the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 30 Jan 2024
Technical Note: Evaluation of the Copernicus Atmosphere Monitoring Service Cy48R1 upgrade of June 2023
Abstract. The Copernicus Atmosphere Monitoring Service (CAMS) is providing daily analyses and forecasts of the composition of the atmosphere, including the reactive gases such as O3, CO, NO2, HCHO, SO2, aerosol species and greenhouse gases. The global CAMS analysis system (IFS-COMPO) is based on the ECMWF Integrated Forecast System (IFS) for numerical weather prediction (NWP), and assimilates a large number of composition satellite products on top of the meteorological observations ingested in IFS. The CAMS system receives regular upgrades, following the upgrades of IFS. The last upgrade, Cy48R1, operational since 27 June 2023, was major with a large number of code changes, both for COMPO and for NWP. The main COMPO innovations include the introduction of full stratospheric chemistry, a major update of the emissions, of the aerosol model, including the representation of secondary organic aerosol, several updates of the dust life cycle and optics, inorganic chemistry in the troposphere, and the assimilation of VIIRS AOD and TROPOMI CO. The CAMS Cy48R1 upgrade was validated using a large number of independent measurement datasets, including surface in situ, surface remote sensing, routine aircraft and balloon and satellite observations. In this paper we present the validation results for Cy48R1 by comparing with the skill of the previous operational system (Cy47R3), with the independent observations as reference, for the period October 2022 to June 2023 during which daily forecasts from both cycles are available. Major improvements in skill are found for the ozone profile in the lower-middle stratosphere and for stratospheric NO2 due to the inclusion of full stratospheric chemistry. Stratospheric trace gases compare well with ACE-FTS observations between 10–200 hPa, with larger deviations between 1–10 hPa. The impact of the updated emissions is especially visible over East Asia and is beneficial for the trace gases O3, NO2, and SO2. The CO column assimilation is now anchored by IASI instead of MOPITT which is beneficial for most of the CO comparisons, and the assimilation of TROPOMI CO data improves the model CO field in the troposphere. In general the aerosol optical depth has improved globally, but the dust evaluation shows more mixed results. The results of the 47 comparisons are summarised in a score card, which shows that 83 % of the evaluation datasets show a neutral or improved performance of Cy48R1 compared to the previous operational CAMS system, while 17 % indicate a (slight) degradation. This demonstrates the overall success of this upgrade.
- Preprint
(26006 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on egusphere-2023-3129', Anonymous Referee #1, 25 Mar 2024
This manuscript documented the performance of the CAMS Cy48R1 in comparison with CAMS Cy47R3. This is an important work, and the paper is well structured. I only have a few minor comments.
Out of curiosity, what’s the meaning of Cy47R3, Cy48R1?
Line 49: add full name of NDACC, WMO-GAW, AERONET, IAGOS, ICOS, IASOA?
As noted by the authors, summer validation over the Northern Hemisphere (winter for the Southern Hemisphere) is missing. Please add some more discussion on this. Is the performance in the missing months expected to be similar to other months?
Do you have any assimilation plan for the geostationary satellites?
Line 230: “smaal” typo?
It will be helpful to have a table of all assimilated satellite products in CAMS in the previous and current version, and their assimilated period.
Figure 2: Why e-control – o-control is not shown over Asia like Figures 1 & 3?
Line 259-261: “At most airports worldwide the bias in the lower troposphere (pressure > 850 hPa) is slightly larger for the e-suite than for the o-suite, and in particular over airports located in Western Africa and Eastern Asia (not shown). Conversely in the free troposphere, the bias is smaller in the e-suite that in the o-suite for most visited airports (Eskes et al., 2023b).” Why is that the case? Overall, the paper focused on evaluating the performance of the e- suite relative to o- suite but often does not explain the differences.
Change “ozone sonde” to “ozonesonde”?
Line 556: change “china” to “China”
I like it that the author mentioned in the conclusion “CAMS is actively acquiring and testing (surface) data from South America, Africa, and Asian countries other than China.” I hope in the future more data from the Global South will be added in the evaluation.
The manuscript is very long with many details. While this might be the nature of such type of papers (Technical Note), it is not easy to follow the whole text for general readers. The abstract and conclusion are helpful because they listed main points.
Citation: https://doi.org/10.5194/egusphere-2023-3129-RC1 -
RC2: 'Comment on egusphere-2023-3129', Anonymous Referee #2, 15 Apr 2024
General comments:
The CAMS global forecasts for atmospheric composition are a major effort by the Earth observation component of the European Union’s Space programme – Copernicus - to include atmospheric chemistry and aerosols into the global NWP modelling and data assimilation framework, here fully integrated into the IFS modelling system at ECMWF. At the same time, it is a pioneer service to the public, providing various composition data in near-real time alongside with traditional weather prediction. Due to the complex nature of atmospheric chemistry and dynamics and its interplay with the earth surface, great importance must be attched to validation using independant observational data sets. In this manuscript, extensive evaluation has been performed with different types of observations (station networks, sondes, in-situ and remote sensing data). With the update to Cy48R1 several important innovations have been introduced, namely integration of full stratospheric chemistry, emission updates, representation of secondary organic aerosol, and assimilation of new satellite observations for AOD and CO.
Given the broad scope of the research topic, the pioneer nature of the service, and the relevace of the data provided by CAMS, the paper should be published in ACP. However, I suggest to take a few points into consideration before publishing:
- As the authors state in the introduction and in a disclaimer, the manuscript is based on an existing (preliminary) CAMS report (Eskes, 2023). Looking at the results section, more than 20% of the findings in this manuscript are not substantiated with own figures but refer to results presented in Eskes (2023). This may be a fair decision, given the length of the paper. But the reader should be able to easily verify the results. E.g., this can be achieved by adding additional figures to an appendix.
- While the general findings are extremely well summarized in the conclusion part, I was not able to fully verify all results, even if there are displayed in the figures. Please refer to my specific comments. A few times, I was also missing the reasoning behind updates or the explanation for differences.
- You also introduce results using the control runs of e-suite and o-suite. While this is nicely introduced in the system overview, later on it often remains unclear why these simulations are shown. Please revise all result sections, keep and explain results which can be achieved with the help of the control runs only and remove those results elsewhere. Ensure that the figures contain relevant information only, but try to enrich the statistical information with some overall values (see specific comments).
Specific comments:
Line 26: Add MACC reference, e.g., Marécal et al. (2015).
Line 32: As reference, you could point to: https://www.ecmwf.int/en/forecasts/documentation-and-support/changes-ecmwf-model
Line 34: To be more precise, change to "The assimilation of satellite data for atmospheric composition …“
Line 38: Remove "high-resolution".
Line 43: TROPOMI and VIIRS instruments need to be introduced, AOD abbreviation needs to be explained.
Lines 47/48: You should add here the work by Katragkou et al. (2015) on evaluation of near-surface ozone over Europe.
Lines 49/50: These organisations need to be introduced. You can do that by refering to Table 1.
Line 50: "real-time“ should be "near-real time“.
Line 51: Please change reference to ECMWF (2023d) to account for the series character.
Line 52: The validation report on the CAMS GHG reanalysis by Ramonet et al. (2021) has not been updated so far. Moreover, this reanalysis seems to have being stopped by the end of 2020. Please clarify or omit.
Line 62: The term "compo-suite“ is not used furtheron and can be omitted.
Lines 100-103: This important introduction into the nature of the control runs should be placed directly after its first mention (after line 92).
Lines 104-106: The relation of IFS-COMPO to IFS-GHG is not well introduced with the description of CAMS services. You can add one or two senteces in the introduction and skip this paragraph here.
Lines 108-109: The cycles have already been explained in the introduction. Skip the sentence here.
Line 118: BASCOE abbreviation needs to be explained here (done in line 124).
Line 125: 64 species are mentioned here to be used in BASCOE, in line 118/119 you talk about 123 active tracers to be included. Following Errera et al. (2019), 58 species are included. As the integration of a full stratospheric chemistry scheme is a major achievement of the latest CAMS update, please be more precise here.
Lines 170-107: MODIS, S-NPP, and NOAA20 need to be introduced.
Line 174: AERONET needs to be introduced (-> Table 1).
Line 194: You should add here that diurnal cycle profiles are partly applied to the emissions, depending on sector and species.
Lines 194-198: Are these emission datasets published and/or citeable? In the IFS documentation for Cy38R1 the following references are given: Granier et al. (2022) and Denier van der Gon (2021), which both refer to unspecified CAMS reports.
Lines 203-206: Latest validation results can be found in ECMWF(2023), while Benedictow et al. (2023) is only one report out of a series. The latter describes validation of Cy47R3. Please change accordingly.
Line 220: How do you define „main trace gases and aerosol“? I‘d argue that this selection is at least partly due to observational constraints.
Line 236: Don’t mention the control runs here as they are not shown in Fig.2.
Line 251: What does manual validation by the PI imply for IAGOS L1 data?
Line 253: For comparison with the figure, better stick to altitude here (e.g., 1.5 to 8 km).
Lines 288-289: I guess you mean "The e-suite shows similar correlations as the o-suite with values exceeding 0.6 for most stations“. I cannot confirm this statement. Correlations are varying from 0 to ~0.8. Please revise or clarify.
Lines 300-301: I can’t see any obvious differences in bias between e-suite and o-suite in Fig.8.
Lines 304-310: This refers to Fig. 8, bottom row (?). I can’t see the differences described here in Fig.8. Please revise the whole section.
Lines 319-321: Please mention already at the beginning of the sentence that you refer to the relative bias in December 2022 (Fig.9, third row). The bias south of 60°S is negative, not positive.
Lines 323-324: There are larger negative biases for e-suite in March and June 2023, particular over the Arctic Sea (March) and North America (June), which are neither described nor explained.
Line 331-343: CO partial columns are not shown in Fig.10. Most of the section describe these partial columns. It remains unclear, if Lines 340-343 refer to Fig.10.
Line 419: I guess that you mean here the introduction of full stratospheric chemistry. Please confirm.
Line 464: Changes in Nitrates AOD between e-suite and o-suite over southeast Asia and Sahel are in the same order than the absolute values for e-suite. Please revise.
Line 466: Add for AE: "… , a parameter which is indicative of the aerosol size distribution, …" and remove this part in line 480.
Lines 489-490: Network abbreviations have been explained before (or should have been).
Line 509: "SDA“ needs to be explained here (done in Fig.24).
Lines 514-516: Both, e-suite and o-suite perform moderately over North Africa. I would not set any preference from Fig. 24 (middle row).
Table 1: The link to https://www.eumetsat.int/iasi reveals "access denied“.
Table 2: What objective criteria were used to produce the relative scores for the evaluation? How is "small“ and "significant“ defined?
Figures 1,2,3,7,11,13,21,22,23: Always specify clearly the two datasets where your statistical evaluation is based on. E.g., "e-suite – observations“ in Fig.1, top row.
Figures 1,2,3,7,11,13,21,22,23: Mean value over all stations of the statistical measure shall be given as numbers in each panel, ideally together with standard deviation.
Figures 1,2,3,7,11,13,21,22,23: Background of the upper left corner in all panels is either light gray or light blue. Panel enumeration would be better visible with black colour letters.
Figure 10: "GB“ (Global Bias?) needs to be introduced or omitted. "GB_origpressure“ should be "pressure“?
Figure 17: Specify blue lines and black dots in the caption.
Figure 18: I assume that the stratospheric NO2 columns shown in the top row are the same than used for the differences between e-suite and TROPOMI. For the bottom row you mention the SREAM-B algorithm explicitly. Is there any reason for that?
Figure 19: Better use "o-suite“ in x-axis labeling.
Figure 20: DU, SS, OM, BC, SU, NI, AM need to be introduced.
Figure 23: Give MNMB and Correlation as titles to the top row. Here, it would be sufficient to show PM2.5 only.
Technical comments:
Line 144: Replace "… that used …“ by "… that is used …“.
Line 159: Put "since cycle 46R1“ in brackets.
Line 165: Add the year 2023 to the date.
Line 177: Replace "VarBC involves introducing …“ by "VarBC introduces …“.
Line 185: "and“ -> "an“
Line 214: "show“ -> "shown“
Line 231: "smaal“ -> "small“
Line 471: Replace "except dust“ by "except for dust“.
Line 486: Put "AeroVal (2023)“ in brackets.
Line 511: Remove , after "AE“.
Line 519: Replace "Capo Verde“ by "Cabo Verde“.
Figure 5: Caption: "covariancein“ -> "covariance in“
References:
Benedictow, A., Arola, A., Bennouna, Y., Bouarar, I., Cuevas, E., Errera, Q., Eskes, H., Griesfeller, J., Basart, S., Kapsomenakis, J., Lange[1]rock, B., Mortier, A., Pison, I., Pitkänen, M., Ramonet, M., Richter, A., Schoenhardt, A., Schulz, M., Tarniewicz, J., Thouret, V., Tsik[1]erdekis, A., Warneke, T., and Zerefos, C.: Validation report of the CAMS near-real-time global atmospheric composition service, Period December – February 2023, https://doi.org/10.24380/i31d-5i54, 2023.
ECMWF: CAMS global validation services, https://atmosphere.copernicus.eu/global-services, 2023.
Errera, Q., Chabrillat, S., Christophe, Y., Debosscher, J., Hubert, D., Lahoz, W., Santee, M. L., Shiotani, M., Skachko, S., von Clarmann, T., and Walker, K.: Technical note: Reanalysis of Aura MLS chemical observations, Atmos. Chem. Phys., 19, 13647–13679, https://doi.org/10.5194/acp-19-13647-2019, 2019.
Eskes, H., Tsikerdekis, A., Benedictow, A., Bennouna, Y., Blake, L., Bouarar, I., Errera, Q., Griesfeller, J., Ilic, L., Kapsomenakis, J., Langerock, B., Mortier, A., Pison, I., Pitkänen, M., Richter, A., Schönhardt, A., Schulz, M., Thouret, V., Warneke, T., and Zerefos, C.: Upgrade verification note for the CAMS near-real time global atmospheric composition service: Evaluation of the e-suite for the CAMS CY48R1 upgrade of 27 June 2023, https://doi.org/10.24380/rzg1-8f3l, 2023.
Katragkou, E., Zanis, P., Tsikerdekis, A., Kapsomenakis, J., Melas, D., Eskes, H., Flemming, J., Huijnen, V., Inness, A., Schultz, M. G., Stein, O., and Zerefos, C. S.: Evaluation of near-surface ozone over Europe from the MACC reanalysis, Geosci. Model Dev., 8, 2299–2314, https://doi.org/10.5194/gmd-8-2299-2015, 2015.
Marécal, V., Peuch, V.-H., Andersson, C., Andersson, S., Arteta, J., Beekmann, M., Benedictow, A., Bergström, R., Bessagnet, B., Cansado, A., Chéroux, F., Colette, A., Coman, A., Curier, R. L., Denier van der Gon, H. A. C., Drouin, A., Elbern, H., Emili, E., Engelen, R. J., Eskes, H. J., Foret, G., Friese, E., Gauss, M., Giannaros, C., Guth, J., Joly, M., Jaumouillé, E., Josse, B., Kadygrov, N., Kaiser, J. W., Krajsek, K., Kuenen, J., Kumar, U., Liora, N., Lopez, E., Malherbe, L., Martinez, I., Melas, D., Meleux, F., Menut, L., Moinat, P., Morales, T., Parmentier, J., Piacentini, A., Plu, M., Poupkou, A., Queguiner, S., Robertson, L., Rouïl, L., Schaap, M., Segers, A., Sofiev, M., Tarasson, L., Thomas, M., Timmermans, R., Valdebenito, Á., van Velthoven, P., van Versendaal, R., Vira, J., and Ung, A.: A regional air quality forecasting system over Europe: the MACC-II daily ensemble production, Geosci. Model Dev., 8, 2777–2813, https://doi.org/10.5194/gmd-8-2777-2015, 2015.
Ramonet, M., Langerock, B., Warneke, T., and Eskes, H.: Validation report of the CAMS greenhouse gas global reanalysis, years 2003-2020,
https://doi.org/10.24380/438C-4597, 2021.Citation: https://doi.org/10.5194/egusphere-2023-3129-RC2
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 340 | 80 | 16 | 436 | 10 | 7 |
- HTML: 340
- PDF: 80
- XML: 16
- Total: 436
- BibTeX: 10
- EndNote: 7
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1