the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 20 Jul 2022
Assessment of tropospheric ozone products from CAMS reanalysis and near-real time analysis using observations over Iran
Abstract. Tropospheric ozone time series consists of the effects of various scales of motion, from meso to large time scales which is often challenging for global models to capture them. This study uses two global datasets, namely reanalysis and analysis of the Copernicus atmospheric Monitoring Service (CAMS), to assess the capability of these models or systems in presenting ozone’s features in small scales. We employ the tropospheric ozone product of the models and in situ measurements at 18 stations over Iran for the year of 2020. Furthermore, we make use of data of ozone, temperature, nitrogen oxides, wind speed, and wind direction at one more station. We decompose the datasets into three spectral components, i.e., short (S), medium (M), and long (L) term. We only evaluate the S and M terms of modelled against those of observed datasets for all stations. We examine the relationship between ozone and the relevant proxies. Results show a correlation coefficient of larger than 0.5 for S and about 0.25 for M term in both models. It turns out that the reanalysis dataset demonstrates more precision for the S component than that for the M. Both models can show the observed correlation between ozone and temperature, whereas some inconsistencies appear in presenting the anti-correlation between ozone and nitrogen oxides.
- Preprint
(1435 KB) - Metadata XML
- BibTeX
- EndNote
Status: closed
-
RC1: 'Comment on gmd-2022-138', Anonymous Referee #1, 05 Sep 2022
The authors present an ambitious study to analyze the information content of surface ozone over Iran (particularly Tehran) in two CAMS-based data products: the CAMS reanalysis, and the CAMS NRT analysis. For this the authors make use of in-situ observations of ozone, along with other chemical and meteorological quantities at various stations across Iran, but particularly in the cities Tehran and Tabriz.
We value their efforts in studying the usefulness of CAMS data for their applications, and their investigation in mathematical methods to analyze, and differentiate errors of various kinds. Reporting on this progress, and its application to locations across the world that may often be overlooked, is much appreciated. Nevertheless, I believe there are several serious flaws which makes it hard to extract useful information from this study, which unfortunately makes me advise to reject this manuscript for publication in ACP, before these aspects are addressed.
My main criticism concerns the plain application of mathematical methods to differentiate and analyze the representativity of various time scales in the data products: Prior to applying these methods, in my view the authors should better justify the selection of input data for this, both on observation and modeling side:
- from the observation side, I notice that the majority of observation stations are very close to one another, i.e. inside the city of Tehran. This makes that only a single grid-box in the global (re-)analysis data is used (and one at Tabriz station). Therefore the statistics is difficult to interpret for the whole of Iran, as suggested in the title of the manuscript. A closer assessment, and possibly corresponding filtering, of the observations that can be used to better represent ozone variability on the scales of the global model, e.g. by using a site classification method (urban / rural / etc) could be beneficial. E.g. the authors should better describe the ‘Geophysics’ station (line 82). A map with distribution of stations across Iran (and/or Tehran) would be beneficial.
- Such assessment of the observation network is the more important because the global (re-)analysis datasets definitely lack ability to describe (ozone) pollution levels on a city-scale, as also remarked by the authors. Therefore this analysis should be done with care, and should be discussed. Also it is not 100% clear which model level is selected, and what is the rationale for selecting a model level above the surface for comparison against surface observations. Such an approach is useful for observations that are situated on (isolated) hill / mountain tops, but otherwise this can add to large discrepancies. It might be best to simply select the surface layer for this assessment.
- The authors should better explain the status of the two CAMS datasets. Different to what the authors describe, I think these can best be described as a Reanalysis, and a NRT Analysis dataset, with various similarities (based on very similar input datasets, and data assimilation systems), but with their differences (in terms of model resolution, chemistry, input emissions). The statement on line 71 (“CAMS provides global forecasts, called analysis”) is confusing in this respect.
- Also the selection of additional quantities that explain the variability in O3, as discussed in Sec. 3.3 (and eqn 6) appears ad-hoc, and it is not clear where these quantities are taken from (also from CAMS model data, or observed?) . E.g. why is radiation (or solar zenith angle / cloud cover) not included as a proxy to explain variability in O3?
Furthermore, the authors present statistics that is difficult to interpret. E.g. their Figure 1 the diurnal cycle plots are illegible, while the Figure 1(d) better should include longer time series. The units in expressions 1-4 are not specified, nor the meaning of variable ‘p’. The naming convention dfo, dfa, fdr (line 90) is needlessly complex for the reader. (why ‘df’ in the first place?). The Table 1 shows numbers for the regression coefficients that cannot be inter compared as such, as they apparently depend on the mean variability, while it is unclear what is the meaning of the second number in the column. Also it is unclear why the authors select station ‘8’ for this analysis. To my judgement the study would benefit from presentation of plain intercomparison of model-observation time series, and corresponding statistics, to justify and analyze these more advanced mathematical methods.
Citation: https://doi.org/10.5194/gmd-2022-138-RC1 -
AC1: 'Reply on RC1', Najmeh Kaffashzadeh, 05 Oct 2022
The comment was uploaded in the form of a supplement: https://gmd.copernicus.org/preprints/gmd-2022-138/gmd-2022-138-AC1-supplement.pdf
-
RC2: 'Comment on gmd-2022-138', Anonymous Referee #2, 11 Sep 2022
The present study applies an interesting approach to decompose time series of surface ozone concentrations over two cities of Iran for two CAMS atmospheric composition global products and observations. Although I find the idea quite interesting, I see essential limitations in the design and application, which make me suggest rejection of publication of the present version of the study in GMD. Please find below my comments:
- My main concern is that the study is quite regional. A country-based focus is itself regional, especially when the study presents results for only two cities. As the other Reviewer also stated, you present results for only two grid-points (at least for CAMS reanalysis), a fact that is not helping in drawing safe conclusions. What is the scientific significance/interest of a paper assessing CAMS O3 (global) products over only two cities? My suggestion is to spatially extend the focus of the study including more stations from Iran (if available) and/or other neighboring countries to extract more robust findings and increase the scientific interest of the paper. Something also interesting would be to apply your approach over Europe and include in the analysis the CAMS-regional forecast of atmospheric composition exploring also the added value that CAMS regional models bring (?).
- The differences in the S term are mostly attributed to chemistry. What about deposition? Uncertainties in emission inventories? Stratospheric ozone contribution is not referred at all, yet, the broader Iran region is a well-known hot spot of stratosphere-to-troposphere transport that might affect day-to-day O3 variability in some cases. More discussion is needed in the interpretation of the results.
- P8L226-228: Both CAMS reanalysis and CAMS NRT are assimilated products. More details are needed here about the differences in assimilation process between the two products that might be related to the diferences in the performance of dfr and dfa.
- What data are used for the MLR models? Do you use the reanalysis and analysis NOx, AT, WS, WD data for the MLR of the respective dfr and dfa?
- P3L67: “Despite the well performance of CAMS in the upper troposphere,..”. This is not the case according to recent evaluation studies like those of Inness et al. (2019), Huijnen et al. (2020), Wagner et al. (2021) and Akritidis et al. (2022). CAMS reanalysis ozone in the upper troposphere is found biased and this is probably associated with the assimilation process. This is something that needs to be noted in the manuscript.
Akritidis, D., Pozzer, A., Flemming, J., Inness, A., Nédélec, P., and Zanis, P.: A process-oriented evaluation of CAMS reanalysis ozone during tropopause folds over Europe for the period 2003–2018, Atmos. Chem. Phys., 22, 6275–6289, https://doi.org/10.5194/acp-22-6275-2022, 2022
Huijnen, V., Miyazaki, K., Flemming, J., Inness, A., Sekiya, T., and Schultz, M. G.: An intercomparison of tropospheric ozone reanalysis products from CAMS, CAMS interim, TCR-1, and TCR-2, Geosci. Model Dev., 13, 1513–1544, https://doi.org/10.5194/gmd-13-1513-2020, 2020
Inness, A., Ades, M., Agustí-Panareda, A., Barré, J., Benedictow, A., Blechschmidt, A.-M., Dominguez, J. J., Engelen, R., Eskes, H., Flemming, J., Huijnen, V., Jones, L., Kipling, Z., Massart, S., Parrington, M., Peuch, V.-H., Razinger, M., Remy, S., Schulz, M., and Suttie, M.: The CAMS reanalysis of atmospheric composition, Atmos. Chem. Phys., 19, 3515–3556, https://doi.org/10.5194/acp-19-3515-2019, 2019
Wagner, A., Bennouna, Y., Blechschmidt, A.-M., Brasseur, G., Chabrillat, S., Christophe, Y., Errera, Q., Eskes, H., Flemming, J., Hansen, K. M., Inness, A., Kapsomenakis, J., Langerock, B., Richter, A., Sudarchikova, N., Thouret, V., and Zerefos, C.: Comprehensive evaluation of the Copernicus Atmosphere Monitoring Service (CAMS) reanalysis against independent observations: Reactive gases, Elementa: Science of the Anthropocene, 9, 00171, https://doi.org/10.1525/elementa.2020.00171, 2021
Citation: https://doi.org/10.5194/gmd-2022-138-RC2 -
AC2: 'Reply on RC2', Najmeh Kaffashzadeh, 05 Oct 2022
The comment was uploaded in the form of a supplement: https://gmd.copernicus.org/preprints/gmd-2022-138/gmd-2022-138-AC2-supplement.pdf
Status: closed
-
RC1: 'Comment on gmd-2022-138', Anonymous Referee #1, 05 Sep 2022
The authors present an ambitious study to analyze the information content of surface ozone over Iran (particularly Tehran) in two CAMS-based data products: the CAMS reanalysis, and the CAMS NRT analysis. For this the authors make use of in-situ observations of ozone, along with other chemical and meteorological quantities at various stations across Iran, but particularly in the cities Tehran and Tabriz.
We value their efforts in studying the usefulness of CAMS data for their applications, and their investigation in mathematical methods to analyze, and differentiate errors of various kinds. Reporting on this progress, and its application to locations across the world that may often be overlooked, is much appreciated. Nevertheless, I believe there are several serious flaws which makes it hard to extract useful information from this study, which unfortunately makes me advise to reject this manuscript for publication in ACP, before these aspects are addressed.
My main criticism concerns the plain application of mathematical methods to differentiate and analyze the representativity of various time scales in the data products: Prior to applying these methods, in my view the authors should better justify the selection of input data for this, both on observation and modeling side:
- from the observation side, I notice that the majority of observation stations are very close to one another, i.e. inside the city of Tehran. This makes that only a single grid-box in the global (re-)analysis data is used (and one at Tabriz station). Therefore the statistics is difficult to interpret for the whole of Iran, as suggested in the title of the manuscript. A closer assessment, and possibly corresponding filtering, of the observations that can be used to better represent ozone variability on the scales of the global model, e.g. by using a site classification method (urban / rural / etc) could be beneficial. E.g. the authors should better describe the ‘Geophysics’ station (line 82). A map with distribution of stations across Iran (and/or Tehran) would be beneficial.
- Such assessment of the observation network is the more important because the global (re-)analysis datasets definitely lack ability to describe (ozone) pollution levels on a city-scale, as also remarked by the authors. Therefore this analysis should be done with care, and should be discussed. Also it is not 100% clear which model level is selected, and what is the rationale for selecting a model level above the surface for comparison against surface observations. Such an approach is useful for observations that are situated on (isolated) hill / mountain tops, but otherwise this can add to large discrepancies. It might be best to simply select the surface layer for this assessment.
- The authors should better explain the status of the two CAMS datasets. Different to what the authors describe, I think these can best be described as a Reanalysis, and a NRT Analysis dataset, with various similarities (based on very similar input datasets, and data assimilation systems), but with their differences (in terms of model resolution, chemistry, input emissions). The statement on line 71 (“CAMS provides global forecasts, called analysis”) is confusing in this respect.
- Also the selection of additional quantities that explain the variability in O3, as discussed in Sec. 3.3 (and eqn 6) appears ad-hoc, and it is not clear where these quantities are taken from (also from CAMS model data, or observed?) . E.g. why is radiation (or solar zenith angle / cloud cover) not included as a proxy to explain variability in O3?
Furthermore, the authors present statistics that is difficult to interpret. E.g. their Figure 1 the diurnal cycle plots are illegible, while the Figure 1(d) better should include longer time series. The units in expressions 1-4 are not specified, nor the meaning of variable ‘p’. The naming convention dfo, dfa, fdr (line 90) is needlessly complex for the reader. (why ‘df’ in the first place?). The Table 1 shows numbers for the regression coefficients that cannot be inter compared as such, as they apparently depend on the mean variability, while it is unclear what is the meaning of the second number in the column. Also it is unclear why the authors select station ‘8’ for this analysis. To my judgement the study would benefit from presentation of plain intercomparison of model-observation time series, and corresponding statistics, to justify and analyze these more advanced mathematical methods.
Citation: https://doi.org/10.5194/gmd-2022-138-RC1 -
AC1: 'Reply on RC1', Najmeh Kaffashzadeh, 05 Oct 2022
The comment was uploaded in the form of a supplement: https://gmd.copernicus.org/preprints/gmd-2022-138/gmd-2022-138-AC1-supplement.pdf
-
RC2: 'Comment on gmd-2022-138', Anonymous Referee #2, 11 Sep 2022
The present study applies an interesting approach to decompose time series of surface ozone concentrations over two cities of Iran for two CAMS atmospheric composition global products and observations. Although I find the idea quite interesting, I see essential limitations in the design and application, which make me suggest rejection of publication of the present version of the study in GMD. Please find below my comments:
- My main concern is that the study is quite regional. A country-based focus is itself regional, especially when the study presents results for only two cities. As the other Reviewer also stated, you present results for only two grid-points (at least for CAMS reanalysis), a fact that is not helping in drawing safe conclusions. What is the scientific significance/interest of a paper assessing CAMS O3 (global) products over only two cities? My suggestion is to spatially extend the focus of the study including more stations from Iran (if available) and/or other neighboring countries to extract more robust findings and increase the scientific interest of the paper. Something also interesting would be to apply your approach over Europe and include in the analysis the CAMS-regional forecast of atmospheric composition exploring also the added value that CAMS regional models bring (?).
- The differences in the S term are mostly attributed to chemistry. What about deposition? Uncertainties in emission inventories? Stratospheric ozone contribution is not referred at all, yet, the broader Iran region is a well-known hot spot of stratosphere-to-troposphere transport that might affect day-to-day O3 variability in some cases. More discussion is needed in the interpretation of the results.
- P8L226-228: Both CAMS reanalysis and CAMS NRT are assimilated products. More details are needed here about the differences in assimilation process between the two products that might be related to the diferences in the performance of dfr and dfa.
- What data are used for the MLR models? Do you use the reanalysis and analysis NOx, AT, WS, WD data for the MLR of the respective dfr and dfa?
- P3L67: “Despite the well performance of CAMS in the upper troposphere,..”. This is not the case according to recent evaluation studies like those of Inness et al. (2019), Huijnen et al. (2020), Wagner et al. (2021) and Akritidis et al. (2022). CAMS reanalysis ozone in the upper troposphere is found biased and this is probably associated with the assimilation process. This is something that needs to be noted in the manuscript.
Akritidis, D., Pozzer, A., Flemming, J., Inness, A., Nédélec, P., and Zanis, P.: A process-oriented evaluation of CAMS reanalysis ozone during tropopause folds over Europe for the period 2003–2018, Atmos. Chem. Phys., 22, 6275–6289, https://doi.org/10.5194/acp-22-6275-2022, 2022
Huijnen, V., Miyazaki, K., Flemming, J., Inness, A., Sekiya, T., and Schultz, M. G.: An intercomparison of tropospheric ozone reanalysis products from CAMS, CAMS interim, TCR-1, and TCR-2, Geosci. Model Dev., 13, 1513–1544, https://doi.org/10.5194/gmd-13-1513-2020, 2020
Inness, A., Ades, M., Agustí-Panareda, A., Barré, J., Benedictow, A., Blechschmidt, A.-M., Dominguez, J. J., Engelen, R., Eskes, H., Flemming, J., Huijnen, V., Jones, L., Kipling, Z., Massart, S., Parrington, M., Peuch, V.-H., Razinger, M., Remy, S., Schulz, M., and Suttie, M.: The CAMS reanalysis of atmospheric composition, Atmos. Chem. Phys., 19, 3515–3556, https://doi.org/10.5194/acp-19-3515-2019, 2019
Wagner, A., Bennouna, Y., Blechschmidt, A.-M., Brasseur, G., Chabrillat, S., Christophe, Y., Errera, Q., Eskes, H., Flemming, J., Hansen, K. M., Inness, A., Kapsomenakis, J., Langerock, B., Richter, A., Sudarchikova, N., Thouret, V., and Zerefos, C.: Comprehensive evaluation of the Copernicus Atmosphere Monitoring Service (CAMS) reanalysis against independent observations: Reactive gases, Elementa: Science of the Anthropocene, 9, 00171, https://doi.org/10.1525/elementa.2020.00171, 2021
Citation: https://doi.org/10.5194/gmd-2022-138-RC2 -
AC2: 'Reply on RC2', Najmeh Kaffashzadeh, 05 Oct 2022
The comment was uploaded in the form of a supplement: https://gmd.copernicus.org/preprints/gmd-2022-138/gmd-2022-138-AC2-supplement.pdf
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 675 | 168 | 26 | 869 | 14 | 17 |
- HTML: 675
- PDF: 168
- XML: 26
- Total: 869
- BibTeX: 14
- EndNote: 17
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1