For air quality management, while numerical tools are mainly evaluated to assess their performances on absolute concentrations, this study assesses the impact of their settings on the robustness of model responses to emission reduction strategies for the main criteria pollutants. The effect of the spatial resolution and chemistry schemes is investigated. We show that whereas the spatial resolution is not a crucial setting (except for NO 2 ), the chemistry scheme has more impact, particularly when assessing hourly values of the absolute potential of concentrations. The analysis of model responses under the various configurations triggered an analysis of the impact of using online models, like WRF-chem or WRF-CHIMERE, which accounts for the impact of aerosol concentrations on meteorology. This study informs the air quality modeling community on what extent some model settings can affect the expected model responses to emission changes. We suggest to not activate online effects when analyzing the effect of an emission reduction strategy to avoid any confusion in the interpretation of results even if an online simulation should represent better the reality.

The sensitivity of air quality model responses to modifications in input data (e.g. emissions, meteorology and boundary conditions) or model configurations is recognized as an important issue for air quality modelling applications in support of air quality plans. In the framework of FAIRMODE (Forum of Air Quality Modelling in Europe, https://fairmode.jrc.ec.europa.eu/ ) a dedicated air quality modelling exercise has been designed to address this issue. The main goal was to evaluate the magnitude and variability of air quality model responses when studying emission scenarios/projections by assessing the changes of model output in response to emission changes. This work is based on several air quality models that are used to support model users and developers, and, consequently, policy makers. We present the FAIRMODE exercise and the participating models, and provide an analysis of the variability of O 3 and PM concentrations due to emission reduction scenarios. The key novel feature, in comparison with other exercises, is that emission reduction strategies in the present work are applied and evaluated at urban scale over a large number of cities using new indicators such as the absolute potential, the relative potential and the absolute potency. The results show that there is a larger variability of concentration changes between models, when the emission reduction scenarios are applied, than for their respective baseline absolute concentrations. For ozone, the variability between models of absolute baseline concentrations is below 10%, while the variability of concentration changes (when emissions are similarly perturbed) exceeds, in some instances 100% or higher during episodes. Combined emission reductions are usually more efficient than the sum of single precursor emission reductions both for O 3 and PM. In particular for ozone, model responses, in terms of linearity and additivity, show a clear impact of non-linear chemistry processes. This analysis gives an insight into the impact of model’ sensitivity to emission reductions that may be considered when designing air quality plans and paves the way of more in-depth analysis to disentangle the role of emissions from model formulation for present and future air quality assessments.

A standardized methodology for the validation of short-term air quality forecast applications was developed in the framework of the Forum for Air quality Modeling (FAIRMODE) activities. The proposed approach, focusing on specific features to be checked when evaluating a forecasting application, investigates the model's capability to detect sudden changes in pollutant concentration levels, predict threshold exceedances and reproduce air quality indices. The proposed formulation relies on the definition of specific forecast modelling quality objectives and performance criteria, defining the minimum level of quality to be achieved by a forecasting application when it is used for policy purposes. The persistence model, which uses the most recent observed value as the predicted value, is used as a benchmark for the forecast evaluation. The validation protocol has been applied to several forecasting applications across Europe, using different modelling paradigms and covering a range of geographical contexts and spatial scales. The method is successful, with room for improvement, in highlighting shortcomings and strengths of forecasting applications. This provides a useful basis for using short-term air quality forecasts as a supporting tool for providing correct information to citizens and regulators.

Abstract. The Eurodelta-Trends multi-model experiment, aimed to assess the efficiency of emission mitigation measures in improving air quality in Europe during 1990–2010, was designed to answer a series of questions regarding European pollution trends. i.e. were there significant trends detected by observations? do the models manage to reproduce observed trends? how close is the agreement between the models and how large are the deviations from observations? In this paper, we address these issues with respect to PM pollution. An in-depth trend analysis has been performed for PM10 and PM2.5 for the period of 2000–2010, based on results from six chemical transport models and observational data from the EMEP (Cooperative Programme for Monitoring and Evaluation of the Long-range Transmission of Air Pollutants in Europe) monitoring network. Given harmonization of set up and main input data, the differences in model results should mainly result from differences in the process formulations within the models themselves, and the spread in the models simulated trends could be regarded as an indicator for modelling uncertainty. The model ensemble simulations indicate overall decreasing trends in PM10 and PM2.5, with reduction by between 2 and 6 μg m−3 m−3 (or between 10 and 30 %) from 2000 to 2010. Compared to PM2.5, relative PM10 trends are weaker due to large inter-annual variability of natural coarse PM within the former. The changes in the concentrations of PM individual components are in general consistent with emission reductions. There is a reasonable agreement in PM trends estimated by the individual models, with the inter-model variability below 30–40 % over most of Europe, increasing to 50–60 % in northern and eastern parts of EDT domain. Averaged over measurement sites (26 for PM10 and 13 for PM2.5), the mean ensemble simulated trends are −0.24 and −0.22 μg m−3 year−1 for PM10 and PM2.5, which are somewhat weaker than the observed trends of −0.35 and −0.40 μg m−3 year−1, respectively, partly due to models underestimation of PM concentrations. The correspondence is better in relative PM10 and PM2.5 trends, which are −1.7 and −2.0 % year−1 from the model ensemble and −2.1 and −2.9 % year−1 from the observations, respectively. The observations identify significant trends for PM10 at 56 % of the sites and for PM2.5 at 36 % of the sites, which is somewhat less that the fractions of significant modelled trends. Further, we find somewhat smaller spatial variability of modelled PM trends with respect to the observed ones across Europe and also within individual countries. The strongest decreasing PM trends and the largest number of sites with significant trends is found for the summer season, according to both the model ensemble and observations. The winter PM trends are very weak and mostly insignificant. One important reason for that is the very modest reductions and even increases in the emissions of primary PM from residential heating in winter. It should be kept in mind that all findings regarding modeled versus observed PM trends are limited the regions where the sites are located. The analysis reveals a considerable variability of the role of the individual aerosols in PM10 trends across European countries. The multi-model simulations, supported by available observations, point to decreases in SO4−2 concentrations playing an overall dominant role. Also, we see relatively large contributions of the trends of NH4+ and NO3− to PM10 decreasing trends in Germany, Denmark, Poland and the Po Valley, while the reductions of primary PM emissions appears to be a dominant factor in bringing down PM10 in France, Norway, Portugal, Greece and parts of the UK and Russia. Further discussions are given with respect to emission uncertainties and the effect of inter-annual meteorological variability on the trend analysis.

The sensitivity of air quality model responses to emission reductions when input data (e.g. emissions, meteorology and boundary conditions) or the model set-up configurations are changed is recognised as an important issue for the optimisation of model simulations. In the framework of FAIRMODE (Forum of Air Quality Modelling in Europe) and, in particular, in its Cross Cutting Task 9 (CT9), a dedicated intercomparison exercise has been designed. The goal is to evaluate the robustness of air quality models when studying emission scenarios/projections and to address the issue of the sensitivity of model responses to emission changes, in particular, to identify, study and possibly reduce discrepancies between models. This will provide robust support to model users and developers, and, consequently, policy makers. The aim of this paper is to present the FAIRMODE CT9 platform exercise, its evaluation tool and the community of models involved and provide a preliminary evaluation on the variability of model responses. The key novel feature (in comparison with other exercises) is that emission reduction strategies are here applied and evaluated at urban scale. Appropriate indicators based on previous studies, such as the absolute potential, the relative potential, the absolute potency are defined. The first results based on emission reductions, show a large variability in the indicator values, more so than for the absolute concentrations. While a variability below 10% between models is observed on Ozone concentrations during episodes, the results on indicators show a variability exceeding 100% and even more looking at the highest values. Also, applying emission reductions on several precursors at the same time seems more beneficial to reduce PM concentrations. Model responses on linearity and additivity show a clear impact of non-linear chemistry processes (particularly for Ozone).

The wet deposition of nitrogen and sulfur in Europe for the period 1990–2010 was estimated by six atmospheric chemistry transport models (CHIMERE, CMAQ, EMEP MSC-W, LOTOS-EUROS, MATCH and MINNI) within the framework of the EURODELTA-Trends model intercomparison. The simulated wet deposition and its trends for two 11-year periods (1990–2000 and 2000–2010) were evaluated using data from observations from the EMEP European monitoring network. For annual wet deposition of oxidised nitrogen (WNOx), model bias was within 30 % of the average of the observations for most models. There was a tendency for most models to underestimate annual wet deposition of reduced nitrogen (WNHx), although the model bias was within 40 % of the average of the observations. Model bias for WNHx was inversely correlated with model bias for atmospheric concentrations of NH3+NH4+, suggesting that an underestimation of wet deposition partially contributed to an overestimation of atmospheric concentrations. Model bias was also within about 40 % of the average of the observations for the annual wet deposition of sulfur (WSOx) for most models. Decreasing trends in WNOx were observed at most sites for both 11-year periods, with larger trends, on average, for the second period. The models also estimated predominantly decreasing trends at the monitoring sites and all but one of the models estimated larger trends, on average, for the second period. Decreasing trends were also observed at most sites for WNHx, although larger trends, on average, were observed for the first period. This pattern was not reproduced by the models, which estimated smaller decreasing trends, on average, than those observed or even small increasing trends. The largest observed trends were for WSOx, with decreasing trends at more than 80 % of the sites. On average, the observed trends were larger for the first period. All models were able to reproduce this pattern, although some models underestimated the trends (by up to a factor of 4) and others overestimated them (by up to 40 %), on average. These biases in modelled trends were directly related to the tendency of the models to under- or overestimate annual wet deposition and were smaller for the relative trends (expressed as % yr−1 relative to the deposition at the start of the period). The fact that model biases were fairly constant throughout the time series makes it possible to improve the predictions of wet deposition for future scenarios by adjusting the model estimates using a bias correction calculated from past observations. An analysis of the contributions of various factors to the modelled trends suggests that the predominantly decreasing trends in wet deposition are mostly due to reductions in emissions of the precursors NOx, NH3 and SOx. However, changes in meteorology (e.g. precipitation) and other (non-linear) interactions partially offset the decreasing trends due to emission reductions during the first period but not the second. This suggests that the emission reduction measures had a relatively larger effect on wet deposition during the second period, at least for the sites with observations.

S1 Analysis of the errors in wet deposition due to errors in the primary particulate matter emissionsErrors were found in the emissions of primary particulate matter for Russia and North African countries and shipping for the period 1991-1999.Unfortunately it was not possible to re-run the simulations since these errors were not detected until late in the data analysis.In order to estimate the impact on the wet deposition estimates, the CHIMERE model was used to simulate wet deposition using the incorrect and corrected emissions for 1998, the year for which the emission error is the largest.Figure S1 shows the relative errors in the model estimates of WNOx, WNHx and WSOx as a result of the errors in emissions.Errors in WNOx and WSOx were less than 0.5% in most of the domain with maximum errors of 0.95% and 1.5%, respectively.Errors in WNHx were also mostly below 0.5% but larger errors were estimated for about a quarter of the domain (mostly in the northeast), with a maximum error of 2.4%.These errors are small compared with the overall uncertainty of the model estimates and the uncertainty of the observations.Errors in the trends calculated from the simulations with erroneous emissions are expected to be smaller than the errors in annual deposition rates.From this analysis we conclude that the error in emissions is unlikely to affect the results and conclusions of the study significantly.Figure S1: Maps of relative difference between the model estimates of WNOx, WNHx and WSOx for the simulations using the original (incorrect) emissions and those using the corrected emissions of primary PM in Russia, North Africa and maritime areas for 1998.Circles show the locations of the sites used to evaluate modelled deposition.Note: Positive values indicate that the emission error resulted in larger values of wet deposition and vice versa. S2 Table and Figures cited in the articleTable S1: Main features of the chemistry-transport models involved in the EURODELTA-Trends deposition modelling exercise (Adapted from Colette et al., 2017a).

Reliable estimates of atmospheric deposition of nitrogen and sulphur are essential for assessing the risks of eutrophication and acidification to sensitive ecosystems. Atmospheric chemistry and transport models are frequently used to provide these estimates. Several models could be used to estimate deposition rates and, therefore, it is desirable that the models’ abilities to estimate atmospheric deposition are evaluated and compared. Since these models are used to study future scenarios (e.g. emission abatement), it is also desirable that the model estimates respond realistically to changing conditions (e.g. emissions, boundary conditions, meteorology, etc.), something that can only be tested by evaluating the models’ abilities to estimate past changes in deposition. As part of the EURODELTA/TFMM trend modelling exercise, wet deposition of sulphur and nitrogen in Europe was simulated by seven models for the period 1990-2010. The model estimates of annual deposition rates and their temporal trends were compared and evaluated for two ten year periods (1990-2000 and 2000-2010) using data from EMEP background measurement stations. Model estimates of wet deposition differ greatly for all species and all models tended to underestimate the observed values, apart from one model that tended to overestimate sulphur deposition. Most of the observed and modelled trends of reduced nitrogen are not significant, reflecting the smaller emission changes for ammonia compared with those of nitrogen oxides and sulphur dioxide. Observed and modelled trends of wet deposition of oxidised nitrogen were more significant, with decreasing trends at most of the measurement stations. The most significant observed and modelled trends were found for sulphur wet deposition, the majority of which were decreasing trends, reflecting the reported European sulphur dioxide emission reductions of approximately 70% for the modelling period. Most of the models underestimated the trends in wet deposition of sulphur and oxidised nitrogen over the two ten year periods, although model performance differed greatly. Model performance was better and less variable for the relative trends (trends with respect to the values at the beginning of the period) than for the absolute trends. This suggests that relative trends could be used to produce more reliable deposition estimates for future scenarios.

This study shows the capabilities of a benchmarking system to identify inconsistencies in emission inventories, and to evaluate the reason behind discrepancies as a mean to improve both bottom-up and downscaled emission inventories. Fine scale bottom-up emission inventories for seven urban areas in Norway are compared with three regional emission inventories, EC4MACS, TNO_MACC-II and TNO_MACC-III, downscaled to the same areas. The comparison shows discrepancies in nitrogen oxides (NOx) and particulate matter (PM2.5 and PM10) when evaluating both total and sectorial emissions. The three regional emission inventories underestimate NOx and PM10 traffic emissions by approximately 20–80% and 50–90%, respectively. The main reasons for the underestimation of PM10 emissions from traffic in the regional inventories are related to non-exhaust emissions due to resuspension, which are included in the bottom-up emission inventories but are missing in the official national emissions, and therefore in the downscaled regional inventories. The benchmarking indicates that the most probable reason behind the underestimation of NOx traffic emissions by the regional inventories is the activity data. The fine scale NOx traffic emissions from bottom-up inventories are based on the actual traffic volume at the road link and are much higher than the NOx emissions downscaled from national estimates based on fuel sales and based on population for the urban areas. We have identified important discrepancies in PM2.5 emissions from wood burning for residential heating among all the inventories. These discrepancies are associated with the assumptions made for the allocation of emissions. In the EC4MACs inventory, such assumptions imply high underestimation of PM2.5 emissions from the residential combustion sector in urban areas, which ranges from 40 to 90% compared with the bottom-up inventories. The study shows that in three of the seven Norwegian cities there is need for further improvement of the emission inventories.

The model trend analysis for PM10 and PM2.5, performed within the Eurodelta-Trends experiment, covers 21 year, from 1990 through 2010, with particular focus on the period 2001-2010 for which appropriate amount of PM observations is available. Eight chemical transport models (CTM) participated in the multi-modal trend analysis: EMEP/MSC-W, CHIMERE, LOTOS-EUROS, MINNI, MATCH, WRF-Chem, CMAQ and Polyphemus (out of which six models performed trend runs for the 21-year period). The average modelled trends are somewhat smaller than the observed, though the models identify significant PM trends at more sites in the period 2001-2010. There are considerable difference in the PM trends between the regions/countries and in different seasons. Investigation of the changes in PM chemical composition during the investigated period shows that the models differ in terms of relative contribution of the individual PM components to the PM trends. For the 2001-2010 period, the effcct of inter-annual meteorological variability appears more important relative to emission changes. Finally, we look at PM trends/changes during the 1990-2010 period.

The present report synthesises the main features of the evolution over the 1990-2012 time period of the concentration and deposition of air pollutants relevant in the context of the Convention on Long-range Transboundary Air Pollution: (i) ozone, (ii) sulfur and nitrogen compounds and particulate matter, (iii) heavy metals and persistent organic pollutants. It is based on observations gathered in State Parties to the Convention within the EMEP monitoring network of regional background stations, as well as relevant modelling initiatives. Joint Report of: EMEP Task Force on Measurements and Modelling (TFMM), Chemical Co-ordinating Centre (CCC), Meteorological Synthesizing Centre-East (MSC-E), Meteorological Synthesizing Centre-West (MSC-W).

PM trends in Europe during the last decades have been assessed using a multi-model approach in the framework of Eurodelta-TRENDS exercise initiated by the European Monitoring and Evaluation Program Task Force on Measurements and Modeling (EMEP-TFMM). Seven regional models (EMEP/MSC-W, CHIMERE, CMAQ, LOTOS-EUROS, MINNI, Polair3D and WRF-Chem, participated in the EURODELTA-Trends exercise, which builds upon previous iterations of the CITYDELTA and EURODELTA projects (Thunis et al., 2007; Cuvelier et al., 2007; Bessagnet et al., 2014). Three of the models performed a 21-year hindcast over Europe for the 1990-2010 period, while all the models made calculations for the years 1990, 2000 and 2010. In addition, a series of sensitivity tests have been performed with the purpose of studying the role of meteorological variability, emission changes and boundary conditions. For the period of 2001-2010, for which enough of PM monitoring data is available, the trends in PM10 and PM2.5 have been studied based on both modelling results and measurement data. The Mann-Kendall test are applied to calculated and measured annual mean concentrations to detect significant (90% probability) trends, whereas the Sen’s slopes are calculated to estimate the absolute and relative declines in PM...

The present report synthesises the main features of the evolution over the 1990-2012 time period of the concentration and deposition of air pollutants relevant in the context of the Convention on Long-range Transboundary Air Pollution: (i) ozone, (ii) sulfur and nitrogen compounds and particulate matter, (iii) heavy metals and persistent organic pollutants. It is based on observations gathered in State Parties to the Convention within the EMEP monitoring network of regional background stations, as well as relevant modelling initiatives. Joint Report of: EMEP Task Force on Measurements and Modelling (TFMM), Chemical Co-ordinating Centre (CCC), Meteorological Synthesizing Centre-East (MSC-E), Meteorological Synthesizing Centre-West (MSC-W).

The present report synthesises the main features of the evolution over the 1990-2012 time period of the concentration and deposition of air pollutants relevant in the context of the Convention on Long-range Transboundary Air Pollution: (i) ozone, (ii) sulfur and nitrogen compounds and particulate matter, (iii) heavy metals and persistent organic pollutants. It is based on observations gathered in State Parties to the Convention within the EMEP monitoring network of regional background stations, as well as relevant modelling initiatives. Joint Report of: EMEP Task Force on Measurements and Modelling (TFMM), Chemical Co-ordinating Centre (CCC), Meteorological Synthesizing Centre-East (MSC-E), Meteorological Synthesizing Centre-West (MSC-W).

Observations of atmospheric reactive nitrogen (Nr) deposition are severely restricted in spatial extent and type. The chain of processes leading to atmospheric deposition emissions, atmospheric dispersion, chemical transformation and eventual loss from the atmosphere is extremely complex and therefore currently, observations can only address part of this chain.

Ozone exposure is associated with negative health impacts, including premature mortality. Observations and modeling studies demonstrate that emissions from one continent influence ozone air quality over other continents. We estimate the premature mortalities avoided from surface ozone decreases obtained via combined 20% reductions of anthropogenic nitrogen oxide, nonmethane volatile organic compound, and carbon monoxide emissions in North America (NA), EastAsia (EA), South Asia (SA), and Europe (EU). We use estimates of ozone responses to these emission changes from several atmospheric chemical transportmodels combined with a health impactfunction. Foreign emission reductions contribute approximately 30%, 30%, 20%, and >50% of the mortalities avoided by reducing precursor emissions in all regions together in NA, EA, SA and EU, respectively. Reducing emissions in NA and EU avoids more mortalities outside the source region than within, owing in part to larger populations in foreign regions. Lowering the global methane abundance by 20% reduces mortality mostin SA,followed by EU, EA, and NA. For some source-receptor pairs, there is greater uncertainty in our estimated avoided mortalities associated with the modeled ozone responses to emission changes than with the health impact function parameters.