Assessment of the Effects of the Spanish National Air Pollution Control Programme on Air Quality

Abstract

During the last few decades, European legislation has driven progress in reducing air pollution in Europe through emission mitigation measures. In this paper, we use a chemistry transport model to assess the impact on ambient air quality of the measures considered for 2030 in the for the scenarios with existing (WEM2030) and additional measures (WAM2030). The study estimates a general improvement of air quality for the WAM2030 scenario, with no non-compliant air quality zones for NO2, SO2, and PM indicators. Despite an improvement for O3, the model still estimates non-compliant areas. For this pollutant, the WAM2030 scenario leads to different impacts depending on the indicator considered. Although the model estimates a reduction in maximum hourly O3 concentrations, small increases in O3 concentrations in winter and nighttime in the summer lead to increases in the annual mean in some areas and increases in other indicators (SOMO35 for health impacts and AOT40 for impacts on vegetation) in some urban areas. The results suggest that the lower NOx emissions in the WEM and WAM scenarios lead to less removal of O3 by NO titration, especially background ozone in winter and both background and locally produced ozone in summer, in areas with high NOx emissions.

Full text

atmosphere
Article
Assessment of the Effects of the Spanish National Air Pollution
Control Programme on Air Quality
Marta G. Vivanco 1,*, Juan Luis Garrido 1, Fernando Martín1, Mark R. Theobald 1, Victoria Gil 1,
José-Luis Santiago 1, Yolanda Lechón2, Ana R. Gamarra 2,3 , Eugenio Sánchez 4, Angelines Alberto 5and
Almudena Bailador 4


Citation: Vivanco, M.G.; Garrido,
J.L.; Martín, F.; Theobald, M.R.; Gil, V.;
Santiago, J.-L.; Lechón, Y.; Gamarra,
A.R.; Sánchez, E.; Alberto, A.; et al.
Assessment of the Effects of the
Spanish National Air Pollution
Control Programme on Air Quality.
Atmosphere 2021,12, 158. https://
doi.org/10.3390/atmos12020158
Academic Editor: Ole Hertel
Received: 1 December 2020
Accepted: 20 January 2021
Published: 26 January 2021
Publisher’s Note: MDPI stays neutral
with regard to jurisdictional claims in
published maps and institutional affil-
iations.
Copyright: © 2021 by the authors.
Licensee MDPI, Basel, Switzerland.
This article is an open access article
distributed under the terms and
conditions of the Creative Commons
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/
4.0/).
1Atmospheric Pollution Division, Environment Department, CIEMAT, Av. Complutense, 40,
28040 Madrid, Spain; juanluis.garrido@ciemat.es (J.L.G.); fernando.martin@ciemat.es (F.M.);
mark.theobald@ciemat.es (M.R.T.); VictoriaE.Gil@ciemat.es (V.G.); jl.santiago@ciemat.es (J.-L.S.)
2Energy Systems Analysis Unit, Energy Department, CIEMAT, Av. Complutense, 40, 28040 Madrid, Spain;
yolanda.lechon@ciemat.es (Y.L.); anarosa.gamarra@ciemat.es (A.R.G.)
3
Chemical and Environmental Engineering Department, Universidad Politécnica de Madrid, C/JoséGutiérrez
Abascal, 2, 28006 Madrid, Spain
4Application and Computer Systems Development Unit, Department of Technology, CIEMAT,
Av. Complutense, 40, 28040 Madrid, Spain; eugenio.sanchez@ciemat.es (E.S.);
almudena.bailador@ciemat.es (A.B.)
5Computer Architecture, Department of Technology, CIEMAT, Av. Complutense, 40, 28040 Madrid, Spain;
angelines.alberto@ciemat.es
*Correspondence: m.garcia@ciemat.es
Abstract:
During the last few decades, European legislation has driven progress in reducing air
pollution in Europe through emission mitigation measures. In this paper, we use a chemistry
transport model to assess the impact on ambient air quality of the measures considered for 2030 in the
for the scenarios with existing (WEM2030) and additional measures (WAM2030). The study estimates
a general improvement of air quality for the WAM2030 scenario, with no non-compliant air quality
zones for NO
2
, SO
2
, and PM indicators. Despite an improvement for O
3
, the model still estimates
non-compliant areas. For this pollutant, the WAM2030 scenario leads to different impacts depending
on the indicator considered. Although the model estimates a reduction in maximum hourly O
3
concentrations, small increases in O
3
concentrations in winter and nighttime in the summer lead to
increases in the annual mean in some areas and increases in other indicators (SOMO35 for health
impacts and AOT40 for impacts on vegetation) in some urban areas. The results suggest that the
lower NO
x
emissions in the WEM and WAM scenarios lead to less removal of O
3
by NO titration,
especially background ozone in winter and both background and locally produced ozone in summer,
in areas with high NOxemissions.
Keywords:
air pollution; air quality modeling; ozone; NEC Directive; national programme; NO-
titration; emission control
1. Introduction
During the last two decades, substantial reductions in the emissions of primary
pollutants that contribute to ambient air concentrations of particulate matter (PM), O
3
and NO
2
, have been made in Europe. Between 2000 and 2018, European emissions of
SO
2
were reduced the most (79%) and those of NH
3
the least (10%). In fact, the latter
actually increased between 2015 and 2018, mainly driven by emissions from agriculture [
1
].
European and national policies focused on air pollution control have, in general, succeeded
in improving air quality in Europe since 2000.
According to a recent report by the European Environment Agency [
1
], a trend assess-
ment study of air pollutant concentrations for the period between 2000 and 2017 shows
that the average NO
2
annual mean concentration has decreased in Europe, as well as the
Atmosphere 2021,12, 158. https://doi.org/10.3390/atmos12020158 https://www.mdpi.com/journal/atmosphere

Atmosphere 2021,12, 158 2 of 21
annual mean of SO
2
, for which almost all of the stations with a significant trend recorded
a decrease in concentrations. PM10 and PM2.5 annual mean concentrations also exhibit
a decreasing trend on average across the stations with data available (23), with a faster
decrease in PM10 between 2000 and 2008 than between 2008 and 2017.
However, for O
3
, the situation is different. In spite of a reduction in the emissions of
O
3
precursors (NO
x
and non-methane volatile organic compounds (NMVOCs)), varied
trends are found for O
3
concentrations, depending on the metric (indicator) used. In the
same report by the European Environment Agency [
1
], no clear trend for SOMO35 (sum of
ozone means over 35 ppb, defined as the yearly sum of the daily maximum of the 8-hour
running mean over 35 ppb) is reported, except for traffic stations, where concentrations
have increased on average. By contrast, peak O
3
, defined as the fourth highest maximum
daily 8-hour mean (98.9th percentile), exhibits a clear decreasing trend from 2000 to 2008
for all station types, except traffic sites. However, from 2009, there is no clear trend, and a
flattening of the curve is seen for all station types. According to the trend analysis for the
period 2009–2018, SOMO35 increased on average for all station types, except for stations
classed as industrial. The same report refers to the results from a generalized additive
model [
2
], which estimates no clear trend for SOMO35 in rural background stations and a
general increasing trend from 2010 to 2016 over the Iberian Peninsula.
In spite of this general improvement in air quality, the European Environment Agency
still considers air pollution to be the largest environmental threat to health. In 2018, the
proportion of the urban population in the European Union (EU) exposed to atmospheric
concentrations above European limit values for PM10, PM2.5, NO
2
, and O
3
was estimated
to be 15%, 4%, 4%, and 3 %, respectively [
1
]. The percentage of population above the WHO
guideline concentrations for health protection is even higher (except for NO
2
, which has the
same value as the EU limit value), i.e., 48%, for PM10, 74% for PM2.5, and 99% for O
3
[
1
]. It
is estimated that these high levels of exposure to PM2.5, NO
2
, and O
3
were responsible for
379,000, 54,000, and 15,700 premature deaths, respectively, in 2018 in the EU [1].
European legislation is still pushing further emission reductions in the near future.
Under the National Emissions Ceilings Directive (NEC Directive, 2016/2284/EU) of the Eu-
ropean Parliament, European countries are committed to reduce the annual emissions of the
following five pollutants: SO
x
, NO
x
, non-methane volatile organic compounds (NMVOC),
NH
3
, and PM2.5, with respect to emissions in 2005. In the case of non-compliance, member
states are required to adopt and put into action air quality plans to remediate the problem.
In addition to the NEC Directive, air quality in Europe is regulated under the Directive on
Ambient Air Quality and Cleaner Air for Europe (Air Quality Directive, 2008/50/EC), in
which limit values, target values, and exposure concentration obligations are defined for
particulate matter (PM), SO2, NOx, O3, CO, lead, and benzene.
Emissions in Spain have decreased in recent decades for some pollutants, especially
SO
2
and NO
x
(Figure 1, solid line). For SO
2
, there was a large decrease around 2007, mainly
produced by the implementation of a national plan for the reduction of emissions from
large combustion plants, including the introduction of new desulfurization technologies.
Reductions in NO
x
emissions have also been substantial, with a decrease driven by lower
emissions from the transport sector, mainly due to cleaner emission standards for new
vehicles. For NMVOCs, an initial decrease was observed, but there was a slight emission
increase after 2013, mainly due to emissions in the solvent-use sector. Emissions of NH
3
have also increased since 2013. Despite the progress made in reducing emissions in Spain,
the limit values set out in the Air Quality Directive for tropospheric ozone, NO
2
, and
particulate matter are still frequently exceeded [
3
], although PM10 and NO
2
concentrations
have decreased slightly in the last 10 years ([
3
], reports of the Spanish Ministry for the
Ecological Transition and Demographic Challenge (METDC)). Further reductions are still
required in order to meet the objectives set for Spain in the NEC Directive (reductions of
88%, 62%, 39%, 16% and 50% in SO
x
, NO
x
, NMVOC, NH
3
, and PM2.5, respectively, dash-
dotted lines in Figure 1). These obligations have been transposed to national legislation in
Spain (Real Decreto 818/2018).

Atmosphere 2021,12, 158 3 of 21
Atmosphere 2021, 12, x FOR PEER REVIEW 3 of 22
ther reductions are still required in order to meet the objectives set for Spain in the NEC
Directive (reductions of 88%, 62%, 39%, 16% and 50% in SOx, NOx, NMVOC, NH3, and
PM2.5, respectively, dash-dotted lines in Figure 1). These obligations have been trans-
posed to national legislation in Spain (Real Decreto 818/2018).
In order to meet these objectives, Spain has developed a national programme, the
first National Air Pollution Control Programme (NAPCP), approved in September 2019
[4], which includes a set of measures that would lead to the emissions shown in Figure 1
(dotted lines) for the scenario with additional measures (WAM), projected for 2020, 2025,
and 2030. The WAM2030 scenario achieves the target emissions reductions set in the
NEC Directive for Spain in 2030 for all pollutants except NMVOCs. Meeting the
NMVOC emission target will remain a challenge, since the NAPCP only achieves a 30%
reduction as compared with the 39% required by the NEC Directive.
Figure 1. Evolution of NOx, SOx (as SO2), NH3, NMVOC, and PM2.5 emissions in Spain taken from
the National Emission Inventory produced by the Ministry for the Ecological Transition and De-
mographic Challenge for the period 2000–2018 (solid lines up to 2018), Commitments under the
NEC Directive (dash-dotted lines, values for 2020 and 2030, with 2005 as the reference year) and
projected emissions for the WEM and WAM scenarios for 2020, 2025, and 2030 (solid and dashed
lines, respectively).
In this context, this paper aims to quantify the impacts of the proposed emission
reductions in the NAPCP (NMVOC, SO2, NOX, PM2.5, and NH3) on the relevant pollu-
tant indicators covered by the Air Quality Directive. These indicators are maximum
hourly and daily maximum 8-hour O3 concentrations, annual mean NO2, annual and
daily mean SO2, annual and daily mean PM10, and annual mean PM2.5. We discuss the
results, with a special focus on O3, due to the fact that the effects of the emission reduc-
tions have been found to be dependent on the indicator or metric calculated.
2. Methodology
2.1. Description of the Scenario with Additional Measures Projected for 2030 in the Spanish
National Air Pollution Control Programme on Air Quality (WAM2030)
The NAPCP considers 50 measures (in parentheses) divided into the following 8
sectorial packages for the WAM2030 scenario:
(1) Energy mix (8);
(2) Transport: road transport, rail, aviation and shipping (6);
Figure 1.
Evolution of NO
x
, SO
x
(as SO
2
), NH
3
, NMVOC, and PM2.5 emissions in Spain taken
from the National Emission Inventory produced by the Ministry for the Ecological Transition and
Demographic Challenge for the period 2000–2018 (solid lines up to 2018), Commitments under the
NEC Directive (dash-dotted lines, values for 2020 and 2030, with 2005 as the reference year) and
projected emissions for the WEM and WAM scenarios for 2020, 2025, and 2030 (solid and dashed
lines, respectively).
In order to meet these objectives, Spain has developed a national programme, the first
National Air Pollution Control Programme (NAPCP), approved in September 2019 [
4
],
which includes a set of measures that would lead to the emissions shown in Figure 1(dotted
lines) for the scenario with additional measures (WAM), projected for 2020, 2025, and 2030.
The WAM2030 scenario achieves the target emissions reductions set in the NEC Directive
for Spain in 2030 for all pollutants except NMVOCs. Meeting the NMVOC emission target
will remain a challenge, since the NAPCP only achieves a 30% reduction as compared with
the 39% required by the NEC Directive.
In this context, this paper aims to quantify the impacts of the proposed emission
reductions in the NAPCP (NMVOC, SO
2
, NO
X
, PM2.5, and NH
3
) on the relevant pollutant
indicators covered by the Air Quality Directive. These indicators are maximum hourly
and daily maximum 8-hour O
3
concentrations, annual mean NO
2
, annual and daily mean
SO
2
, annual and daily mean PM10, and annual mean PM2.5. We discuss the results, with a
special focus on O
3
, due to the fact that the effects of the emission reductions have been
found to be dependent on the indicator or metric calculated.
2. Methodology
2.1. Description of the Scenario with Additional Measures Projected for 2030 in the Spanish
National Air Pollution Control Programme on Air Quality (WAM2030)
The NAPCP considers 50 measures (in parentheses) divided into the following
8 sectorial packages for the WAM2030 scenario:
(1)
Energy mix (8);
(2)
Transport: road transport, rail, aviation and shipping (6);
(3)
Improved energy efficiency in the industrial and manufacturing sectors (3);
(4)
Improved energy efficiency (7);
(5)
Waste management (8);
(6)
Use of fertilizer plans (9);
(7)
Reduction in emissions from burning of pruning (2);

Atmosphere 2021,12, 158 4 of 21
(8)
Manure and housing management for cattle, pigs and poultry (7).
The NAPCP also includes 5 other packages with 7 additional measures that are
considered to be target measures, with a special focus on NMVOC emissions, which are not
expected to meet the NEC objectives in the WAM30 scenario. In summary, the programme
contains a total of 13 packages with 57 measures. All the packages and measures are
summarized in Appendix A(sectorial measures) and Appendix B(target measures).
The main effort of the emission reductions is focused on the transport sector (T.1,
Appendix Aand Figure 2), mainly through the promotion of electric vehicles, which affects
NO
x
emissions and, to a lesser extent, those of NMVOC. Other measures such as the use of
biofuels for transport and changes in transportation modes are also included in the NAPCP.
Measures related to the energy mix (E.1, Appendix Aand Figure 2) and energy efficiency
(I.1 and EE.1, Appendix Aand Figure 2), with the promotion of renewable technologies,
will result in a decrease in SO
2
emissions and, to a lesser extent, NO
x
emissions. For
NMVOC, the reductions are generally quite small, mainly focused on transport and energy
efficiency in the residential sector (EE.1). For NH
3
, the measures in the NAPCP affect the
agriculture sector (mainly A.1 and A.3, Appendix A). The reduction expected for each
package and pollutant is shown in Figure 2(categories are explained in Appendix A).
Atmosphere 2021, 12, x FOR PEER REVIEW 5 of 23
Change of Emissions (kt) in the WAM2030 scenario with Respect to 2016
Figure 2. Contributions of the different measures to the reduction in emissions for each pollutant.
2.2. Modeling Methodology
The CHIMERE chemistry transport model [5] was used to evaluate the impacts of
the emission reductions described in Section 2.1 on the concentrations of the pollutants
regulated under EU legislation that are affected by these reductions (NO
2
, O
3
, SO
2
,
PM2.5, and PM10). The CHIMERE model has been extensively used and evaluated in
Europe [6] and, in particular, in Spain [7–10]. Model performance for estimating atmos-
pheric concentrations has been shown to be comparable to that of other air quality mod-
els applied in Europe [11].
The model was applied to a domain covering the Iberian Peninsula with a spatial
resolution of 0.1° × 0.1°, nested within a European domain (0.15° × 0.15°). The assessment
of the impacts on air quality was carried out by using 2016 as the reference year.
Three simulations were performed, each using a different emission scenario, in
2016, with emissions taken from the national emission inventory (NEI) for that year and
the WEM2030 and WAM2030 scenarios. The reductions for WEM2030 and WAM2030
were provided by the METDC (in collaboration with TRAGSATEC). The emissions for
the other countries within the domains were obtained from the EMEP website [12] for
2016. The NEI uses a 0.1° × 0.1°grid resolution (the same as EMEP grid).
The meteorological fields were adapted from simulations at the European Centre
for Medium-Range Weather Forecasts, ECMWF, (www.ecmwf.int) known as Integrated
Forecasting System (IFS), for 2016, and obtained from the MARS archive at ECMWF
through the access provided by AEMET for research projects. All the simulations used
the meteorology for 2016,in order to only evaluate the impacts due to changes in emis-
sions. Boundary conditions for the European domain were taken from LMDZ-INCA [13]
and GOCART [14] global model climatology. All the simulations used the same bound-
ary conditions.
In this modeling approach, the atmospheric concentrations estimated by the CHI-
MERE model were combined with observations in order to minimize biases and uncer-
Figure 2. Contributions of the different measures to the reduction in emissions for each pollutant.
The implementation of these measures, according to METDC estimates, represents
emissions reductions of 0.4%, 20%, 33%, 39%, and 55% in NMVOCs, NH
3
, NO
x
, PM2.5,
and SO
2
, respectively, with respect to emissions in 2016 (used as the reference year). As
can be seen, major reductions are proposed for SO
2
, PM2.5, and NO
x
, whereas the smallest
reductions are planned for NMVOC. The specific emission reductions for each sector are
included in Appendix C.
2.2. Modeling Methodology
The CHIMERE chemistry transport model [
5
] was used to evaluate the impacts of
the emission reductions described in Section 2.1 on the concentrations of the pollutants

Atmosphere 2021,12, 158 5 of 21
regulated under EU legislation that are affected by these reductions (NO
2
, O
3
, SO
2
, PM2.5,
and PM10). The CHIMERE model has been extensively used and evaluated in Europe [6]
and, in particular, in Spain [
7
–
10
]. Model performance for estimating atmospheric concen-
trations has been shown to be comparable to that of other air quality models applied in
Europe [11].
The model was applied to a domain covering the Iberian Peninsula with a spatial
resolution of 0.1
◦×
0.1
◦
, nested within a European domain (0.15
◦×
0.15
◦
). The assessment
of the impacts on air quality was carried out by using 2016 as the reference year.
Three simulations were performed, each using a different emission scenario, in 2016,
with emissions taken from the national emission inventory (NEI) for that year and the
WEM2030 and WAM2030 scenarios. The reductions for WEM2030 and WAM2030 were
provided by the METDC (in collaboration with TRAGSATEC). The emissions for the other
countries within the domains were obtained from the EMEP website [
12
] for 2016. The NEI
uses a 0.1◦×0.1◦grid resolution (the same as EMEP grid).
The meteorological fields were adapted from simulations at the European Centre for
Medium-Range Weather Forecasts, ECMWF, (www.ecmwf.int) known as Integrated Forecasting
System (IFS), for 2016, and obtained from the MARS archive at ECMWF through the access
provided by AEMET for research projects. All the simulations used the meteorology for 2016,in
order to only evaluate the impacts due to changes in emissions. Boundary conditions for
the European domain were taken from LMDZ-INCA [
13
] and GOCART [
14
] global model
climatology. All the simulations used the same boundary conditions.
In this modeling approach, the atmospheric concentrations estimated by the CHIMERE
model were combined with observations in order to minimize biases and uncertainties
in the concentrations estimates. First, the differences between model and observations
were calculated, separately for urban and rural stations, to take into account the differ-
ent spatial distribution patterns of air pollution concentrations. These differences were
interpolated (using Kriging and a spherical variogram model) and then the resulting map
of residuals was added to the model estimates. The population density was used as a
surrogate indicator to merge urban and rural air pollution maps. More detailed information
can be found in [
15
,
16
]. This methodology takes into account observations and potential
underestimations of the peak values due to model uncertainties and resolution (see [
7
,
9
]
for O
3
and NO
2
underprediction) and has been shown to improve concentration estimates,
as shown in [15,16].
For the WEM2030 and WAM2030 simulations, we also applied a correction to the
modeled concentrations, assuming that relative errors estimated for the 2016 simulation
(the residuals, as explained above) could be assigned to future scenarios. The corrected
concentration was calculated as:
Ca,r =Ca1+R2016
C2016 (1)
where C
a
is the uncorrected future concentration, C
2016
is the concentration for 2016, and
R
2016
is the model error for 2016 (modeled minus observed concentration) interpolated
over the domain.
The NEC Directive requests that all European countries provide information on
the number of air quality zones (zones defined by each member state for air quality
management) that will not be compliant with the air quality limit values set in the European
legislation. As predicting, peak value is essential to assess compliance, we considered that
this correction was necessary.
3. Results: Impacts on Air Quality
In this section, we show the impacts of the different emission scenarios on the relevant
pollutant indicators covered by the Air Quality Directive. The following indicators are
considered: (1) for ozone, maximum hourly concentration, daily maximum 8-hour concen-
tration and AOT40 (accumulated O
3
over the threshold value of 40 ppb, calculated for only

Atmosphere 2021,12, 158 6 of 21
one growing season); (2) for NO
2
, annual mean and maximum hourly concentration; (3) for
SO
2
, annual and daily mean concentration; (4) for PM10, annual and daily mean concentra-
tion; (5) for PM2.5, annual mean concentration. In addition, for ozone, we considered other
indicators that are often used in some applications, such as annual mean and SOMO35.
Figure 3shows the total number of air quality zones for each pollutant and the number of
non-compliant air quality zones for the three scenarios (2016, WEM2030, and WAM2030)
for the metrics with the main problems of non-compliance, namely O
3
hourly values, O
3
daily maximum 8-hour mean, AOT40 (calculated for only one year), NO
2
annual mean,
SO
2
annual mean, and PM10 daily mean. For the WEM2030 scenario, there is a slight
improvement with respect to 2016, with a reduction in non-compliant air quality zones
for PM10 daily mean and NO
2
annual mean, but very little impact on ozone (only a small
reduction in the number of non-compliant zones for the hourly values). For the WAM2030
scenario, there is larger improvement, although non-compliance is still expected for the
ozone metrics. For the rest of the metrics, that is, NO
2
maximum hourly concentration,
SO
2
daily mean concentration, PM10 annual mean concentrations, and PM2.5 annual mean
concentration, the modeling approach does not estimate non-compliant air quality zones
in 2016. In this section, these impacts are described in more detail.
Figure 3.
Number of non-compliant air quality zones for the base scenario (2016) and the WEM and
WAM scenarios for 2030.
3.1. NO2
The implementation of the measures in the NAPCP is expected to decrease NO
2
concentrations. For the annual mean limit value (40
µ
g m
−3
) (Figure 4, bottom) some
exceedances were estimated by the model in the base case 2016 for the city of Madrid,
Cataluña (Barcelona area), and Andalucía. These non-compliant air quality zones disappear
for the WAM2030 emission scenario under the same conditions as 2016 (meteorology and
boundary conditions). For the NO
2
maximum hourly indicator, the limit value of 200
µ
g
m
−3
cannot be exceeded more than 18 times to comply with the Air Quality Directive.
According to the model approach, there are no non-compliant air quality zones in 2016,
WEM2030, or WAM2030. The 19th highest maximum hourly value is presented in
Figure 4
,
to highlight the non-compliant areas (>200
µ
g m
−3
). Despite this improvement in air quality
for NO
2
, very local problems could occur close to traffic-dense areas in the biggest cities,

Atmosphere 2021,12, 158 7 of 21
although these would be expected to decrease with NO
x
emission reductions. It’s important
to note that specific local measures that could be implemented by local authorities have
not been considered in these estimates.
Atmosphere 2020, 11, x FOR PEER REVIEW 7 of 21
The implementation of the measures in the NAPCP is expected to decrease NO
2
concentrations.
197
For the annual mean limit value (Figure 5, bottom) some exceedances were estimated by the model
198
in the base case 2016 for the city of Madrid, Cataluña (Barcelona area) and Andalucía. These
199
non-compliant areas disappear for the WAM2030 emission scenario under the same conditions as
200
2016 (meteorology and boundary conditions). For the NO
2
maximum hourly limit value, no
201
non-compliant (more than 18 exceedances of 200 µg/m3) zones are estimated for 2016 WEM2030 or
202
WAM2030. However very local problems could occur close to traffic-dense areas in the biggest cities,
203
although this would be expected to decrease. It’s important to note that local measures that could be
204
implemented by local authorities have not been considered in these estimates.
205
206
Figure 5. Concentration maps showing the 19th highest hourly values of NO2 (top row) and the
207
annual mean (bottom row). Maps on the left show the concentration in 2016 and relative differences
208
(%) with respect to 2016 are shown for the WEM2030 scenario (middle) and WAM2030 (right).
209
4.2. O
3
210
Figure 6 includes maps of O
3
concentration for different metrics: maximum hourly value, daily
211
maximum 8-hourly value (26
th
), AOT40, SOMO35 and annual mean. The 26th value of the daily
212
maximum 8-hourly value of ozone is presented to indicate non-compliant areas, since, to comply
213
with EU Directive on ambient air quality and cleaner air for Europe (2008/50/EC), the limit value of
214
125 µg m
-3
cannot be exceeded more than 25 times. Figure 7 shows the non-compliant areas for O
3
215
(hourly and daily maximum 8-hourly values) (red areas) for 2016 and the WEM2030 and WAM2030
216
emission scenarios. These maps show that the number of non-compliant areas decreases, although
217
they do not disappear completely in spite of the significant decrease in NOx emissions.
218
For hourly values, the combination of model and observations estimates concentrations above
219
180 µg m
-3
in 2016 for some parts of Galicia, northern Spain (Principado de Asturias, Cantabria, País
220
Vasco), Cataluña and Madrid, and a very small area in Extremadura (eastern Spain, close to
221
Figure 4.
Concentration maps (
µ
g m
−3
) showing the 19th highest hourly values of NO
2
(
top
row,
left) and the annual mean (
bottom
row, left) in 2016. Maps on the middle and right show the relative
differences (%) with respect to 2016 for the WEM2030 (middle) and WAM2030 (right) scenarios.
3.2. O3
Figure 5includes maps of O
3
concentration for the following metrics: maximum
hourly value, daily maximum 8-hour value (26th), AOT40, SOMO35, and annual mean.
The 26th value of the daily maximum 8-hour value of ozone is presented to indicate non-
compliant air quality areas, since, to comply with EU Directive on ambient air quality and
cleaner air for Europe (2008/50/EC), the limit value of 120
µ
g m
−3
cannot be exceeded more
than 25 times. Figure 6shows, in red, the non-compliant areas for O
3
(hourly and daily
maximum 8-hour values) for 2016 and the WEM2030 and WAM2030 emission scenarios.
These maps show that the number of non-compliant areas decreases, although they do not
disappear completely in spite of the significant decrease in NOxemissions.
For hourly values, concentrations above 180
µ
g m
−3
in 2016 are found for some parts
of Galicia, northern Spain (Principado de Asturias, Cantabria, País Vasco), Cataluña, and
Madrid, and a very small area in Extremadura (western Spain, close to Portugal) (Figures 5
and 6, red areas). These zones still persist for the WEM2030 scenario, but they are reduced,
with some of them disappearing in WAM2030.

Atmosphere 2021,12, 158 8 of 21
Atmosphere 2021, 12, x FOR PEER REVIEW 8 of 22
For hourly values, concentrations above 180 µg m−3 in 2016 are found for some parts
of Galicia, northern Spain (Principado de Asturias, Cantabria, País Vasco), Cataluña, and
Madrid, and a very small area in Extremadura (western Spain, close to Portugal) (Figures 5
and 6, red areas). These zones still persist for the WEM2030 scenario, but they are re-
duced, with some of them disappearing in WAM2030.
For the maximum daily 8-hour value (Figure 6) several non-compliant air quality
zones were found in 2016 in the center (Madrid-Guadalajara), northeast (Cataluña, part
of Huesca), the Mediterranean coast (Comunidad Valenciana), some zones in the south-
ern Spain (Andalucía: Sevilla, Málaga and Jaén), and one air quality zone in the west
(Extremadura, Cáceres). Non-compliant areas are also estimated for the Mediterranean
coast, southern Atlantic coast, Balearic Islands and some parts of the Cantabrian coast,
although these correspond to model grid cells with a large fraction of sea, for which
deposition rates are lower, leading to increased air concentrations in the model. The
coastal areas are also the most affected by international shipping emissions, which are
not considered in the national programme. The number of non-compliant air quality
zones is reduced in the WAM2030 scenario (as was the case for the hourly values) but
not for WEM2030.
Figure 5. Concentration maps (µg m−3) for different ozone metrics, for 2016, estimated by the model
approach.
The maps of SOMO35 and AOT40 (that are based on cumulative concentrations) in
Figure 6 show that the areas with the highest values are mainly located in southern, cen-
tral, and eastern Spain.
Figure 5.
Concentration maps (
µ
g m
−3
) for different ozone metrics, for 2016, estimated by the model
approach.
Atmosphere 2021, 12, x FOR PEER REVIEW 9 of 22
Figure 6. Maps (µg m−3) showing the non-compliant grid cells estimated by the modeling approach
for O3 (daily maximum 8-hour value (top) and hourly values (bottom)) and for 2016 (left);
WEM2030 (middle); WAM2030 (right).
The analysis of the relative differences between the 2016 and WEM2030 (Figure 7)
and WAM2030 (Figure 8) scenarios is interesting. For hourly and maximum daily 8-hour
O3, the maps in the figure indicate an improvement of air concentration levels (blue col-
ors) over some regions and no significant change (white) for others. However, for SO-
MO35, AOT40, and annual mean, the metrics increase in some areas (red colors), mainly
in the areas with the largest emissions in 2016. The differences, both increases and de-
creases, are larger for the WAM2030 scenario than for WEM2030 scenario. The increases
in AOT40 for WAM2030 occur mainly around Barcelona and the industrial areas of As-
turias and, although there are also small increases, over Madrid. The increases in SO-
MO35 for this scenario are also located in these areas, as well as some other industrial
areas in Galicia (Northwestern Spain) and in the North (around Bilbao). Increases in the
annual mean concentrations are located in the previously mentioned areas of Asturias,
Madrid, and Barcelona, as well as other small areas in northwestern Spain (Galicia),
northern Spain, southern Spain (Andalucía), and the Mediterranean coast.
Figure 6.
Maps (
µ
g m
−3
) showing the non-compliant grid cells estimated by the modeling approach
for O
3
(daily maximum 8-hour value (
top
) and hourly values (
bottom
)) and for 2016 (left); WEM2030
(middle); WAM2030 (right).
For the maximum daily 8-hour value (Figure 6) several non-compliant air quality
zones were found in 2016 in the center (Madrid-Guadalajara), northeast (Cataluña, part of

Atmosphere 2021,12, 158 9 of 21
Huesca), the Mediterranean coast (Comunidad Valenciana), some zones in the southern
Spain (Andalucía: Sevilla, Málaga and Jaén), and one air quality zone in the west (Ex-
tremadura, Cáceres). Non-compliant areas are also estimated for the Mediterranean coast,
southern Atlantic coast, Balearic Islands and some parts of the Cantabrian coast, although
these correspond to model grid cells with a large fraction of sea, for which deposition rates
are lower, leading to increased air concentrations in the model. The coastal areas are also
the most affected by international shipping emissions, which are not considered in the
national programme. The number of non-compliant air quality zones is reduced in the
WAM2030 scenario (as was the case for the hourly values) but not for WEM2030.
The maps of SOMO35 and AOT40 (that are based on cumulative concentrations) in
Figure 6show that the areas with the highest values are mainly located in southern, central,
and eastern Spain.
The analysis of the relative differences between the 2016 and WEM2030 (Figure 7)
and WAM2030 (Figure 8) scenarios is interesting. For hourly and maximum daily 8-h O
3
,
the maps in the figure indicate an improvement of air concentration levels (blue colors)
over some regions and no significant change (white) for others. However, for SOMO35,
AOT40, and annual mean, the metrics increase in some areas (red colors), mainly in the
areas with the largest emissions in 2016. The differences, both increases and decreases, are
larger for the WAM2030 scenario than for WEM2030 scenario. The increases in AOT40
for WAM2030 occur mainly around Barcelona and the industrial areas of Asturias and,
although there are also small increases, over Madrid. The increases in SOMO35 for this
scenario are also located in these areas, as well as some other industrial areas in Galicia
(Northwestern Spain) and in the North (around Bilbao). Increases in the annual mean
concentrations are located in the previously mentioned areas of Asturias, Madrid, and
Barcelona, as well as other small areas in northwestern Spain (Galicia), northern Spain,
southern Spain (Andalucía), and the Mediterranean coast.
Atmosphere 2021, 12, x FOR PEER REVIEW 10 of 22
Figure 7. Relative difference (%) maps between the WEM2030 scenario and the reference case 2016
for different O3 metrics.
Figure 8. Relative difference (%) maps between the WAM2030 scenario and the reference case 2016
for different O3 metrics.
3.3. SO2
For the annual mean SO2 concentration limit value for the protection of vegetation
(20 µg m−3), one non-compliant air quality zone is estimated for 2016 in an industrial ar-
ea in northern Spain (Avilés). General improvements are expected with the implementa-
tion of the measures in the WEM2030 and WAM2030 scenarios, leading to the disap-
pearance of this non-compliant area in both scenarios. In the WEM2030 scenario, a few
Figure 7.
Relative difference (%) maps between the WEM2030 scenario and the reference case 2016
for different O3metrics.

Atmosphere 2021,12, 158 10 of 21
Atmosphere 2021, 12, x FOR PEER REVIEW 10 of 22
Figure 7. Relative difference (%) maps between the WEM2030 scenario and the reference case 2016
for different O3 metrics.
Figure 8. Relative difference (%) maps between the WAM2030 scenario and the reference case 2016
for different O3 metrics.
3.3. SO2
For the annual mean SO2 concentration limit value for the protection of vegetation
(20 µg m−3), one non-compliant air quality zone is estimated for 2016 in an industrial ar-
ea in northern Spain (Avilés). General improvements are expected with the implementa-
tion of the measures in the WEM2030 and WAM2030 scenarios, leading to the disap-
pearance of this non-compliant area in both scenarios. In the WEM2030 scenario, a few
Figure 8.
Relative difference (%) maps between the WAM2030 scenario and the reference case 2016
for different O3metrics.
3.3. SO2
For the annual mean SO
2
concentration limit value for the protection of vegetation
(20
µ
g m
−3
), one non-compliant air quality zone is estimated for 2016 in an industrial area
in northern Spain (Avilés). General improvements are expected with the implementation
of the measures in the WEM2030 and WAM2030 scenarios, leading to the disappearance
of this non-compliant area in both scenarios. In the WEM2030 scenario, a few small areas
in red can be seen in Figure 9, indicating an increase in SO
2
. This is due to an increase in
fugitive emissions from oil refining and storage occurring in a few industrial areas, when
only existing measures are implemented. For the daily value, there are no non-compliant
areas estimated by the model for 2016. Further improvements are still expected from the
implementation of measures in the NAPCP (WAM2030), as it can be inferred from the
relative differences in Figure 9(blue colors indicate a decrease in air concentration).
Atmosphere 2021, 12, x FOR PEER REVIEW 11 of 22
small areas in red can be seen in Figure 9, indicating an increase in SO2. This is due to an
increase in fugitive emissions from oil refining and storage occurring in a few industrial
areas, when only existing measures are implemented. For the daily value, there are no
non-compliant areas estimated by the model for 2016. Further improvements are still
expected from the implementation of measures in the NAPCP (WAM2030), as it can be
inferred from the relative differences in Figure 9 (blue colors indicate a decrease in air
concentration).
Figure 9. SO2 annual concentration map for 2016 (µg m−3) and relative differences (%) with respect
to 2016 for the WEM2030 scenario (middle) and WAM2030 scenario (right).
3.4. PM2.5 and PM10
No non-compliant air quality zones for the annual limit value for PM2.5 (25 µg m−3)
were estimated by the modeling approach in 2016 (Figure 10, first row, left column),
with annual mean values even lower than 20 µg m−3 (the PM2.5 exposure concentration
obligation based on 3-year average). The modeling approach estimates even lower con-
centrations in the WEM2030 and WAM2030 scenarios in the NAPCP, as a result of the
decrease in PM2.5 emissions (Figure 10, blue colors in middle and right columns indi-
cating a decrease in concentration). Similarly, no exceedances were estimated for the
PM10 annual mean (limit value 40 µg m−3) in 2016 (Figure 10, second row, left column).
Since the model resolution was approximately 10 × 10 km2, local exceedances could be
present in 2016, and even in WAM2030, at sites with very low spatial representativeness,
such as some industrial areas in Northern Spain. In the same model grid cell, several air
quality stations can be present and the process of combining the model with the mean
value of the stations within the cell could mask the high localized concentrations.
With regards to the daily mean PM10 concentrations (50 µg m−3), some
non-compliant areas were estimated in 2016 in Granada (southern Spain), Asturias in the
north, and Galicia in the northwest. These non-compliant areas disappear for both 2030
projected scenarios.
There is no specific consideration of dust events in these simulations.
Figure 9.
SO
2
annual concentration map for 2016 (
µ
g m
−3
) (
left
) and relative differences (%) with
respect to 2016 for the WEM2030 scenario (middle) and WAM2030 scenario (right).

Atmosphere 2021,12, 158 11 of 21
3.4. PM2.5 and PM10
No non-compliant air quality zones for the annual limit value for PM2.5 (25 µg m−3)
were estimated by the modeling approach in 2016 (Figure 10, first row, left column), with
annual mean values even lower than 20
µ
g m
−3
(the PM2.5 exposure concentration obliga-
tion based on 3-year average). The modeling approach estimates even lower concentrations
in the WEM2030 and WAM2030 scenarios in the NAPCP, as a result of the decrease in
PM2.5 emissions (Figure 10, blue colors in middle and right columns indicating a decrease
in concentration). Similarly, no exceedances were estimated for the PM10 annual mean
(limit value 40
µ
g m
−3
) in 2016 (Figure 10, second row, left column). Since the model
resolution was approximately 10
×
10 km
2
, local exceedances could be present in 2016,
and even in WAM2030, at sites with very low spatial representativeness, such as some
industrial areas in Northern Spain. In the same model grid cell, several air quality stations
can be present and the process of combining the model with the mean value of the stations
within the cell could mask the high localized concentrations.
Atmosphere 2021, 12, x FOR PEER REVIEW 12 of 22
Figure 10. PM2.5 annual mean concentrations (top row, left, µg m−3). PM10 annual mean concen-
trations (middle row, left) and PM10 daily mean concentrations (36th value, bottom row, left) for
2016 (µg m−3). Relative differences (%) with respect to 2016 are shown for the WEM2030 scenario
(middle column) and WAM2030 (right column).
4. Discussion
In the previous section, we described the improvements in air quality expected for
several pollutants for two projected emission scenarios (WEM2030 and WAM2030). A
general improvement in air quality across Spain is expected for NO2, SO2, and PM, with
the largest improvements expected for WAM2030, the emission scenario with the largest
reductions.
The case of ozone deserves special attention. As we have shown in the previous
section, although the number of non-compliant areas is expected to be reduced, the im-
provement over some areas is not as large as may have been expected. Different metrics
correspond to different air quality impacts. For instance, the annual mean does not im-
prove over large areas, and even increases in some areas. Although this metric is not
considered in the current EU Directive on ambient air quality and cleaner air for Europe
(2008/50/EC), the analysis of the expected impact on this metric of the two emission sce-
narios considered here provides useful insight into the O3 photochemistry of the scenar-
ios. As shown in the results, the implementation of the measures in the WAM2030 sce-
nario (and to a lesser extent those of WEM2030) is expected to decrease maximum hour-
ly O3 concentrations across the country but increase annual mean ozone concentrations
in many areas.
Figure 10.
PM2.5 annual mean concentrations (
top
row, left,
µ
g m
−3
). PM10 annual mean concen-
trations (
middle
row, left) and PM10 daily mean concentrations (36th value,
bottom
row, left) for
2016 (
µ
g m
−3
). Relative differences (%) with respect to 2016 are shown for the WEM2030 scenario
(middle column) and WAM2030 (right column).

Atmosphere 2021,12, 158 12 of 21
With regards to the daily mean PM10 concentrations (50
µ
g m
−3
), some non-compliant
areas were estimated in 2016 in Granada (southern Spain), Asturias in the north, and Galicia
in the northwest. These non-compliant areas disappear for both 2030 projected scenarios.
There is no specific consideration of dust events in these simulations.
4. Discussion
In the previous section, we described the improvements in air quality expected for
several pollutants for two projected emission scenarios (WEM2030 and WAM2030). A
general improvement in air quality across Spain is expected for NO
2
, SO
2
, and PM, with
the largest improvements expected for WAM2030, the emission scenario with the largest
reductions.
The case of ozone deserves special attention. As we have shown in the previous section,
although the number of non-compliant areas is expected to be reduced, the improvement
over some areas is not as large as may have been expected. Different metrics correspond to
different air quality impacts. For instance, the annual mean does not improve over large
areas, and even increases in some areas. Although this metric is not considered in the
current EU Directive on ambient air quality and cleaner air for Europe (2008/50/EC), the
analysis of the expected impact on this metric of the two emission scenarios considered
here provides useful insight into the O
3
photochemistry of the scenarios. As shown in
the results, the implementation of the measures in the WAM2030 scenario (and to a lesser
extent those of WEM2030) is expected to decrease maximum hourly O
3
concentrations
across the country but increase annual mean ozone concentrations in many areas.
During sunny hours, the photolysis of NO
2
is the dominant reaction, leading to the
formation of O
3
. The reduction in NO
x
emissions in WAM2030 will result in less NO
2
photolysis and, consequently, less O
3
formation. During the night the dominant reaction is
the one between O
3
and NO (NO + O
3→
NO
2
+ O
2
). In this case, the reduction in NO
x
emissions in WAM2030 reduces NO concentrations, which in turn reduces O
3
removal
through this NO titration reaction. Therefore, this leads to an increase in O
3
during the
periods with no or low levels of sunlight (summer nights and winter). This behavior is
illustrated in Figures 11 and 12, which show the differences in hourly O
3
concentrations for
WAM2030 with respect to 2016 for one day in January 2016 (Figure 11) and one day in July
(Figure 12). While the maps in Figure 11 (winter) show that O3concentrations increase or
remain the same for all hours of the day, the maps in Figure 12 (summer) show decreases
in concentrations during the middle part of the day but increases are expected in some
areas (mostly urban and industrial areas) during the night.
In addition to this titration effect, we also need to consider whether there could be an
effect of changes in NMVOC emissions. At a national scale, emissions change very little for
the WAM2030 scenario (decrease of 0.4% fin NMVOC emissions versus a decrease of 34%
in NO
x
emissions). Nevertheless, there are some spatial differences, as shown in the maps
in Appendix E, with slight increases in northeastern Spain and the Mediterranean area
(mainly due to road transport, combustion in manufacturing industry, and agriculture and
farming sectors). These increases could potentially lead to an increase in O
3
concentrations.
However, two facts support the hypothesis that changes in NO titration have the greatest
influence on the increase in O
3
concentrations during summer nights and winter. The first
is that the model estimates an increase in daytime O
3
during the winter, which corresponds
to the period when there is a reduction in NMVOC concentration over the whole domain
(Appendix F, showing monthly means in January and July), as a result of emission temporal
profiles for the different emission sectors. The second is that, in summer, the increases in
O
3
are produced at night, evening, or early in the morning, when O
3
formation rates are
low or zero, and are strongly outweighed by the removal processes, mainly NO titration,
but also others such as the reaction with alkenes [
17
]. Nevertheless, in the scenarios for
2030, summer NMVOC concentrations are estimated to be higher than in 2016 over some
areas, which would decrease (not increase) nighttime O
3
concentrations due to more active
removal of O
3
by (more) alkenes. More research is currently being done by the authors

Atmosphere 2021,12, 158 13 of 21
to investigate the role of chemistry and the contributions of different source types using
higher resolution simulations for different regions of Spain.
Atmosphere 2021, 12, x FOR PEER REVIEW 14 of 22
large reduction in NOx directly causes an increase in O3 due to a reduction in titration
with NO (in addition to the fact that a warming climate may lead to increased emissions
of biogenic VOCs). For AOT40 and SOMO35, which are based on longer time periods
than the hourly maximum or daily maximum 8-hour metrics and are calculated using a
threshold value, the modeling approach estimates increases in some areas (mainly in
Madrid, Barcelona, and Asturias), despite a general improvement in the rest of the
country.
Figure 11. Hourly O3 concentration difference maps (µg m−3) between the WAM2030 scenario and
2016 for one day in winter (18 January 2016) (µg m−3). Red colors indicate an increase in ozone for
the WAM2030 scenario.
Figure 12. Hourly O3 concentration difference maps (µg m−3) between the WAM2030 scenario and
2016 for one day in summer (18 July 2016). Red colors indicate an increase in ozone for the
WAM2030 scenario and blue colors indicate a reduction.
Since NOx (and NMVOC) emissions have been reduced in the past, it is interesting
to evaluate the effect of these emission reductions on the observed O3 and NO2 concen-
trations. To look at this, we analyzed data for stations within the Spanish air quality
Figure 11. Hourly O3concentration difference maps (µg m−3) between the WAM2030 scenario and
2016 for one day in winter (18 January 2016) (
µ
g m
−3
). Red colors indicate an increase in ozone for
the WAM2030 scenario.
Atmosphere 2021, 12, x FOR PEER REVIEW 14 of 22
large reduction in NOx directly causes an increase in O3 due to a reduction in titration
with NO (in addition to the fact that a warming climate may lead to increased emissions
of biogenic VOCs). For AOT40 and SOMO35, which are based on longer time periods
than the hourly maximum or daily maximum 8-hour metrics and are calculated using a
threshold value, the modeling approach estimates increases in some areas (mainly in
Madrid, Barcelona, and Asturias), despite a general improvement in the rest of the
country.
Figure 11. Hourly O3 concentration difference maps (µg m−3) between the WAM2030 scenario and
2016 for one day in winter (18 January 2016) (µg m−3). Red colors indicate an increase in ozone for
the WAM2030 scenario.
Figure 12. Hourly O3 concentration difference maps (µg m−3) between the WAM2030 scenario and
2016 for one day in summer (18 July 2016). Red colors indicate an increase in ozone for the
WAM2030 scenario and blue colors indicate a reduction.
Since NOx (and NMVOC) emissions have been reduced in the past, it is interesting
to evaluate the effect of these emission reductions on the observed O3 and NO2 concen-
trations. To look at this, we analyzed data for stations within the Spanish air quality
Figure 12. Hourly O3concentration difference maps (µg m−3) between the WAM2030 scenario and
2016 for one day in summer (18 July 2016). Red colors indicate an increase in ozone for the WAM2030
scenario and blue colors indicate a reduction.
Previous studies by the author’s group [
18
] have indicated that the largest contribu-
tion to O
3
concentrations in Spain come from the global background (model boundary
conditions), especially during the winter, when it can exceed 80
µ
g m
−3
. This suggests that
lower NO titration is indeed decreasing the removal of O
3
background over large areas.
More studies are being done by the authors to investigate this aspect. Other authors have
also highlighted the relevance of background contributions. In summer, [19] also showed
the contribution of local emissions (and also confirmed by on-going work of the authors
of this paper). In this case, over areas with high local emissions, according to the model,
locally formed ozone will be less efficiently removed when NO titration is reduced.

Atmosphere 2021,12, 158 14 of 21
Consequently, the reduction in NO
x
emissions, therefore, could lead to opposing
effects, depending on the metric considered. For annual mean concentrations, which
are influenced by winter and nighttime concentrations, the reduced NO titration can be
important, leading to increased annual mean concentrations of O
3
over some areas. For
metrics less influenced by nighttime concentrations (maximum 1 h, daily maximum 8 h,
SOMO35, AOT40) the effect is smaller, leading to a more general reduction in the metric
(the reduction in NO
x
emissions at higher radiation hours reduce O
3
formation, leading to
an improvement in air quality). These results are along the lines of the observed effect of
COVID-19 lockdown emission reductions in the UK [
20
]. The authors found that a large
reduction in NO
x
directly causes an increase in O
3
due to a reduction in titration with NO
(in addition to the fact that a warming climate may lead to increased emissions of biogenic
VOCs). For AOT40 and SOMO35, which are based on longer time periods than the hourly
maximum or daily maximum 8-hour metrics and are calculated using a threshold value,
the modeling approach estimates increases in some areas (mainly in Madrid, Barcelona,
and Asturias), despite a general improvement in the rest of the country.
Since NO
x
(and NMVOC) emissions have been reduced in the past, it is interesting to
evaluate the effect of these emission reductions on the observed O
3
and NO
2
concentrations.
To look at this, we analyzed data for stations within the Spanish air quality network for
the period 2002–2018. During this period, NO
x
and NMVOC emissions were reduced
by 45 and 29% respectively, according to the NEI. After filtering for data coverage (sites
with >75% of the years in the selected period with each year having >75% data coverage),
140 stations were selected. The trends in daily minimum, daily mean, and daily maximum
concentrations were calculated for each station using the seasonal Mann-Kendall method
([
21
,
22
] with modifications from [
23
]) and the relative trends calculated according to [
24
].
For NO
2
, a general decrease in minimum, mean, and maximum concentrations was ob-
served at the vast majority of rural, suburban, and urban sites (Figure 13). However, for O
3
,
large increases (>2% per year) in minimum O
3
concentrations were observed at most sites.
Smaller increases were observed for most sites for annual mean concentrations and the
smallest trends were observed for the maximum concentrations, with decreasing trends at
many sites. These observed past trends are consistent with the changes in O
3
concentrations
estimated in our study for the WAM2030 scenario (decrease in peak concentrations and an
increase in mean concentrations, especially in areas with larger precursor emissions), which
should give some confidence in the results of the present study. In fact, in this study of past
trends, NMVOCs emissions decreased, which avoids a potential impact of an increase in
O
3
due to a slight increase in NMVOCS (seen in some areas for the WAM2030 scenario,
as discussed before). However, it should be noted that the changes in past concentrations
were not solely affected by changes in emissions but were also influenced by other factors,
such as changes in background O3, meteorological conditions.
Atmosphere 2021, 12, x FOR PEER REVIEW 15 of 22
network for the period 2002–2018. During this period, NOx and NMVOC emissions were
reduced by 45 and 29% respectively, according to the NEI. After filtering for data cov-
erage (sites with >75% of the years in the selected period with each year having >75%
data coverage), 140 stations were selected. The trends in daily minimum, daily mean,
and daily maximum concentrations were calculated for each station using the seasonal
Mann–Kendall method ([21] and [22] with modifications from [23]) and the relative
trends calculated according to [24]. For NO2, a general decrease in minimum, mean, and
maximum concentrations was observed at the vast majority of rural, suburban, and ur-
ban sites (Figure 13). However, for O3, large increases (>2% per year) in minimum O3
concentrations were observed at most sites. Smaller increases were observed for most
sites for annual mean concentrations and the smallest trends were observed for the
maximum concentrations, with decreasing trends at many sites. These observed past
trends are consistent with the changes in O3 concentrations estimated in our study for
the WAM2030 scenario (decrease in peak concentrations and an increase in mean con-
centrations, especially in areas with larger precursor emissions), which should give
some confidence in the results of the present study. In fact, in this study of past trends,
NMVOCs emissions decreased, which avoids a potential impact of an increase in O3 due
to a slight increase in NMVOCS (seen in some areas for the WAM2030 scenario, as dis-
cussed before). However, it should be noted that the changes in past concentrations
were not solely affected by changes in emissions but were also influenced by other factors,
such as changes in background O3, meteorological conditions.
Figure 13. Box plots showing the slope of the trends of daily minimum, daily mean, and daily
maximum concentrations observed during the period 2002–2018 for all stations meeting the data
coverage criteria. O3 (red) has mostly positive slopes, indicating an increase for this period,
whereas the trends of NO2 (blue) are mostly negative, indicating decreases in NO2 concentrations.
It is important to bear in mind that the methodology applied for the estimation of
air quality impacts has several limitations. First, the use of meteorology from one single
year limits the applicability of the results to those conditions. Other meteorological con-
ditions should be considered in order to provide more robust conclusions on the effects
of changes in emissions (this is currently being done by the authors). The model spatial
resolution (~10 × 10 km) is another source of uncertainty for the model estimates, and
also for the correction procedure, since some of the observed exceedances are for sites
influenced by traffic emissions or very local effects, which the modeling approach can-
not take into account, unless a higher spatial resolution is used. Another source of un-
certainty is related to the fact that emission reductions were provided by the Ministry for
the Ecological Transition and Demographic Challenge, (MITERD), for the third level of
NFR emission categories (Appendix D) and were uniformly applied across the nation.
This approach will add uncertainty to the spatial distribution of future emissions. It also
should be noted that reductions in PM10 are not considered in the NAPCP, and there-
fore, only reductions in the fine fraction (PM2.5) are included in PM10 reductions. Fi-
nally, the correction of model results for future scenarios in order to estimate
non-compliant zones can produce deviations and potentially overestimate future con-
Figure 13.
Box plots showing the slope of the trends of daily minimum, daily mean, and daily
maximum concentrations observed during the period 2002–2018 for all stations meeting the data
coverage criteria. O
3
(red) has mostly positive slopes, indicating an increase for this period, whereas
the trends of NO2(blue) are mostly negative, indicating decreases in NO2concentrations.

Atmosphere 2021,12, 158 15 of 21
It is important to bear in mind that the methodology applied for the estimation
of air quality impacts has several limitations. First, the use of meteorology from one
single year limits the applicability of the results to those conditions. Other meteorological
conditions should be considered in order to provide more robust conclusions on the
effects of changes in emissions (this is currently being done by the authors). The model
spatial resolution (~10
×
10 km) is another source of uncertainty for the model estimates,
and also for the correction procedure, since some of the observed exceedances are for
sites influenced by traffic emissions or very local effects, which the modeling approach
cannot take into account, unless a higher spatial resolution is used. Another source of
uncertainty is related to the fact that emission reductions were provided by the Ministry
for the Ecological Transition and Demographic Challenge, (MITERD), for the third level
of NFR emission categories (Appendix D) and were uniformly applied across the nation.
This approach will add uncertainty to the spatial distribution of future emissions. It also
should be noted that reductions in PM10 are not considered in the NAPCP, and therefore,
only reductions in the fine fraction (PM2.5) are included in PM10 reductions. Finally,
the correction of model results for future scenarios in order to estimate non-compliant
zones can produce deviations and potentially overestimate future concentrations. A more
complete simulation considering changes in emissions from other countries and projected
emissions for international shipping would provide a more realistic view of a future
situation, although this would not be useful for estimating the effects of the implementation
of measures in the national programme (the objective of the present study).
5. Conclusions
The potential impacts on air quality of the implementation of the Spanish National
Air Pollution Control Programme (NAPCP), approved in Spain, in September 2019 [
4
],
have been evaluated using an air quality model. The scenario with additional measures
projected for 2030 (WAM2030) in the NAPCP, includes eight packages of measures (and
five additional ones, considered as target packages), containing 50 measures (plus seven
additional ones in the target packages). With these measures, a reduction of 0.4% and 34%
in NMVOCs and NOx, respectively, is expected at a national scale.
The study of air quality impacts carried out using the CHIMERE model indicates a
general improvement in air quality for the WAM2030 scenario. Under this scenario no
non-compliant air quality zones are expected for annual SO
2
, NO
2
, and PM10 indicators
(except for very local areas that cannot be assessed with the methodology applied). Despite
an improvement, there are still expected to non-compliant zones for O
3
. Although NO
x
emissions are reduced by 32.9% in the WAM2030 scenario, there are still expected to be 31
non-compliant zones for maximum daily 8-hour concentrations (down from 42 in 2016)
and six non-compliant zones for hourly concentrations (down from 22 in 2016). A decline
of NO
x
emissions reduces the removal of O
3
by NO titration in winter (mainly reducing
the removal of background O
3
), and also in summer nights over areas with high NO
x
emissions (e.g., Madrid and Barcelona), potentially leading to increased concentrations
of O
3
. By contrast, during the hours with high solar radiation, there is a reduction in
O
3
over these areas, as a consequence of reduced O
3
formation, mainly driven by lower
NO
x
emissions. Therefore, different net effects can be expected from a reduction in NO
x
emissions. For metrics least influenced by winter and nighttime values (e.g., maximum
hourly concentrations) the highest benefits are expected, whereas the annual mean will
not change or even increase for some areas, especially those with notable NO titration
of local emissions at night (Asturias, Madrid, and Barcelona, as well as other small areas
in northwestern Spain (Galicia), northern Spain, southern Spain (Andalucía) and the
Mediterranean coast). Although an almost net-zero change in NMVOC emissions is
estimated for the WAM2030 scenario, there are some small emission increases in some
areas, mainly around Barcelona and the Mediterranean area. This increase in NMVOC
emissions in WAM2030 could, nevertheless, reduce the effectiveness of the NOxemission
reduction measures during the day in these areas and should be investigated further. It

Atmosphere 2021,12, 158 16 of 21
is also important to investigate interactions of pollutants from different sources, such as
biogenic emissions, shipping and transboundary air pollution transport, especially over
some affected areas.
These results are in line with observed trends in O
3
over the last two decades, for
which daily maximum hourly values have changed very little (except at rural background
stations where there has been slight decrease) but a slight increase has been found for the
daily minimum concentrations.
Finally, it should be noted that the methodology used is subject to several sources
of uncertainty related to the air quality modeling (fixed meteorology (2016), emission
reductions at a national scale, fixed boundary conditions, and spatial resolution) and the
procedure to correct model estimates using observations. Further studies are being done
to address some of these uncertainties, for instance, considering other meteorological
conditions and increasing the spatial resolution of the model.
Author Contributions:
Conceptualization, M.G.V. and F.M.; methodology, M.G.V. and M.R.T.; soft-
ware, M.G.V. and M.R.T.; validation, M.R.T. and J.L.G.; formal analysis, M.G.V., M.R.T., V.G., and
J.L.G.; investigation, M.G.V., M.R.T., and F.M.; resources, A.A., E.S., and A.B.; data curation, M.G.V.,
M.R.T., V.G., and J.L.G.; writing—original draft preparation, M.G.V.; writing—review and editing,
M.G.V., M.R.T., F.M., J.-L.S., J.L.G., A.R.G., and Y.L.; supervision, M.G.V. and F.M.; project administra-
tion, M.G.V. and J.-L.S.; funding acquisition, M.G.V., J.-L.S., F.M., and M.R.T. All authors have read
and agreed to the published version of the manuscript.
Funding:
This research was funded by the National Agency for Scientific Research of the Spanish
Ministry of Science and Innovation, through the Retos-AIRE project (Air pollution mitigation actions
for environmental policy support. Air quality multiscale modeling and evaluation of health and
vegetation impacts), with grant number “RTI2018-099138-B-100. Partial funding was also provided
by the Ministry for the Ecological Transition and Demographic Challenge through a contract with
TRAGSATEC.
Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.
Acknowledgments:
We thank the Ministry for the Ecological Transition and Demographic Challenge
(MITERD) and TRAGSATEC for the provision of the emission reduction for the measures in the
NAPCP and related discussion. We also acknowledge MITERD for providing data from air quality
stations and information. We are also grateful for the services offered by the European Center for
Medium-Range Weather Forecasts (ECMWF), including the provision of meteorological modeling
data with thanks also to AEMET for managing access to this information.
Conflicts of Interest:
The authors declare no conflict of interest. The Ministry for the Ecological
Transition and Demographic Challenge (MITERD) designed the measures in the National Air Pol-
lution Control Programme (NAPCP). MITERD and TRAGSATEC provided the pollutant emission
reduction estimates for the measures in the NAPCP.

Atmosphere 2021,12, 158 17 of 21
Appendix A
Atmosphere 2021, 12, x FOR PEER REVIEW 18 of 22
Appendix A
Figure A1. Measures Included in the WAM2030 scenario.
E.1 Energy mix

New renewable energy installations
 Integration of renewable energy into the grid
 Self-production and distributed generation
 Increased use of renewable gases
 Refurbishment of existing renewable energy installations
 Power purchase agreements between energy provider and consumer
 Specific programs for biomass use
 Unique projects and renewable energy for islands
T.1 Emission
reductions for
road transport,
rail, aviation and
shipping

Advanced biofuels for transport
 Changes in mode of transportation
 More efficient use of transport
 Fleet renewal
 Promotion of electric vehicles
 Refuelling/recharging points for alternative fuels
I.1 Improved energy
efficiency in the
industrial and
manufacturing
sectors

Support for industry
 Framework for the development of renewable thermal technologies
 Improvement of the technology and management systems for
industrial processes
EE.1 Improved energy
efficiency in the
residential,
commercial,
institutional and
other sectors

Integration of renewable thermal technologies
 Subsidies for installations in buildings and heating networks
 Improved energy efficiency in the residential sector
 Renewal of installations in residential buildings
 Improved energy efficiency in public buildings and the tertiary
sector
 Improved energy efficiency of large heating/air conditioning
systems in the tertiary sector and public infrastructure
 Improved energy efficiency of agricultural installations, irrigation
communities and agricultural machinery
RS.1 Waste

Increased domestic and community composting
 Renovation of composting infrastructure
 Separation of biowastes for biomethanization
 Reduction of food waste
 Increased paper collection in the municipal channel
 Increased collection of domestic cooking oil
 Increased collection of textiles
 Use of oxidising covers on landfill sites
A.1 Use of fertiliser
plans

Set a limit of 30% of plant N requirements by urea
 Set conditions for urea application
 Ban the surface application of slurries and other substances with a
water content > 40%
 Low emission application technologies
 Incorporation of solid organic fertilisers following application
 Use of fertiliser plans
 Soil nitrogen balance calculations
 Inclusion of environmental objectives in fertiliser plans
 Registro de operaciones en el cuaderno de explotación
 Registering activities in the farm logbook
Figure A1. Measures Included in the WAM2030 scenario.

Atmosphere 2021,12, 158 18 of 21
Appendix B
Atmosphere 2021, 12, x FOR PEER REVIEW 19 of 22
Appendix B
Figure A2. Target Measures (not included yet in the WAM2030 scenario emission reductions).
Appendix C
Figure A3. Emissions in 2016 (kt/year) and reduction of emissions (%) for the measures included in WEM2030 and
WAM2030 scenarios for the NFR (nomenclature for reporting) source categories (described in Appendix D), relative to
2016.
O.1 Reduction of emissions from
residential wood burning

Reduce fine particulate emissions from fires
and wood-burning stoves in rural areas
O.2 Reduction of emissions from the
domestic use of solvents and
paints

Responsible domestic use of solvents and
paints
O.3 Analysis of the potential
pollution from small and
medium combustion plants

Analysis of the potential pollution reduction
for small and medium combustion plants
(between 500 kW and 50 MW)
O.4 Reduction of harbour emissions

O.4.1.- Promotion of alternative and
renewable energies in ports
 O.4.2.- Control of diffuse emissions from
ports
 O.4.3.- Fiscal measures
O.5 Public awareness raising

Public awareness raising
NH3 NMVOC NOx PM2.5 SOx
kt % % kt % % kt % % kt % % kt % %
NFR CODE 20 16 WEM2030 WAM2030 2016 WEM2030 WAM20 30 2016 WEM2030 WAM2030 2016 WEM203 0 WAM2030 2016 WEM203 0 WAM2030
1A1 0.1 2% 27% 6.4 -39% -146% 107.9 26% 58% 4.1 -18% -63% 89.5 37% 83%
1A2 1.2 -9% -70% 16.4 -10% -40% 106.2 1% 20% 8.2 -3% -28% 58.0 10% 39%
1A3a,c,d,e 0.0 -27% -46% 2.2 -37% -52% 44.8 -4% -7% 1.4 18% 1 8% 5.8 4 8% 57%
1A3bi 2.2 -118% -48% 7.6 -89 % -25% 145.7 32% 70% 4.0 38% 75% 0 .2 -5% 41%
1A3bii 0.0 3 4% 37% 0 .6 38% 65% 21.8 37% 65% 0.6 38% 66% 0 .0 38% 66 %
1A3biii 0.2 -6% 8% 1.7 -251% 17% 9 0.2 -38% 1 1% 1.1 6% 20% 0 .1 -2% 12%
1A3biv 0.0 -142% -82 % 12.7 -142% -82% 2.0 -141% -81 % 0.2 -142% -81% 0.0 -142 % -81%
1A3bv 0.0 0 % 0% 1.8 -96 % -7% 0.0 0% 0% 0.0 0% 0% 0.0 0 % 0%
1A3bvi 0.0 0% 0% 0.0 0% 0 % 0.0 0% 0% 3.1 5% 41% 0.0 0% 0%
1A3bvii 0.0 0% 0% 0 .0 0% 0% 0.0 0% 0% 1.8 6% 42% 0.0 0 % 0%
1A4 7.4 24% -6 0% 48.6 20 % 41% 136.9 9% 19% 56.4 31% 54% 19.1 58 % 67%
1A5 0.0 -19% -19% 0.1 -76% -76% 5.0 -1 9% -19% 0.0 -19% -19 % 0 .1 -19% -19%
1B 0.0 48% 6 5% 35.7 -16 % 4% 4.2 -1 5% 5% 0.4 10% 29% 28 .5 -15% 5%
2A,B,C,H,I,J,K,L
30.4 -2% -2% 62.7 -2% -2% 4.9 -11% -11% 7.5 -7% -7% 14 .7 21% 21 %
2D, 2G 0.0 -2 9% -29% 279.0 2% 2 % 0 .0 -29 % -29% 0.3 -25% -25% 0.0 -29% -29 %
3B1a 19.9 6 % 39% 10.2 6% 6% 0.5 6% 6% 0.1 6% 6% 0.0 0% 0%
3B1b 41.3 5% 37% 25 .8 0% 0% 1.2 -2% -2% 0.4 0% 0% 0.0 0% 0%
3B2 9.5 16% 16% 3.4 18% 18% 0.4 17% 1 7% 0.1 18% 18% 0 .0 0% 0%
3B3 79.7 -9% 9% 15.6 -10% -10% 0.4 -9% -9% 0.1 -10% -10% 0 .0 0% 0%
3B4d 0.8 10 % 10% 1.5 10% 10% 0.0 10 % 10% 0.0 10% 1 0% 0.0 0% 0 %
3B4e 5.2 1 % 1% 2.8 -4% -4% 0.2 -4% -4% 0 .0 -4% -4% 0 .0 0% 0%
3B4f 0.1 19% 19 % 0.1 11% 11% 0.0 12% 12% 0.0 11% 11 % 0.0 0% 0%
3B4g 43.8 1 % 20% 24.4 1% 1% 1.8 0% 0% 0.6 1% 1% 0.0 0% 0%
3D 247.8 7% 26% 22 .1 0% 0% 54.5 -1% 3% 1.5 0% 0% 0.0 0 % 0%
3F,I 0.7 5% 5% 0.2 5% 5% 0 .7 5% 5% 1.6 5% 5% 0.2 5% 5%
5 1.8 -7% -28% 12.2 1 0% 52% 36.2 0 % 58 % 34.7 1% 56% 1.7 16% 6 4%
TOTAL 492 .2 3% 20% 593.9 -4.6% 0.4% 765.5 7% 33 % 128.4 14% 39% 218 24% 5 5%
Figure A2.
Target Measures (not included yet in the WAM2030 scenario emission reduc-
tions).
Appendix C
Atmosphere 2021, 12, x FOR PEER REVIEW 19 of 22
Appendix B
Figure A2. Target Measures (not included yet in the WAM2030 scenario emission reductions).
Appendix C
Figure A3. Emissions in 2016 (kt/year) and reduction of emissions (%) for the measures included in WEM2030 and
WAM2030 scenarios for the NFR (nomenclature for reporting) source categories (described in Appendix D), relative to
2016.
O.1 Reduction of emissions from
residential wood burning

Reduce fine particulate emissions from fires
and wood-burning stoves in rural areas
O.2 Reduction of emissions from the
domestic use of solvents and
paints

Responsible domestic use of solvents and
paints
O.3 Analysis of the potential
pollution from small and
medium combustion plants

Analysis of the potential pollution reduction
for small and medium combustion plants
(between 500 kW and 50 MW)
O.4 Reduction of harbour emissions

O.4.1.- Promotion of alternative and
renewable energies in ports
 O.4.2.- Control of diffuse emissions from
ports
 O.4.3.- Fiscal measures
O.5 Public awareness raising

Public awareness raising
NH3 NMVOC NOx PM2.5 SOx
kt % % kt % % kt % % kt % % kt % %
NFR CODE 20 16 WEM2030 WAM2030 2016 WEM2030 WAM20 30 2016 WEM2030 WAM2030 2016 WEM203 0 WAM2030 2016 WEM203 0 WAM2030
1A1 0.1 2% 27% 6.4 -39% -146% 107.9 26% 58% 4.1 -18% -63% 89.5 37% 83%
1A2 1.2 -9 % -70% 16.4 -10% -40% 106 .2 1% 20% 8.2 -3% -28% 58.0 10 % 39%
1A3a,c,d,e 0.0 -27% -46% 2.2 -37% -52% 44.8 -4% -7% 1.4 18% 1 8% 5.8 4 8% 57%
1A3bi 2.2 -118% -48% 7.6 -89 % -25% 145.7 32% 70% 4.0 38% 75% 0 .2 -5% 41%
1A3bii 0.0 3 4% 37% 0 .6 38% 65% 21.8 37% 65% 0.6 38% 66% 0 .0 38% 66 %
1A3biii 0.2 -6% 8% 1.7 -251% 17% 9 0.2 -38% 1 1% 1.1 6% 20% 0 .1 -2% 12%
1A3biv 0.0 -142% -82 % 12.7 -142% -82% 2.0 -141% -81 % 0.2 -142% -81% 0.0 -142 % -81%
1A3bv 0.0 0 % 0% 1.8 -96 % -7% 0.0 0% 0% 0.0 0% 0% 0.0 0 % 0%
1A3bvi 0.0 0% 0% 0.0 0% 0 % 0.0 0% 0% 3.1 5% 41% 0.0 0% 0%
1A3bvii 0.0 0% 0% 0 .0 0% 0% 0.0 0% 0% 1.8 6% 42% 0.0 0 % 0%
1A4 7.4 24% -6 0% 48.6 20 % 41% 136.9 9% 19% 56.4 31% 54% 19.1 58 % 67%
1A5 0.0 -19% -19% 0.1 -76% -76% 5.0 -1 9% -19% 0.0 -19% -19 % 0 .1 -19% -19%
1B 0.0 48% 6 5% 35.7 -16 % 4% 4.2 -1 5% 5% 0.4 10% 29% 28 .5 -15% 5%
2A,B,C,H,I,J,K,L
30.4 -2% -2% 62.7 -2% -2% 4.9 -11% -11% 7.5 -7% -7% 14 .7 21% 21 %
2D, 2G 0.0 -2 9% -29% 279.0 2% 2 % 0 .0 -29 % -29% 0.3 -25% -25% 0.0 -29% -29 %
3B1a 19.9 6 % 39% 10.2 6% 6% 0.5 6% 6% 0.1 6% 6% 0.0 0% 0%
3B1b 41.3 5% 37% 25 .8 0% 0% 1.2 -2% -2% 0.4 0% 0% 0.0 0% 0%
3B2 9.5 16% 16% 3.4 18% 18% 0.4 17% 1 7% 0.1 18% 18% 0 .0 0% 0%
3B3 79.7 -9% 9% 15.6 -10% -10% 0.4 -9% -9% 0.1 -10% -10% 0 .0 0% 0%
3B4d 0.8 10 % 10% 1.5 10% 10% 0.0 10 % 10% 0.0 10% 1 0% 0.0 0% 0 %
3B4e 5.2 1 % 1% 2.8 -4% -4% 0.2 -4% -4% 0 .0 -4% -4% 0 .0 0% 0%
3B4f 0.1 19% 19 % 0.1 11% 11% 0.0 12% 12% 0.0 11% 11 % 0.0 0% 0%
3B4g 43.8 1 % 20% 24.4 1% 1% 1.8 0% 0% 0.6 1% 1% 0.0 0% 0%
3D 247.8 7% 26% 22 .1 0% 0% 54.5 -1% 3% 1.5 0% 0% 0.0 0 % 0%
3F,I 0.7 5% 5% 0.2 5% 5% 0 .7 5% 5% 1.6 5% 5% 0.2 5% 5%
5 1.8 -7% -28% 12.2 1 0% 52% 36.2 0 % 58 % 34.7 1% 56% 1.7 16% 6 4%
TOTAL 492 .2 3% 20% 593.9 -4.6% 0.4% 765.5 7% 33 % 128.4 14% 39% 218 24% 5 5%
Figure A3.
Emissions in 2016 (kt/year) and reduction of emissions (%) for the measures included in WEM2030 and
WAM2030 scenarios for the NFR (nomenclature for reporting) source categories (described in Appendix D), relative to 2016.

Atmosphere 2021,12, 158 19 of 21
Appendix D
Atmosphere 2021, 12, x FOR PEER REVIEW 20 of 22
Appendix D
Figure A4. NFR source categories (nomenclature for reporting (NFR)).
NFR Code Longname
1A1 Energy industries (Comb us tion in po wer
plan ts & Energy Production)
1A2 Manufacturing Industries and
Construction (Com bus tion in industry
inclu ding Mobile)
1A3b Road Transport
1A3bi Pas senger cars
1A3bii Light duty vehicles
1A3biii Heavy duty vehicles
1A3biv Mopeds & Motorcycles
1A3bv Gasoline evaporatio n
1A3bvi Autom obil e tyre and brake wea r
1A3bvii Automobile road abrasi on
1A3a,c,d,e Off-road transport
1A4 Other sectors (Co mm ercial, institutional,
res iden tial, agriculture and fishing
statio nary and mo bile comb ustion)
1A5 Other
1B Fugitive emiss ions (Fugitive emis si ons
from fuels)
2A,B,C,H,I,J,K
,L
Industrial Processes
2D, 2G Solvent and other product use
3B Animal husbandry and manure
management
3B1a Cattle Dairy
3B1b Cattle Non-Dairy
3B2 Shee p
3B3 Swin e
3B4a Buffalo
3B4d Goats
3B4e Ho rse s
3B4f Mules and as ses
3B4g Pou ltry
3B4h Other
3D Plant production and agricultural soils
3F,I Field burning and other agriculture
5 Was te
Figure A4. NFR source categories (nomenclature for reporting (NFR)).
Appendix E
Atmosphere 2021, 12, x FOR PEER REVIEW 21 of 22
Appendix E
NOx emissions
Relative Differences (%)
WAM20302016
NMVOCs emissions
Relative Differences (%)
WAM20302016

Figure A5. Relative differences (%) of annual NOx (left) and NMVOC (right) emissions between the WAM2030 scenario
and 2016.
Appendix F
NMVOC concentrations
Mean Differences (µg m−3)
WAM20302016 January
NMVOC concentrations
Mean Differences (µg m−3)
WAM20302016 July
Figure A6. Mean monthly differences (µg m−3) between the WAM2030 scenario and 2016 in January
and July.
References
1. European Environment Agency. Air Quality in Europe—2020 Report; EEA Report No 09/2020; European Environment Agency:
Copenhagen, Denmark, 2020.
2. European Environment Agency; Solberg, S.; Walker, S.-E.; Guerreiro, C.; Colette, A. ETC/ATNI: Statistical Modelling for
Long-Term Trends of Pollutants—Use of a GAM Model for the Assessment of Measurements of O3, NO2 and PM; Eionet Report
ETC/ATNI 14/2019; European Topic Centre on Air Pollution, Transport, Noise and Industrial Pollution: Kjeller, Norway, 2019.
3. Ministerio Para la Transición Ecológica y Reto Demográfico. 2020 EVALUACIÓN DE LA CALIDAD DEL AIRE EN ESPAÑA
Año. 2019. Available online:
https://www.miteco.gob.es/images/es/informeevaluacioncalidadaireespana2019_tcm30-510616.pdf (accessed on 25 January
2021).
4. Ministerio Para la Transición Ecológica y Reto Demográfico. 2019. Available online:
https://www.miteco.gob.es/images/es/primerpncca_2019_tcm30-502010.pdf (accessed on 25 January 2021).
5. Menut, L.; Bessagnet, B.; Khvorostyanov, D.; Beekmann, M.; Blond, N.; Colette, A.; Coll, I.; Curci, G.; Foret, G.; Hodzic, A.; et al.
CHIMERE 2013: A model for regional atmospheric composition modelling. Geosci. Model Dev. 2013, 6, 981–1028,
doi:10.5194/gmd-6-981-2013.
6. Pirovano, G.; Balzarini, A.; Bessagnet, B.; Emery, C.; Kallos, G.; Meleux, F.; Mitsakou, C.; Nopmongcol, U.; Riva, G.M.;
Yarwood, G. Investigating impacts of chemistry and transport model formulation on model performance at European scale.
Atmos. Environ. 2012, 53, 93–109, doi:10.1016/j.atmosenv.2011.12.052
Figure A5.
Relative differences (%) of annual NO
x
(left) and NMVOC (right) emissions between the WAM2030 scenario
and 2016.

Atmosphere 2021,12, 158 20 of 21
Appendix F
Atmosphere 2021, 12, x FOR PEER REVIEW 21 of 22
Appendix E
NOx emissions
Relative Differences (%)
WAM20302016
NMVOCs emissions
Relative Differences (%)
WAM20302016

Figure A5. Relative differences (%) of annual NOx (left) and NMVOC (right) emissions between the WAM2030 scenario
and 2016.
Appendix F
NMVOC concentrations
Mean Differences (µg m−3)
WAM20302016 January
NMVOC concentrations
Mean Differences (µg m−3)
WAM20302016 July
Figure A6. Mean monthly differences (µg m−3) between the WAM2030 scenario and 2016 in January
and July.
References
1. European Environment Agency. Air Quality in Europe—2020 Report; EEA Report No 09/2020; European Environment Agency:
Copenhagen, Denmark, 2020.
2. European Environment Agency; Solberg, S.; Walker, S.-E.; Guerreiro, C.; Colette, A. ETC/ATNI: Statistical Modelling for
Long-Term Trends of Pollutants—Use of a GAM Model for the Assessment of Measurements of O3, NO2 and PM; Eionet Report
ETC/ATNI 14/2019; European Topic Centre on Air Pollution, Transport, Noise and Industrial Pollution: Kjeller, Norway, 2019.
3. Ministerio Para la Transición Ecológica y Reto Demográfico. 2020 EVALUACIÓN DE LA CALIDAD DEL AIRE EN ESPAÑA
Año. 2019. Available online:
https://www.miteco.gob.es/images/es/informeevaluacioncalidadaireespana2019_tcm30-510616.pdf (accessed on 25 January
2021).
4. Ministerio Para la Transición Ecológica y Reto Demográfico. 2019. Available online:
https://www.miteco.gob.es/images/es/primerpncca_2019_tcm30-502010.pdf (accessed on 25 January 2021).
5. Menut, L.; Bessagnet, B.; Khvorostyanov, D.; Beekmann, M.; Blond, N.; Colette, A.; Coll, I.; Curci, G.; Foret, G.; Hodzic, A.; et al.
CHIMERE 2013: A model for regional atmospheric composition modelling. Geosci. Model Dev. 2013, 6, 981–1028,
doi:10.5194/gmd-6-981-2013.
6. Pirovano, G.; Balzarini, A.; Bessagnet, B.; Emery, C.; Kallos, G.; Meleux, F.; Mitsakou, C.; Nopmongcol, U.; Riva, G.M.;
Yarwood, G. Investigating impacts of chemistry and transport model formulation on model performance at European scale.
Atmos. Environ. 2012, 53, 93–109, doi:10.1016/j.atmosenv.2011.12.052
Figure A6. Mean monthly differences (µg m−3) between the WAM2030 scenario and 2016 in January and July.
References
1.
European Environment Agency. Air Quality in Europe—2020 Report; EEA Report No 09/2020; European Environment Agency:
Copenhagen, Denmark, 2020.
2.
European Environment Agency; Solberg, S.; Walker, S.-E.; Guerreiro, C.; Colette, A. ETC/ATNI: Statistical Modelling for Long-Term
Trends of Pollutants—Use of a GAM Model for the Assessment of Measurements of O3, NO2 and PM; Eionet Report ETC/ATNI 14/2019;
European Topic Centre on Air Pollution, Transport, Noise and Industrial Pollution: Kjeller, Norway, 2019.
3.
Ministerio Para la Transición Ecológica y Reto Demográfico. 2020 EVALUACIÓN DE LA CALIDAD DEL AIRE EN ESPAÑA Año.
2019. Available online: https://www.miteco.gob.es/images/es/informeevaluacioncalidadaireespana2019_tcm30-510616.pdf
(accessed on 25 January 2021).
4.
Ministerio Para la Transición Ecológica y Reto Demográfico. 2019. Available online: https://www.miteco.gob.es/images/es/
primerpncca_2019_tcm30-502010.pdf (accessed on 25 January 2021).
5.
Menut, L.; Bessagnet, B.; Khvorostyanov, D.; Beekmann, M.; Blond, N.; Colette, A.; Coll, I.; Curci, G.; Foret, G.; Hodzic, A.; et al.
CHIMERE 2013: A model for regional atmospheric composition modelling. Geosci. Model Dev. 2013,6, 981–1028. [CrossRef]
6.
Pirovano, G.; Balzarini, A.; Bessagnet, B.; Emery, C.; Kallos, G.; Meleux, F.; Mitsakou, C.; Nopmongcol, U.; Riva, G.M.; Yarwood,
G. Investigating impacts of chemistry and transport model formulation on model performance at European scale. Atmos. Environ.
2012,53, 93–109. [CrossRef]
7.
Vivanco, M.G.; Palomino, I.; Vautard, R.; Bessagnet, B.; Martín, F.; Menut, L.; Jiménez, S. Multi-year assessment of photochemical
air quality simulation over Spain. Environ. Model. Softw. 2009,24, 63–73. [CrossRef]
8.
Vivanco, M.G.; Palomino, I.; Martín, F.; Palacios, M.; Jorba, O.; Jiménez, P.; Baldasano, J.M.; Azula, O. An evaluation of the
performance of the CHIMERE model over spain using meteorology from MM5 and WRF models. In Computational Science and Its
Applications—ICCSA 2009; Lecture Notes in Computer Science (Including Subseries Lecture Notes in Artificial Intelligence and
Lecture Notes in Bioinformatics); Springer: Berlin/Heidelberg, Germany, 2009; Volume 5592, pp. 107–117.
9.
Vivanco, M.G.; Correa, M.; Azula, O.; Palomino, I.; Martín, F. Influence of model resolution on ozone predictions over Madrid
Area (Spain). In Computational Science and Its Applications—ICCSA 2008; Lecture Notes in Computer Science (Including Subseries
Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Springer: Berlin/Heidelberg, Germany, 2008; Volume
5072, pp. 165–178.
10.
Vivanco, M.G.; Theobald, M.; Garrido, J.L.; Gil, V.; Martín, F. Evaluación de la Calidad del Aire en España Utilizando Modelización Com-
binada con Mediciones; Preevaluación Año 2017; Acuerdo de Encomienda de Gestión 2014–2018 Entre el Ministerio de Agricultura,
Alimentación y Medio Ambiente y CIEMAT; CIEMAT Report; CIEMAT: Madrid, Spain, 2018.
11.
Bessagnet, B.; Pirovano, G.; Mircea, M.; Cuvelier, C.; Aulinger, A.; Calori, G.; Ciarelli, G.; Manders, A.; Stern, R.; Tsyro, S.; et al.
Presentation of the EURODELTA III intercomparison exercise—Evaluation of the chemistry transport models’ performance on
criteria pollutants and joint analysis with meteorology. Atmos. Chem. Phys. 2016,16, 12667–12701. [CrossRef]
12.
EMEP Emissions Database for European Emissions. Available online: https://www.ceip.at/webdab-emission-database/
emissions-as-used-in-emep-models (accessed on 25 January 2021).
13.
Hauglustaine, D.A.; Hourdin, F.; Jourdain, L.; Filiberti, M.-A.; Walters, S.; Lamarque, J.-F.; Holland, E.A. Interactive chemistry in
the Laboratoire de Meteorologie Dynamique general circulation model: Description and background tropospheric chemistry
evaluation. J. Geophys. Res. 2004,109. [CrossRef]

Atmosphere 2021,12, 158 21 of 21
14.
Ginoux, P.; Chin, M.; Tegen, I.; Prospero, J.M.; Holben, B.; Dubovik, O.; Lin, S.J. Sources and distributions of dust aerosols
simulated with the GOCART model. J. Geophys. Res. 2001,106, 20255–20273. [CrossRef]
15.
Martín, F.; Palomino, I.; Vivanco, M.G. Application of a method for combining measured data and modelling results in air quality
assessment in Spain. Rev. Física Tierra 2009,21, 65–78.
16.
Martín, F.; Palomino, I.; Vivanco, M.G. Combination of measured and modelling data in air quality assessment in Spain. Int. J.
Environ. Pollut. 2012,49, 36–44. [CrossRef]
17.
Finlayson–Pitts, B.J.; Pitts, J.N.J. Chemistry of the Upper and Lower Atmosphere: Theory, Experiments, and Applications; Academic
Press: Cambridge, MA, USA, 2000; ISBN 978-0-12-257060-5. [CrossRef]
18.
Theobald, M.R.; Vivanco, M.G.; Gil, V.; Garrido, J.L.; Martín, F. Analysis of the zero-out method of source apportionment for
air quality modelling in Spain. In Proceedings of the 37th International Technical Meeting on Air Pollution Modelling and Its
Application, Katholische Akademie, Hamburg, Germany, 23–27 September 2019; Extended Abstract to Be Published by Springer
in Air Pollution Modeling and Its Application. ITM Springer Book: Cham, Switzerland, 2021; Volume XXVII.
19.
Pay, M.-T.; Gangoiti, G.; Guevara, M.; Napelenok, S.; Querol, X.; Jorba, O.; García-Pando, C.P. Ozone source apportionment
during peak summer events over southwestern Europe. Atmos. Chem. Phys. 2019,19, 5467–5494. [CrossRef] [PubMed]
20.
Lee, J.; Drysdale, W.; Finch, D.P.; Wilde, S.; Palmer, P.I. Palmer UK surface NO2 levels dropped by 42% during the COVID-19
lockdown: Impact on surface O3. Atmos. Chem. Phys. 2020,20, 15743–15759. [CrossRef]
21. Mann, H.B. Nonparametric tests against trend. Economet. J. 1945,13, 245–259. [CrossRef]
22. Kendall, M.G. Rank Correlation Methods; Griffin: London, UK, 1970.
23.
Hirsch, R.M.; Slack, J.R. A nonparametric trend test for seasonal data with serial dependence. Water Resour. Res.
1984
,20, 727–732.
[CrossRef]
24. Sen, P.K. Estimates of the regression coefficient based on Kendall’s tau. J. Am. Stat. Assoc. 1968,63, 1379–1389. [CrossRef]