suggestions for modelers to mitigate these biases would enhance the value of the study for readers seeking to improve background O₃ modeling accuracy.
Specific comments:
Abstract. The current form of the abstract is notably objective, it lacks quantitative results detailing the biases and their spatial-temporal characteristics. Additionally, what can we do to reduce these biases? The abstract could benefit from discussing potential strategies to mitigate these biases.

Methods. The overall model configuration needs to be briefly outlined in the main text, such as modelling domain, WRF configuration, CMAQ gas-phase mechanism, IC/BC, vertical layers, etc.

Table 1. The descriptions provided in Table 1 could benefit from clarification regarding emissions from other regions. For instance, in the case of "ZROW," where all international anthropogenic emissions are removed, it's unclear whether this includes emissions from the United States, Canada, and Mexico. To enhance clarity, it is recommended to use a clear table format that separates regions with emissions using symbols such as "√" to indicate the presence of emissions and "×" to denote the absence of emissions. This approach will facilitate a more straightforward understanding of the emission scenarios across different regions. Addtionally, Table 1 and Table S1 is exactly the same.

In the main text, Tables S4 and S5 (Line 95) comes before Table S3 (Line 101), this is strange.

Line 115, “STRAT” should be spelled out at the first time it appears.

CTM results: can you explain further why the 12km simulations have the best performance? Additionally, what about the model performance for NO2 simulations?

Table 2 and 3, what’s the unit of the data in this table?

section 3.3. The authors extensively compared the differences in deviations between different scenarios. Are these differences in deviations related to the zero-out method neglecting nonlinear ozone production?

section 3.4. The authors extensively described the results of model bias. These results seem to be an extension of section 3.3, but what can these results further illustrate?
Citation: https://doi.org/10.5194/egusphere-2024-554-RC1
RC2: 'Comment on egusphere-2024-554', Anonymous Referee #2, 19 Apr 2024

Major comments:
1. This study analyzed surface ozone from a suite of hemispheric and regional-scale CMAQ simulations for 2016 and 2017 and attempted to attribute the biases in model simulated total surface ozone to different components, including ozone produced from US anthropogenic emissions, natural sources, intercontinental transport, and stratospheric intrusions. Understanding US background ozone and its components is of broad interest because they are directly relevant to the setting and implementation of US ozone air quality standards. However, the manuscript needs to be substantially revised before it can be published. Description of the methodology used and discussions in many sections are incomplete. The authors should also discuss the model biases in the context of published literature. The referee’s main concern is on the methodology used to attribute the model biases to different components. The description of the data fusion model in Section 2.3 is hard to understand. Is the data fusion model trained using one set of simulations and applied to another set of simulations for the bias attribution? How do you know the sources of biases in the two sets of simulations are the same? There are a couple of places where the authors refer to Skipper et al. (ES & T, 2021) for the method, but that study did not discuss the different USB components.
2. The title of this paper is about the bias of US background ozone, but in the abstract and in the paper, there is substantial discussion on the biases of US anthropogenic O3 and the influence of model resolution. The authors stated “The estimated correction factors suggest a seasonally consistent positive bias in US anthropogenic O3 in the eastern US, with the bias becoming higher with coarser model resolution and with higher simulated total O3 though the bias does not increase much with higher observed O3.” This statement seems to imply that coarser model resolution always produces higher US anthropogenic O3, which is not true. There is clearly a seasonal dependence. During winter when ozone production is in NOx-saturated regime, coarser model resolution leads to artificial dilution of NOx and thus higher O3 due to less NOx titration. During summer, however, when ozone production at most locations is in NOx-limited regime, coarser model resolution may lead to lower ozone concentrations produced from regional anthropogenic emissions. Increasing model resolution may lead to higher simulated US anthropogenic O3, leading to better agreement with observations, such as in the Central Valley of California. These seasonal characteristics of model resolution impacts on US anthropogenic ozone are clearly demonstrated in the published literature, including the recent studies of Schwantes et al. (2021) and Lin et a. (2024).
Figure 6: the authors should present results for different seasons, not annual averages.
Figure 13: What is the horizontal resolution of PA and EQUATES simulations presented in this figure? Are the differences driven by differences in model configurations or model resolution? The authors should show the comparison from the same configuration but at different resolutions.
Schwantes, R. H., Lacey, F. G., Tilmes, S., Emmons, L. K., Lauritzen, P. H., Walters, S., et al. (2022). Evaluating the impact of chemical complexity and horizontal resolution on tropospheric ozone over the conterminous US with a global variable resolution chemistry model. Journal of Advances in Modeling Earth Systems, 14(6), e2021MS002889. https://doi.org/10.1029/2021MS002889
Lin, M., L. W. Horowitz, M. Zhao, L. Harris, P. Ginoux, J. P. Dunne, S. Malyshev, E. Shevliakova, H. Ahsan, S. Garner, F. Paulot, A. Pouyaei, S. J. Smith, Y. Xie, N. Zadeh, L. Zhou. The GFDL Variable-Resolution Global Chemistry-Climate Model for Research at the Nexus of US Climate and Air Quality Extremes. Journal of Advances in Modeling Earth Systems, in press, https://doi.org/10.1029/2023MS003984, 2024
3. Discussion on stratospheric contribution and CMAQ low-O3 bias in spring should be placed in the broader published literature, including those using dynamic stratospheric ozone tracers with explicit stratospheric chemistry and evaluation with intensive ozone profiling during western US field campaigns. The stratospheric contribution estimated by CMAQ appears to be much lower than the estimates from these prior studies:
A.O. Langford, R.J. Alvarez II, J. Brioude, R. Fine, M. Gustin, J.S. Holloway, M.Y. Lin, R.D. Marchbanks, R.B. Pierce, S.P. Sandberg, C.J. Senff, A.M. Weickmann, E.J. Williams, Entrainment of stratospheric air and Asian pollution by the convective boundary layer in the Southwestern U.S., J. Geophys. Res., 122 (2), doi:10.1002/2016JD025987, 2017.
Lin M., A. M. Fiore , O. R. Cooper , L. W. Horowitz , A. O. Langford , Hiram Levy II , B. J. Johnson , V. Naik , S. J. Oltmans , C. Senff (2012): Springtime high surface ozone events over the western United States: Quantifying the role of stratospheric intrusions, Journal of Geophysical Research, 117, D00V22, doi:10.1029/2012JD018151
Langford, A.O., C.J. Senff, R.J. Alvarez II, J. Brioude, O.R. Cooper, J.S. Holloway, M.Y. Lin, R.D. Marchbanks, R.B. Pierce, S.P. Sandberg, A.M. Weickmann , E.J. Williams (2015): An overview of the 2013 Las Vegas Ozone Study (LVOS): Impact of stratospheric intrusions and long-range transport on surface air quality. Atmos. Environ, doi:10.1016/j.atmosenv.2014.08.040
Langford, A. O., Senff, C. J., Alvarez II, R. J., Aikin, K. C., Baidar, S., Bonin, T. A., Brewer, W. A., Brioude, J., Brown, S. S., Burley, J. D., Caputi, D. J., Conley, S. A., Cullis, P. D., Decker, Z. C. J., Evan, S., Kirgis, G., Lin, M., Pagowski, M., Peischl, J., Petropavlovskikh, I., Pierce, R. B., Ryerson, T. B., Sandberg, S. P., Sterling, C. W., Weickmann, A. W., and Zhang, L.: The Fires, Asian, and Stratospheric Transport-Las Vegas Ozone Study (FAST-LVOS), Atmos. Chem. Phys., https://doi.org/10.5194/acp-2021-690, 2022.

Other comments:
Lines 29-30: “USB O3 … is a larger portion of total observed O3 as anthropogenic precursor emissions decline”. This statement needs a few references, such as Lin et al. (2017):
Lin, M.., W. Horowitz, R. Payton, A.M. Fiore, G. Tonnesen (2017). US surface ozone trends and extremes from 1980 to 2014: Quantifying the roles of rising Asian emissions, domestic controls, wildfires, and climate. Atmos. Chem. Phys., doi:10.5194/acp-17-2943-2017

Lines 38-48: Need references. Could also discuss the difficulty to separate the anthropogenic and natural driver of wildfire impacts on ozone air quality, as ozone production is enhanced due to mixing of wildfire VOC emissions with urban NOx?

Table 1: The referee agrees with Referee #1 that the authors should list more detailed information regarding model version, simulations types, horizontal and vertical resolution, US anthropogenic emissions, international emissions, fire emissions (including temporal frequency and injection height), and other natural emissions.
For all of the figures, please indicate in the figure caption whether MDA8 O3 or 24-h O3 is shown. There are some discussions of the metric in the first paragraph of Section 2.1. But it is much easier for readers if you label them as “MDA8 O3” directly in the figure captions.
Figures 2 to 5: Results in the maps look pretty similar in their current form. Please use a different colorbar so that the spatial distribution of different model configurations can be better illustrated!

Citation: https://doi.org/10.5194/egusphere-2024-554-RC2
AC1: 'Comment on egusphere-2024-554', Nash Skipper, 16 Jun 2024

We would like to thank the reviewers for their comments. Responses are provided in the attached pdf.

Citation: https://doi.org/10.5194/egusphere-2024-554-AC1

Interactive discussion

Status: closed

RC1:
'Comment on egusphere-2024-554', Anonymous Referee #1, 16 Apr 2024
Review comments on the manuscript titled “Source specific bias correction of US background ozone modeled in CMAQ” by T. Nash Skipper et al.,
Background O₃ constitutes a significant portion of total surface O₃ and the contribution becomes higher when O₃ levels decline. This study used a measurement-model data fusion approach to assess CTM biases in USB O₃ and attribute these biases to different sources. Two sets of CMAQ simulations, PA and EQUATES, are conducted to analyze the contributions of different sources to USB O₃. While the study is of scientific significance in assessing model biases in background O₃ estimation, two major concerns need to be addressed by the authors.
(1) The two sets of simulations are for different years, the model configuration and inputs are different, different versions of emissions inventory are adopted, which will introduce substantial uncertainties. The PA simulations cover the year of 2016; while the EQUATES simulations span between 2002-2019. Additional simulations using the EQUATES modeling framework were conducted for 2016–2017 to estimate USB O₃ and USA O₃ using the zero-out method. CMAQ v5.2.1 was used for the PA simulations while CMAQ v5.3.2 was used for the EQUATES simulations. It seems this study is not outlined within a comprehensive framework but combine different modelling work to do the current study. The differences of biases caused by different model configurations and model setup between the two sets of scenarios need to be fully discussed. (2) While biases in the study are fully discussed, the reasons behind these biases remain ambiguous. Factors such as uncertainties in emissions inventory, meteorology simulations, and chemical mechanisms could contribute to biases. Providing insights into the main drivers of biases and offering suggestions for modelers to mitigate these biases would enhance the value of the study for readers seeking to improve background O₃ modeling accuracy.
Specific comments:
Abstract. The current form of the abstract is notably objective, it lacks quantitative results detailing the biases and their spatial-temporal characteristics. Additionally, what can we do to reduce these biases? The abstract could benefit from discussing potential strategies to mitigate these biases.

Methods. The overall model configuration needs to be briefly outlined in the main text, such as modelling domain, WRF configuration, CMAQ gas-phase mechanism, IC/BC, vertical layers, etc.

Table 1. The descriptions provided in Table 1 could benefit from clarification regarding emissions from other regions. For instance, in the case of "ZROW," where all international anthropogenic emissions are removed, it's unclear whether this includes emissions from the United States, Canada, and Mexico. To enhance clarity, it is recommended to use a clear table format that separates regions with emissions using symbols such as "√" to indicate the presence of emissions and "×" to denote the absence of emissions. This approach will facilitate a more straightforward understanding of the emission scenarios across different regions. Addtionally, Table 1 and Table S1 is exactly the same.

In the main text, Tables S4 and S5 (Line 95) comes before Table S3 (Line 101), this is strange.

Line 115, “STRAT” should be spelled out at the first time it appears.

CTM results: can you explain further why the 12km simulations have the best performance? Additionally, what about the model performance for NO2 simulations?

Table 2 and 3, what’s the unit of the data in this table?

section 3.3. The authors extensively compared the differences in deviations between different scenarios. Are these differences in deviations related to the zero-out method neglecting nonlinear ozone production?

section 3.4. The authors extensively described the results of model bias. These results seem to be an extension of section 3.3, but what can these results further illustrate?
Citation: https://doi.org/10.5194/egusphere-2024-554-RC1
RC2: 'Comment on egusphere-2024-554', Anonymous Referee #2, 19 Apr 2024

Major comments:
1. This study analyzed surface ozone from a suite of hemispheric and regional-scale CMAQ simulations for 2016 and 2017 and attempted to attribute the biases in model simulated total surface ozone to different components, including ozone produced from US anthropogenic emissions, natural sources, intercontinental transport, and stratospheric intrusions. Understanding US background ozone and its components is of broad interest because they are directly relevant to the setting and implementation of US ozone air quality standards. However, the manuscript needs to be substantially revised before it can be published. Description of the methodology used and discussions in many sections are incomplete. The authors should also discuss the model biases in the context of published literature. The referee’s main concern is on the methodology used to attribute the model biases to different components. The description of the data fusion model in Section 2.3 is hard to understand. Is the data fusion model trained using one set of simulations and applied to another set of simulations for the bias attribution? How do you know the sources of biases in the two sets of simulations are the same? There are a couple of places where the authors refer to Skipper et al. (ES & T, 2021) for the method, but that study did not discuss the different USB components.
2. The title of this paper is about the bias of US background ozone, but in the abstract and in the paper, there is substantial discussion on the biases of US anthropogenic O3 and the influence of model resolution. The authors stated “The estimated correction factors suggest a seasonally consistent positive bias in US anthropogenic O3 in the eastern US, with the bias becoming higher with coarser model resolution and with higher simulated total O3 though the bias does not increase much with higher observed O3.” This statement seems to imply that coarser model resolution always produces higher US anthropogenic O3, which is not true. There is clearly a seasonal dependence. During winter when ozone production is in NOx-saturated regime, coarser model resolution leads to artificial dilution of NOx and thus higher O3 due to less NOx titration. During summer, however, when ozone production at most locations is in NOx-limited regime, coarser model resolution may lead to lower ozone concentrations produced from regional anthropogenic emissions. Increasing model resolution may lead to higher simulated US anthropogenic O3, leading to better agreement with observations, such as in the Central Valley of California. These seasonal characteristics of model resolution impacts on US anthropogenic ozone are clearly demonstrated in the published literature, including the recent studies of Schwantes et al. (2021) and Lin et a. (2024).
Figure 6: the authors should present results for different seasons, not annual averages.
Figure 13: What is the horizontal resolution of PA and EQUATES simulations presented in this figure? Are the differences driven by differences in model configurations or model resolution? The authors should show the comparison from the same configuration but at different resolutions.
Schwantes, R. H., Lacey, F. G., Tilmes, S., Emmons, L. K., Lauritzen, P. H., Walters, S., et al. (2022). Evaluating the impact of chemical complexity and horizontal resolution on tropospheric ozone over the conterminous US with a global variable resolution chemistry model. Journal of Advances in Modeling Earth Systems, 14(6), e2021MS002889. https://doi.org/10.1029/2021MS002889
Lin, M., L. W. Horowitz, M. Zhao, L. Harris, P. Ginoux, J. P. Dunne, S. Malyshev, E. Shevliakova, H. Ahsan, S. Garner, F. Paulot, A. Pouyaei, S. J. Smith, Y. Xie, N. Zadeh, L. Zhou. The GFDL Variable-Resolution Global Chemistry-Climate Model for Research at the Nexus of US Climate and Air Quality Extremes. Journal of Advances in Modeling Earth Systems, in press, https://doi.org/10.1029/2023MS003984, 2024
3. Discussion on stratospheric contribution and CMAQ low-O3 bias in spring should be placed in the broader published literature, including those using dynamic stratospheric ozone tracers with explicit stratospheric chemistry and evaluation with intensive ozone profiling during western US field campaigns. The stratospheric contribution estimated by CMAQ appears to be much lower than the estimates from these prior studies:
A.O. Langford, R.J. Alvarez II, J. Brioude, R. Fine, M. Gustin, J.S. Holloway, M.Y. Lin, R.D. Marchbanks, R.B. Pierce, S.P. Sandberg, C.J. Senff, A.M. Weickmann, E.J. Williams, Entrainment of stratospheric air and Asian pollution by the convective boundary layer in the Southwestern U.S., J. Geophys. Res., 122 (2), doi:10.1002/2016JD025987, 2017.
Lin M., A. M. Fiore , O. R. Cooper , L. W. Horowitz , A. O. Langford , Hiram Levy II , B. J. Johnson , V. Naik , S. J. Oltmans , C. Senff (2012): Springtime high surface ozone events over the western United States: Quantifying the role of stratospheric intrusions, Journal of Geophysical Research, 117, D00V22, doi:10.1029/2012JD018151
Langford, A.O., C.J. Senff, R.J. Alvarez II, J. Brioude, O.R. Cooper, J.S. Holloway, M.Y. Lin, R.D. Marchbanks, R.B. Pierce, S.P. Sandberg, A.M. Weickmann , E.J. Williams (2015): An overview of the 2013 Las Vegas Ozone Study (LVOS): Impact of stratospheric intrusions and long-range transport on surface air quality. Atmos. Environ, doi:10.1016/j.atmosenv.2014.08.040
Langford, A. O., Senff, C. J., Alvarez II, R. J., Aikin, K. C., Baidar, S., Bonin, T. A., Brewer, W. A., Brioude, J., Brown, S. S., Burley, J. D., Caputi, D. J., Conley, S. A., Cullis, P. D., Decker, Z. C. J., Evan, S., Kirgis, G., Lin, M., Pagowski, M., Peischl, J., Petropavlovskikh, I., Pierce, R. B., Ryerson, T. B., Sandberg, S. P., Sterling, C. W., Weickmann, A. W., and Zhang, L.: The Fires, Asian, and Stratospheric Transport-Las Vegas Ozone Study (FAST-LVOS), Atmos. Chem. Phys., https://doi.org/10.5194/acp-2021-690, 2022.

Other comments:
Lines 29-30: “USB O3 … is a larger portion of total observed O3 as anthropogenic precursor emissions decline”. This statement needs a few references, such as Lin et al. (2017):
Lin, M.., W. Horowitz, R. Payton, A.M. Fiore, G. Tonnesen (2017). US surface ozone trends and extremes from 1980 to 2014: Quantifying the roles of rising Asian emissions, domestic controls, wildfires, and climate. Atmos. Chem. Phys., doi:10.5194/acp-17-2943-2017

Lines 38-48: Need references. Could also discuss the difficulty to separate the anthropogenic and natural driver of wildfire impacts on ozone air quality, as ozone production is enhanced due to mixing of wildfire VOC emissions with urban NOx?

Table 1: The referee agrees with Referee #1 that the authors should list more detailed information regarding model version, simulations types, horizontal and vertical resolution, US anthropogenic emissions, international emissions, fire emissions (including temporal frequency and injection height), and other natural emissions.
For all of the figures, please indicate in the figure caption whether MDA8 O3 or 24-h O3 is shown. There are some discussions of the metric in the first paragraph of Section 2.1. But it is much easier for readers if you label them as “MDA8 O3” directly in the figure captions.
Figures 2 to 5: Results in the maps look pretty similar in their current form. Please use a different colorbar so that the spatial distribution of different model configurations can be better illustrated!

Citation: https://doi.org/10.5194/egusphere-2024-554-RC2
AC1: 'Comment on egusphere-2024-554', Nash Skipper, 16 Jun 2024

We would like to thank the reviewers for their comments. Responses are provided in the attached pdf.

Citation: https://doi.org/10.5194/egusphere-2024-554-AC1

Peer review completion

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

AR by Nash Skipper on behalf of the Authors (16 Jun 2024) Author's response Author's tracked changes Manuscript

ED: Referee Nomination & Report Request started (01 Jul 2024) by Jason Williams

RR by Anonymous Referee #3 (03 Jul 2024)

Suggestions for revision or reasons for rejection

I was not a reviewer for the original draft, and thus will only make overarching comments.

I found the acronyms of the model runs to be very confusing and not at all intuitive. For that reason, this manuscript was very hard to follow. I highly recommend the authors spell the words out rather than use acronyms. NAT=Natural, LINTL=Long-range international CANMEX=Canada-Mexico USB =U.S. Background, USA=U.S. Anthropogenic (which is different than USA the country, so confusing!) Please spell out these acronyms.

There is a mismatch between the annual/seasonal averages of MDA8 and how the MDA8 is regulated which is the 4th highest concentration (essentially 99th percentile of annual concentrations). This is not a relevant comment in all aspects of the manuscript, however, this is particularly relevant for Tables 2 & 3 and Figures 2-5: the U.S. anthropogenic component, ~10 ppb, which makes sense for an annual average, but not relevant for a 99th percentile day. If I had reviewed the original manuscript, I would've had you plot the 99th percentile of Figure 2 and insert/discuss between Figures 2 & 3. That is still my recommendation, although I realize I did not review this in Round 1, so incorporating this comment may take a lot of effort. I will ultimately leave that decision up to the Editor. At the absolute minimum it would be important to acknowledge several times throughout the manuscript that the annual average is not relevant to the current NAAQS, especially near the text referring to Tables 2 & 3, Figures 2-5 (more than a brief mention in Line 764 Track Changes version at the end of the manuscript).

The biases in the Spring and Summer (Lines 814-816 Tracked Changes version) seem like they warrant further investigation, especially since O3 is regulated based on the 4th highest value which typically occurs in May-Sept. I think Lines 817-832 suggests stratospheric influence is underestimated and missing particulate nitrate photolysis, but I think there may be other reasons. There is no mention of O3 produced from lightning and soil in the Discussion section, and I'm sure both have large uncertainties. Also no mention of O3 deposition uncertainties, or ozone production efficiency uncertainties. All of these uncertainties need to be discussed.

Hide

RR by Anonymous Referee #4 (12 Aug 2024)

Suggestions for revision or reasons for rejection

This study highlights biases as seen in US background ozone and its contribution to total ozone as predicted by two sets of CMAQ simulations, using a model-measurement fusion model utilizing a multivariate least square regression approach. The authors highlight that such a data fusion approach leverages the strengths of different data sources to reduce modeled US background (USB) ozone (O3) biases. The reduction in model USB O3 biases is critical, given the NAAQS for ozone will become more stringent, increasing the importance of background O3 (and its sources) to total O3. The manuscript is quite comprehensive and acceptable for publication, given the following main conceptual comments are addressed:
Comments:
1) Abstract + Lines 61-63: “We extend the bias correction method to estimate biases in separate components of USB O3. Separating the USB O3 components provides new insights into the inferred CTM error in USB O3 that was not possible when USB O3 was treated as a lumped quantity”.
Comments from the erstwhile reviewer(s) to make the abstract more quantitative have been addressed to some extent. However, the findings (or its summary) about ‘source-specific US background O3 biases’ (or quantitatively how much of them are corrected by this work?) are lacking in the abstract, given the authors propose this manuscript as an update to Skipper et al., 2021 in that specific direction only.
Please consider adding 1-2 sentences on the main findings on source-specific USB O3 biases upfront in the abstract. Currently, the variation in USB O3 bias seasonally and with model resolution is quantitatively mentioned in the abstract. Summarize source-specific bias findings as discussed in the conclusions, in the Abstract as well.

2) Lines 143-145: “The EQUATES hemispheric simulations therefore include losses of O3 over seawater that are not present in the PA hemispheric simulations which could affect O3 transported over the Pacific in particular.”
Can you elaborate more on which source-specific US background O3 bias (for example, LINTL? Or some other scenario is more relevant to this?) is this effect the most here (describe using Labels in Table S2)?

3) Lines 174-185: “For hemispheric-scale simulations, biogenic VOC emissions are from the Model of Emissions of Gases and Aerosols from Nature version 2.1 (MEGAN2.1) (Guenther et al., 2012). The PA simulations additionally replace MEGAN emissions with emissions from the Biogenic Emission Inventory System (BEIS) (Bash et al., 2016) over North America (US EPA, 2019a).. The EQUATES MEGAN emissions are obtained from a compilation by Sindelarova et al. (2014). Soil NOx emissions for the PA hemispheric simulations are also from MEGAN with replacement by BEIS soil NOx over North America. Soil NOx emissions for the hemispheric EQUATES simulations are from a dataset by the Copernicus Atmosphere Monitoring Service (CAMS, 2018) based on methods by Yienger and Levy (1995). Lightning NO emissions for both the PA and EQUATES hemispheric simulations are from monthly climatology obtained from the Global Emissions Initiative (GEIA) and are based on Price et al. (1997). Lightning NOx was not included in the PA continental-scale simulations, while lightning NOx for the EQUATES continental-scale simulations is calculated using an inline module in CMAQ (Kang et al., 2019).”
In the above manuscript text, the authors highlight the differences in emission inputs for the PA versus EQUATES simulations. Can the authors enunciate/comment more on the impact of these differences on the model ozone biases, as per the scope of the manuscript, say: 1) with (Continental-scale EQUATES) and without Lightning NOx (Continental-scale PA), 2) Different biogenic VOC and NOx (Soil) emissions used in different simulations. For instance, impact of difference in Soil NOx emission inputs on USB O3 bias might be inferred from time series analysis in growing season (say, JJA) in agricultural regions such as, Central Valley and US Midwest.
The impact (or possible impact) of these differences (biogenic emission inputs being different between simulations) needs to be linked more clearly to the findings in the results and/or Conclusions. Also, a small typo here: North America (US EPA, 2019a). (please review and correct of any other such typos throughout the manuscript)

Hide

ED: Reconsider after major revisions (12 Aug 2024) by Jason Williams

AR by Nash Skipper on behalf of the Authors (20 Sep 2024) Author's response Author's tracked changes Manuscript

ED: Publish as is (02 Oct 2024) by Jason Williams

AR by Nash Skipper on behalf of the Authors (05 Oct 2024) Author's response Manuscript

Journal article(s) based on this preprint

26 Nov 2024

Source-specific bias correction of US background and anthropogenic ozone modeled in CMAQ

T. Nash Skipper, Christian Hogrefe, Barron H. Henderson, Rohit Mathur, Kristen M. Foley, and Armistead G. Russell

Geosci. Model Dev., 17, 8373–8397, https://doi.org/10.5194/gmd-17-8373-2024,https://doi.org/10.5194/gmd-17-8373-2024, 2024

Short summary

T. Nash Skipper, Christian Hogrefe, Barron H. Henderson, Rohit Mathur, Kristen M. Foley, and Armistead G. Russell

Supplement

https://doi.org/10.5194/egusphere-2024-554-supplement

T. Nash Skipper, Christian Hogrefe, Barron H. Henderson, Rohit Mathur, Kristen M. Foley, and Armistead G. Russell

Viewed

Total article views: 507 (including HTML, PDF, and XML)

HTML	PDF	XML	Total	Supplement	BibTeX	EndNote
381	95	31	507	45	26	28

HTML: 381
PDF: 95
XML: 31
Total: 507
Supplement: 45
BibTeX: 26
EndNote: 28

Views and downloads (calculated since 25 Mar 2024)

Month	HTML	PDF	XML	Total
Mar 2024	87	13	3	103
Apr 2024	109	29	9	147
May 2024	24	11	3	38
Jun 2024	57	13	6	76
Jul 2024	21	10	3	34
Aug 2024	31	5	2	38
Sep 2024	18	7	2	27
Oct 2024	12	3	1	16
Nov 2024	22	4	2	28

Cumulative views and downloads (calculated since 25 Mar 2024)

Month	HTML	PDF	XML	Total
Mar 2024	87	13	3	103
Apr 2024	109	29	9	147
May 2024	24	11	3	38
Jun 2024	57	13	6	76
Jul 2024	21	10	3	34
Aug 2024	31	5	2	38
Sep 2024	18	7	2	27
Oct 2024	12	3	1	16
Nov 2024	22	4	2	28

Viewed (geographical distribution)

Total article views: 493 (including HTML, PDF, and XML) Thereof 493 with geography defined and 0 with unknown origin.

Country	#	Views	%

Latest update: 26 Nov 2024

Download

The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.

Preprint (3141 KB)
Metadata XML

Short summary

Chemical transport model simulations are combined with ozone observations to estimate the bias in ozone attributable to US anthropogenic sources as well as several individual sources of US background ozone: natural sources, non-US anthropogenic sources, and stratospheric ozone. Results indicate a positive bias correlated with US anthropogenic emissions during summer in the eastern US and a negative bias correlated with stratospheric ozone during spring.


Total:	0
HTML:	0
PDF:	0
XML:	0