the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 19 Jan 2024
An improved representation of aerosol acidity in the ECMWF IFS-COMPO 49R1 through the integration of EQSAM4Climv12
Abstract. The atmospheric composition forecasting system used to produce the CAMS forecasts of global aerosol and trace gases distributions, IFS-COMPO, undergoes periodic upgrades. In this paper we describe the development of the future operational cycle 49R1, and focus on the implementation of the thermodynamical model EQSAM4Clim version 12 for describing gas-aerosol partitioning processes for nitrate and ammonium and for providing diagnostic aerosol, cloud and precipitation pH values at global scale. This information on aerosol acidity influences tropospheric chemistry processes associated with aqueous phase chemistry and wet deposition. The other updates to cycle 49R1 include modifications to the description of Desert Dust, Sea-salt aerosols, Carbonaceous aerosols and the size description for the calculation of aerosol optics. The implementation of EQSAM4Clim significantly improves the partitioning of reactive nitrogen compounds decreasing surface concentrations of both nitrate and ammonium, which reduces PM2.5 biases for Europe, U.S. and China, especially during summertime. For aerosol optical depth there is generally a decrease in the simulated biases for wintertime, and for some regions an increase in the bias for summertime. Improvements in the simulated Ångström exponent is noted for almost all regions, resulting in generally a good agreement with observations. The diagnostic aerosol and precipitation pH calculated by EQSAM4Clim have been compared against results from previous simulations (for aerosol pH) and against ground observations (for precipitation pH), with the temporal distribution in the annual mean values showing good agreement against the regional observational datasets. The use of aerosol acidity only has a relatively smaller impact on the aqueous-phase production of sulphate when compared to the changes in gas-to-particle partitioning brought by the use of EQSAM4Clim.
- Preprint
(5817 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on egusphere-2023-3072', Anonymous Referee #1, 15 Feb 2024
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2024/egusphere-2023-3072/egusphere-2023-3072-RC1-supplement.pdf
-
AC1: 'Partial reply on RC1 - simulation data', Samuel Remy, 20 Feb 2024
Dear reviewer,
Thanks a lot for your review. We'll adapt the manuscript to take into account your comments and remarks. In the meantime, the simulation data of the article has been put into a zenodo repository: https://zenodo.org/records/10679832
Kind regards,
Samuel Remy
Citation: https://doi.org/10.5194/egusphere-2023-3072-AC1
-
AC1: 'Partial reply on RC1 - simulation data', Samuel Remy, 20 Feb 2024
-
RC2: 'Comment on egusphere-2023-3072', Anonymous Referee #2, 09 Mar 2024
As a preamble I must say that the double standard about code availability in GMD is increasingly frustrating. Either GMD is a journal whose articles describe open-access models or it is not. I do not see the rationale for making an exception for articles describing models that are not open-access for institutional reasons (these definitely fall in the category of non-open-access models). This said the availability of the code of the EQSAM box model version 12 is welcomed.
This manuscript describes the improvements made to the representation of atmospheric aerosols in the ECMWF IFS-COMPO cycle 49R1 in comparison to previous cycles and with observations. In particular this new cycle includes the coupling to the EQSAM4Clim aerosol module for gas-particulate and a calculation of aerosol and precipitation pH.
Overall the manuscript is informative and brings useful information to potential users of the Copernicus aerosol products. Nevertheless the manuscript in its current form suffers from a few weaknesses:
1/ The model description is essentially qualitative (in that it lists the parametrization that have been assembled together) but lacks a more mathematical description of the parametrizations used and the associated numerical schemes. Anything that would provide further mathematical and algorithmic details would be welcomed in the revised manuscript.
2/ The language and presentation of the manuscript need to be greatly improved. There are many small issues to be fixed, including with tables and figures. The grammar needs to thoroughly checked. Regarding the figures, the labels are too small. The captions often lack details. And the color scales are inappropriate and difficult to read. Some of these issues should have been fixed before publication in EGUsphere.
3/ The novelty of this manuscript is to predict aerosol and precipitation pH. The pH of atmospheric aerosols is hardly measurable with state-of-the-art instrumentation so a comparison to direct observations is not possible. A comparison of the IFS model output with indirect observations shows a mixed bag but mostly an acidic bias. In contrast there are many measurements of precipitation pH. The authors focus on only three networks in US, Europe and South-East Asia but ignore other sources of data in Africa, India and some remote locations (e.g. Amsterdam Island). I strongly encourage the authors to look at data from the International Network to study Deposition and Atmospheric chemistry in AFrica (INDAAF) that has long-term measurements of precipitation pH (see https://indaaf.obs-mip.fr/measurements/precipitation/) and could potentially be very useful for evaluating the model performance. It is well known that precipitation pH is often not so acidic in regions with significant emission of soil particles. Precipitation pH over India is known to be not very acidic (doi: 10.4209/aaqr.2015.06.0423) or even alkaline due to crustal material neutralizing the acidity (doi: 10.1016/0004-6981(89)90476-9). Precipitation pH is also not very acidic in some places in Australia (doi: 10.1007/BF01056198). Therefore it is a bit surprising to see the most acidic precipitation over continents in desertic regions on Fig 13a. Further discussion of the model biases on acidity would be highly welcomed. The model captures some features of precipitation pH (eg the east-west gradient in North America) but still has many shortcomings.
Other major comments
Title: I am not sure this is the best title for the article that is broader than just about aerosol acidity.
The language of the abstract needs to be more accurate (see minor comments). I strongly encourage the authors to wordsmith the abstract and also discuss the biases in aerosol and precipitation pH.
Bibliography is a little shallow. Here are a few articles that would certainly deserve a citation: doi: 10.1021/acs.accounts.0c00303, 10.1038/s43247-021-00164-0, 10.1021/acs.jpca.8b10676 but I am sure there must be many other relevant papers.
The sentence “The code revisions that are integrated into the operational version of IFS-COMPO must satisfy the two conditions (one qualitative, one quantitative) that they bring the model closer to "physical" reality, i.e. that more processes and/or species are represented, and that they improve the skill scores against 40 observations” appears to ignore the fact that a particular model development may increase the physical consistency of a model but deteriorate the skill scores if the previous model version relied on error compensation. This is why there is a paradigm shift that consists in retuning the model physics every time a new code revision is made so that a particular model development is given a “fairer” chance to improve the skill scores. I see the current practice as a weakness of the model development process at ECMWF and in NWP centres in general.
The terms “rain pH” and “precipitation pH” seem to be used interchangeably. Can the authors clarify if they consider rain only (liquid precipitation) or precipitation (liquid and solid)? In some places cloud pH is also mentioned but no result is presented. Is cloud pH in scope or not for this manuscript?
Minor comments:
Lines 2, 40, 45, 76 & 568: paper => study
Line 4: as nitrate and ammonium are not in the gas phase per se, the sentence needs modifying.
Line 5: “THE global scale”
Line 5: What matters is whether aerosol acidity affects tropospheric chemistry *in the model*. Does it?
Line 9: nitrate and ammonium *in the particulate phase*
Line 16: brought => induced ? It is unclear what are the two quantities which are compared in this sentence.
Lines 42 and 45: acidity => aerosol acidity ? aerosol, cloud and precipitation acidity ?
Lines 74-75: this is also true of phosphorus deposition (doi: 10.1038/ncomms3934 and 10.1111/gcb.13766).
Line 84: delete “input” ?
Lines 84-85: The International Network to study Deposition and Atmospheric chemistry in AFrica (INDAAF) has long-term measurements of precipitation pH (see https://indaaf.obs-mip.fr/measurements/precipitation/) that could potentially be very useful for evaluating the model performance.
Line 94: “with three bins for each of these two species”
Line 116: consists => consist
Lines 123-125: “CY45R1 and earlier IFS cycles”, “provided to the aerosol scheme”
Line 137: lagrangian => Lagrangian
Line 156: T being a symbol for temperature, it should italicized.
Figure 1: please indicate the ion valences on the figure.
Line 172 & 188: dependant => dependent
Line 190: what do you mean by “domain-dependent” ? IFS is a global model so there is no simulation domain in the usual sense of the term. Maybe the authors mean “regionally-dependent” ?
Lines 194-195: “… contributions … are …”
Line 213: “ … and smaller than …”
Line 220: “but are both including” => “but both include”
Line 221, “which is used operationally”: isn’t all of IFS-COMPO used operationally?
Line 225: convection scavenges chemical species but also transports them upwards and the two processes are intrinsically coupled to each other. How is this dealt with in IFS-COMPO?
Line 226: what is meant by “precipitation fraction”? Is it the fraction of the grid-box where precipitation occurs? How does this vary with model resolution?
Line 231, “mixed clouds, ie for temperature below the freezing point”: this is a weird sentence as it is well known that 1/ liquid clouds can persist below freezing point and 2/ mixed clouds do not occur below say -40°C.
Line 232: subjected => subject
Line 235: “… to follow more closely the particle size dependency of Croft et al … This involved … ”
Line 241: what does “a measure” mean in this context?
Line 261: “… aging … slower than … lifetime”: a process can be slower than another process but it cannot be slower than a lifetime. A lifetime can be smaller than another lifetime.
Section 2.3.3: the reader is missing equations here and what are the many areas of positive results.
Line 280: it may be worth saying PM is a concentration, hence the multiplication by air density. It should be said that rho and the aerosol mixing ratios [] are taken in the surface layer.
Line 315: remove dot after 1.5, ie “level 1.5 AOD”
Line 317: Angstrom => Ångström, integrated => evaluated
Lines 358-359: bring together the references into a single parenthesis
Table 1: ion valence is missing.
Line 383: I am surprised you put India and China on a same footing when it comes to surface PM. Aren’t surface aerosol concentrations much larger over India than China, even back in 2019?
Table 2: please specify the wavelength for the AOD.
Line 387: an increases => an increase
Lines 419 & 476: c.f. has a specific meaning in English => change to “see Fig. 3” or “see Table 1”
Line 425: close parenthesis
Lines 435 & 445: table 2 => Table 2
Line 490: delete first occurrence of “both”
Table 3, caption: delete leading “A”
Line 505: concentration => concentrations
Line 511: is => are
Line 515: not well phrased
Line 516: is => are
Lines 524-525: “using ISORROPIA … and …” w
Line 528: the web link is not effective (at least it didn’t respond when I tried)
Line 532: as mentioned above, there is also a measurement network in Africa that would be very complementary to those used for this evaluation.
Line 552: “shown in Fig. 14”, word Fig is missing.
Line 565: are in closer => are closer
Line 589: doesn’t => does not
Please review the bibliography, eg lines 624 & 693 (remove capitals that are not necessary), line 633 (last author), line 645 (ion valence as exponent)
Figures: many of the labels are too small. Please enlarge axis legends, labels and ticklabels.
Figs. 2, 3, 5, 6, 12, 13: the bluish-greenish-greenish-yellowish color scale is particularly difficult to read. I doubt as well that it is legible by blind color readers. The authors are advised to redraw all plots with a better color scale.
Figs. 2, 3 & 11: please repeat the unit in the caption as it is hard to read from the figures themselves.
Fig. 4: please specify unit.
Figs. 9 and 10: please specify in the caption what the different panels are.
Fig 14c: please add the 1:1 line.
Citation: https://doi.org/10.5194/egusphere-2023-3072-RC2
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 231 | 62 | 14 | 307 | 9 | 9 |
- HTML: 231
- PDF: 62
- XML: 14
- Total: 307
- BibTeX: 9
- EndNote: 9
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1