Sensitivity of air quality indicators to emission inventories (EDGAR, EMEP, CAMS-REG) in Europe through FAIRMODE benchmarking methodology

de Meij, Alexander; Cuvelier, Cornelis; Thunis, Philippe; Pisoni, Enrico; Bessagnet, Bertrand

doi:https://doi.org/10.5194/egusphere-2023-1168

Preprints

https://doi.org/10.5194/egusphere-2023-1168

Preprints

29 Jun 2023

| 29 Jun 2023

Status: this preprint is open for discussion.

Sensitivity of air quality indicators to emission inventories (EDGAR, EMEP, CAMS-REG) in Europe through FAIRMODE benchmarking methodology

Alexander de Meij, Cornelis Cuvelier, Philippe Thunis, Enrico Pisoni, and Bertrand Bessagnet

Abstract. This study evaluates the sensitivity of four different emission inventories (EDGAR 5.0, EMEPG, CAMS version 2.2.1 and CAMS version 4.2 + condensables) on calculated air quality indices using the EMEP chemistry transport model. Emissions are reduced by 25 % and 50 % for different air pollutants for six cities and two regions in Europe to study the impact on particulate matter (PM) ozone (O3) formation. We performed the simulations for the year 2015 over Europe.

Emission at each location for all precursors are quite consistent for EMEP and CAMS inventories (as i.e. all these inventories share the same country total emissions), while EDGAR (being developed using a completely independent approach) generally show different urban scale emissions. Similar emission totals can however hide large variations in sectorial allocation. Our results stress the importance of the sectorial repartition of the emissions, given their different vertical distribution. In terms of non-linear behavior, the relationship between emission reduction and PM10 concentration change shows the largest non-linearity for NOx and in a lesser extend for NH3 whereas it remains mostly linear for the other precursors (VOC, SOx and PPM).

For O3, NOx emission reductions are the most efficient, likely because of the urban focus of this work and the abundance of NOx emission in this type of areas. In terms of non-linear behaviour, the relationship between emission reduction and O3 concentration change shows the largest non-linearity for NOx (concentration increase) and a quasi linear behaviour for VOC (concentration decrease). Potencies and potentials can show differences that are as large between inventories (EMEPC2 vs EMEPG) than between inventory versions (EMEPC2 vs. EMEPC42C). Precursor emission ratios (e.g. VOC/NOx for ozone or NOx/NH3 for PM10) show important differences among emission inventories. This emphasizes the importance of the accuracy of emission estimates since these differences can lead to changes of chemical regimes, directly affecting the responses of O3 or PM10 concentrations to emission reductions.

It is also important to understand that the choice of the indicator (for example mean or P95 values) can lead to different outcomes. It is therefore important to assess the variability of the results around the choice of the indicator to avoid misleading interpretations of the results.

This work takes part of the Forum for Air Quality Modelling (FAIRMODE), that provides air quality modellers a permanent forum to address air quality modelling issues (https://fairmode.jrc.ec.europa.eu).

Received: 31 May 2023 – Discussion started: 29 Jun 2023

Download & links

Preprint (PDF, 2285 KB)

Supplement (4362 KB)

Download & links

Alexander de Meij et al.

Status: open (extended)

Post a comment Subscribe to comment alert

RC1: 'Comment on egusphere-2023-1168', Anonymous Referee #1, 24 Jul 2023 reply

General Comments:
The authors conduct year-long model simulations over Europe with the EMEP CTM to intercompare four emission inventories in order to capture the uncertainty in FAIRMODE air quality metrics for PM and Ozone, focusing on emission control scenarios on specific urban centres and conglomerates. This is timely given the recent update of WHO guideline values, as well as the EU policy targets for improving air quality, and relevant within the scope of ACP (Methods for assessment of models).
The language throughout the manuscript should be improved to be made more fluent and precise, and avoid repetition.
In particular the abstract should be re-written to more clearly state the objectives and outcomes, avoid repetitions and include quantitative as well as qualitative comparisons.
The authors should make clear how this study is related and complementary to Thunis et al. (2022) where emission inventories for 150 cities are investigated?
Specific Comments:
In Sec. 2.2 it would be good for the reader if a table is added summarising the species present in each inventory and highlighting the differences, e.g. resolutions, bottom-up/country-totals methodology, compilation year etc.
In 2.2.4 could you please make more clear in the text how the "condensables" are different to previously reported PM2.5/10 and to quantify the expected differences, also for the examples of Poland and Turkey?
In Sec. 2.4 the screening method should be better motivated. With which criteria were the user-defined thresholds decided? The first 3 paragraphs read as a user guide for the FAIRMODE output diagram in Fig. 3 rather than a discussion on statistical screening - please consider rephrasing.
Sec. 3.1: What were the selection criteria for the cities, regions included in the study? Would it be possible to include additional locations (even to expand to all major European cities) to make the study even more encompassing and robust? (For example in Thunis et al. (2022) emission inventories for 150 cities are investigated - can this be done here or in a subsequent study to statistically assess the effectiveness and impacts of EU-wide measures?)
Sec. 3.2: Are the aerosol processes (secondary production) reported included and captured by the CTM model used? Are natural aerosols such as dust included in the modelled PM10?
Sec. 3.5 can be merged with Sec. 3.6: the ratio of NOx/VOC is important - are the urban centres studied here NOx/VOC limited in terms of O3 production/sink? Is there a different seasonal dependence in the model results from city to city that would be important for air quality plans?
Sec. 3.7: Given the difference in behaviour in specific regions how are the results of this study to be interpreted? Can the EMEP model provide a map of the different chemical regimes across European cities/regions for each inventory? What is the non-linear behaviour regarding reduction of both NOx and VOC?
Sec.4 Please consider using paragraphs rather than bullet points to present the findings.
Please clearly explain early in the text how is PMcoarse different to PM10? What is PPM and how is PPM10 and PPM2.5 different to PM10 and PM2.5? Please make sure all acronyms are properly defined and used.
P95 should be defined when first used.
Technical Corrections: (page, line number):
p3l53 newer, better -> more elaborate
p3l56 uncertainties associated to certain processes -> associated with
p3l63 One of FAIRMODE’s goal -> goals
p3l64 explain -> investigate
p4l121 What does it mean "only differ in terms of version"? Please clarify.
p13l434 Start a new subsection so it's clear the discussion is not about VOC.
p13l440: Can you explain why NOx has to compete with NH3 to form PM?
p14l486: Please consider the use of a different word than dilution which has a specific meaning in chemistry that might confuse the reader (ie. rephrase "diluted by other emitted species")
For the plots in the supplement pages 3-10 it would be advisable to use raster graphics showing the output on the native grid, rather than interpolated values on the map.

Reply

Citation: https://doi.org/10.5194/egusphere-2023-1168-RC1

Alexander de Meij et al.

Supplement

https://doi.org/10.5194/egusphere-2023-1168-supplement

Alexander de Meij et al.

Viewed

Total article views: 223 (including HTML, PDF, and XML)

HTML	PDF	XML	Total	Supplement	BibTeX	EndNote
158	56	9	223	22	5	5

HTML: 158
PDF: 56
XML: 9
Total: 223
Supplement: 22
BibTeX: 5
EndNote: 5

Views and downloads (calculated since 29 Jun 2023)

Month	HTML	PDF	XML	Total
Jun 2023	33	10	3	46
Jul 2023	77	30	5	112
Aug 2023	24	5	0	29
Sep 2023	24	11	1	36

Cumulative views and downloads (calculated since 29 Jun 2023)

Month	HTML	PDF	XML	Total
Jun 2023	33	10	3	46
Jul 2023	77	30	5	112
Aug 2023	24	5	0	29
Sep 2023	24	11	1	36

Viewed (geographical distribution)

Total article views: 225 (including HTML, PDF, and XML) Thereof 225 with geography defined and 0 with unknown origin.

Country	#	Views	%

Latest update: 30 Sep 2023

Short summary

In our study the robustness of the model responses to emission reductions in EU is assessed when the emission data are changed. Our findings are particularly important to better understand the uncertainties associated to the emission inventories and how these uncertainties impact the level of accuracy of the resulting air quality modelling, which is a key model output for designing air quality plans. Also crucial is the choice of indicator to avoid misleading interpretations of the results.


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