the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Sources of Surface O3 in the UK: Tagging O3 within WRF-Chem
- 1School of Environmental Sciences, University of East Anglia, Norwich, UK
- 2Institute for Advanced Sustainability Studies (IASS), Potsdam, Germany
- 3Centre for Atmospheric Sciences, School of Earth, Atmospheric and Environmental Sciences, University of Manchester, Manchester, UK
- 4Institute of Environmental Science and Technology (ICTA), Universitat Autònoma de Barcelona, Barcelona, Spain
- 5Centre for Atmospheric Science, Department of Chemistry, University of Cambridge, Cambridge, UK
- 6Freie Universität Berlin, Institut für Meteorologie, Berlin, Germany
- anow at: Department of Chemistry, University of Colorado Boulder, Boulder, USA
- bnow at: Cooperative Institute for Research in Environmental Sciences (CIRES), University of Colorado Boulder, Boulder, CO, USA
- cnow at: NOAA Global Systems Laboratory (GSL), Boulder, CO, USA
- 1School of Environmental Sciences, University of East Anglia, Norwich, UK
- 2Institute for Advanced Sustainability Studies (IASS), Potsdam, Germany
- 3Centre for Atmospheric Sciences, School of Earth, Atmospheric and Environmental Sciences, University of Manchester, Manchester, UK
- 4Institute of Environmental Science and Technology (ICTA), Universitat Autònoma de Barcelona, Barcelona, Spain
- 5Centre for Atmospheric Science, Department of Chemistry, University of Cambridge, Cambridge, UK
- 6Freie Universität Berlin, Institut für Meteorologie, Berlin, Germany
- anow at: Department of Chemistry, University of Colorado Boulder, Boulder, USA
- bnow at: Cooperative Institute for Research in Environmental Sciences (CIRES), University of Colorado Boulder, Boulder, CO, USA
- cnow at: NOAA Global Systems Laboratory (GSL), Boulder, CO, USA
Abstract. Tropospheric ozone (O3) concentrations are known to depend on a combination of hemispheric, regional, and local-scale processes. Estimates of how much O3 is produced locally vs. transported from further afield are essential in air quality management and regulatory policies. Here, a tagged-ozone mechanism within the WRF-Chem model is used to quantify the contributions to surface O3 in the UK from anthropogenic nitrogen oxide (NOx) emissions from inside and outside the UK during May–August 2015. The contribution of the different source regions to three regulatory O3 metrics is also examined. It is shown that model simulations predict the concentration and spatial distribution of domain-wide surface O3 with a mean bias of -3.7 ppbv. Anthropogenic NOx emissions from the UK and Europe account respectively for 13 % and 16 % of the monthly mean surface O3 in the UK, as the majority (71 %) of O3 comes from the hemispheric background. The north and the west of the UK experience the largest contributions from hemispheric O3 with peaks in May, whereas European and UK contributions are most significant in the east and south-east, intensifying towards June and July. It is demonstrated that more stringent emission controls over continental Europe, particularly in western Europe, would be necessary to improve health-related metrics, such as MDA8 O3 above 50 and 60 ppbv. Emission controls over larger areas, e.g., the northern hemisphere, are instead required to lessen the impacts on ecosystems as quantified by metrics such as the AOT40.
- Preprint
(2460 KB) -
Supplement
(1194 KB) - BibTeX
- EndNote
Johana Romero-Alvarez et al.
Status: final response (author comments only)
-
RC1: 'Comment on acp-2022-173', Anonymous Referee #1, 08 May 2022
This paper describes a modeling study to investigate the sources of ozone over the UK in the spring-summer period in 2015 using a tagged approach. It is a competent study using an established technique, and while the results are not unexpected, they provide a valuable quantification of source contributions that constitute one of the first available in the literature. In particular, the study highlights the importance of sources outside the region in influencing ozone, and provides a thorough quantification of local and regional contributions across different parts of the UK. The finding that different measures would need to be taken to address ozone as represented by the MDA8 and AOT40 metrics is interesting, and this finding could be exploited better in the paper. It also feels as though model evaluation has been skipped over lightly, and inclusion of a brief assessment to convince the reader of the quality of the model simulations would strengthen the paper. Once these issues have been addressed, along with the points below, I feel that the manuscript would make a valuable addition to the literature and would be suitable for publication in ACP.
General CommentsAn original aspect of this study is consideration of impacts over different parts of the UK and using a number of different ozone metrics. Neither of these aspects is fully exploited in the results/discussion section, however. Which regions matter most from a population exposure perspective, for example? Which regions are currently close to regulatory limits? The exploration of different metrics is interesting, but how sensitive are the results likely to be to the meteorology in 2015? Sources in BEL/LUX/NET/GER may be more important than FRA in other years. Some consideration of these issues is needed.
Evaluation of the model simulation is consigned to the supplement, but I feel that something is needed in the paper to convince the reader that the model is up to the task, particularly given that "a good representation of O3 in the European domain" is expressly stated in the conclusions. Please adapt the existing section 2.4 to provide a more quantitative summary of the model performance, particularly for O3 and NOx.
It would also be useful to show a 4-month timeseries of ozone at at least one location to demonstrate the seasonal and diurnal variability (this could be hourly ozone or alternatively daily MDA8). This is important to show the relative importance of episodes, which are investigated in the latter part of the study.
While the manuscript presents a case study from 2015, it would be valuable to speculate on how general the results are likely to be for other years.
Specific CommentsLine 40: narrow concentration window: this might be rephrased, as three orders of magnitude isn't particularly narrow.
Line 49: "European" -> "UK and European"
Line 56: Reductions in European NOx emissions would be expected to give a reduction in rural ozone concentrations in the UK, as this is far from the source region.
Line 65: As stated, tagged-ozone methods are better than perturbation approaches for attribution studies quantifying the contribution of different sources at a given place/time. However, they are less well suited for quantifying the effect of emission controls which involve changing sources (which is how this concept was introduced in line 60). Some rephrasing is needed to avoid undermining the approach adopted here.
Line 75: Is the tagged ozone mechanism used here existing or new? Please make any novel aspects of the current study clear.
Line 107-8: It would be helpful to add a sentence here to suggest why nudging led to poorer simulations.
Line 126: "The method used here is based on...." Is the Lupascu and Butler approach used here directly or are there any developments or changes in implementation? It is important to be clear about the scientific contributions of the present study. Is any element of this new?
Line 136: How important is reentry of ozone into the model domain likely to be?
Line 151: This sentence does not describe how the contribution of tagged O3 to AOT40 was calculated, it just describes how AOT40 is calculated.
Line 156: Equation 1 is incorrect: max(O3-40, 0)
Note that this is summed over specific hours, not all hoursFigures 7 and 8 show the same variable (O3 chemical production) and it would be helpful to combine them so that they can be compared more easily.
Figures 9-12: It is not clear that all four figures are required; presenting results for two contrasting months would be sufficient, with the others placed in the supplement. Note that use of contrasting color palettes would allow the reader to separate the inset pie chart more easily, and that separating the legend into two sections would make interpretation of the charts easier.
Figures 13-15 could also be presented a lot more clearly, ideally with the panels arranged in a more geographically-intuitive layout. Flipping x and y axes would make the figures easier to read (so key sectors LB and UK are first rather than bottom of the list), truncating the O3 axis at 25 or 30 would make values more readable, and coloring bars consistent with Figs 9-12 would make contributions stand out better.
Typos and minor issuesLine 88: is -> are
Line 94: citation error "G. a."
Line 100: citation format for Mar et al.
Line 147: stablished (also exceeds -> exceed)
Line 151: remove "concentration of"
Line 169: units needed for the mean bias
Line 201: remove "from"
Line 260: Remove subsection, as there is no 3.1.2
Line 321: Units on ozone mixing ratios
Line 385: positive and negative bias in what/where?
Line 506: The Romero-Alvarez reference is out of sequenceThe coastlines in Fig 1a are drawn at very low resolution, and the figure would look tidier if the resolution was improved. Consider adding the model grid to give the reader an indication of the model resolution.
Fig 6 caption: Closed up -> Close up
Data availability: key output data should be made available through a publicly accessible repository such as CEDA
Author contributions: A clearer statement of author contributions in needed.
Several entries in the reference list refer to discussion papers that are now published (e.g., Lupascu and Butler; Kuik et al.). Please update these.
Lines 798, 818: number not indicated in header, remove comment?
Supplement:S1.1: Person -> Pearson
p.5: particulatly -> particularly
p.6: Fig 5S-> Fig S5, Fig 4S -> Fig S4Most of the figures in the supplement are not of publication quality, and the timeseries in particular need to larger and more clearly labelled so that the comparison of measured and observed concentrations is clearer. In the spatial maps (Figs S6-S8) the results would be much clearer if a more appropriate color scale was used for the difference plots (ideally dichromatic).
I do not find the composition comparison very convincing. While the analysis points to a number of model weaknesses, the causes remain unclear, so the comparison does not lend confidence in the performance of the model. While derived metrics, particulary those based on thresholds, are challenging to match well, I would have expected diurnal variation in NO, NO2 and O3 to be represented better.
-
RC2: 'Comment on acp-2022-173', Anonymous Referee #2, 17 May 2022
This paper describes the application of a regional chemical transport model using an ozone tagging scheme to quantify source contributions to surface tropospheric ozone in the UK during May-Aug 2015. The application of such a scheme in this context is novel, and the paper provides useful insight into the local, wider European, and extra-European contributions to ozone, broken down by local region within the UK. The paper explores differences in source contributions during episodes of higher surface ozone concentrations, and explores contributions using air quality and vegetation-relevant metrics, which provide some policy-relevant context. The paper is well written, and the methods applied appear robust and well described. There are some aspects of the model information and evaluation, and well as improvements in the discussion of results that would improve the manuscript. I recommend that once these issues (described below) are addressed, that the paper be published in ACP where it will be a valuable addition to the literature on European ozone air quality.
General comments
- For high ozone episodes in summer, biogenic emissions may be an important driver of ozone formation (e.g. see point made in Introduction on Page 7). Even if it is not possible to evaluate the model-simulated isoprene with observations, it might be informative to include a supplementary plot of isoprene during high ozone and more average conditions. The authors could also refer to previous studies evaluating MEGAN isoprene emissions in WRF-Chem, if relevant.
- Is it possible to calculate population-weighted MDA8 ozone contributions using population data and the model output? This would really strengthen the relevance of the results to air quality and human health. At the moment the discussion does not differentiate based on population distributions among the different regions, so it is difficult to interpret the relevance of the results to air quality.
- During ozone episodes (presented as when MDA8 O3 exceeds 50 or 60 ppbv), it would be informative to provide more in-depth discussion of meteorological conditions alongside the source region contributions. Are these periods dominated by anticyclonic conditions? What are the atmospheric transport pathways that dominate the France-sourced O3 influence on UK ozone? Are there any specific features that characterise the MDA8 > 60 ppbv episodes from the more moderate 50 ppbv exceedances?
Specific Comments
- Introduction - be more explicit about describing ozone production dependencies in NOx and VOC-limited conditions, and importance of NO+O3 in high NOx environment. This effect is variously referred to as ‘titration’ and ‘scavenging’. It would help the reader to point out the reaction specifically.
- Line 79: Not clear what is meant by “the second warmest year in a row in Europe”.
- Line 80: “EU information threshold of 1 hour (h) average mixing ratio of 180 μg m-3”: the value of 180 μg m-3 is a concentration not a mixing ratio. Is the threshold defined as the 90 ppbv mixing ratio, or the 180 μg m-3 concentration? These are not necessarily equivalent (dependent on local meteorological conditions).
- Line 97-99: Please clarify how the IC concentrations are applied. The phrase implies that they are used to initialise the model simulation at the outset, however the text implies that they are applied every 3 hours. Does this mean that the model fields are essentially overwritten with MOZART fields every 3 hours? Please clarify.
- Line 103: Presumably aerosol are also simulated in the model? Please provide information on the aerosol scheme used in the simulations.
- Line 169: Mean bias in μg m-3, ppb, or %? Please clarify.
- Figure S1 - Do you have an explanation for the lack of diurnal cycle in the model surface temperature at coastal sites? Does this imply issues regarding diurnal variation in mixing height / boundary layer? Is there any potential link to biases in the NOx and ozone shown? It would be helpful to expand more on some of these evaluations and comparisons in the main text.
- Fig. 3, 4, 6 captions: the plots depict mixing ratio, not concentration. Please change wording to reflect this.
Typographical errors:
Line 35: “Concentration of …” -> “The concentration of..’
Line 94: Erroneous “G. a.”?
Line 100: “shipping lines’ -> “shipping lanes”?
Johana Romero-Alvarez et al.
Johana Romero-Alvarez et al.
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 325 | 87 | 10 | 422 | 29 | 5 | 6 |
- HTML: 325
- PDF: 87
- XML: 10
- Total: 422
- Supplement: 29
- BibTeX: 5
- EndNote: 6
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1