Research article Atmos. Chem. Phys., 23, 10751–10774, 2023 https://doi.org/10.5194/acp-23-10751-2023 © Author(s) 2023. This work is distributed under the Creative Commons Attribution 4.0 License. Modelling the impacts of emission changes on O3 sensitivity, atmospheric oxidation capacity, and pollution transport over the Catalonia region Alba Badia1 , Veronica Vidal1,2 , Sergi Ventura1 , Roger Curcoll3 , Ricard Segura1 , and Gara Villalba1,4 1 Sostenipra Research Group, Institute of Environmental Sciences and Technology, Universitat Autònoma de Barcelona, 08193 Bellaterra, Barcelona, Spain 2 Departament d’Arquitectura de Computadors i Sistemes Operatius (CAOS), Escola d’Enginyeria, Universitat Autònoma de Barcelona 08193 Bellaterra, Barcelona, Spain 3 Institut de Tècniques Energètiques (INTE), Universitat Politècnica de Catalunya, Barcelona, Spain 4 Department of Chemical, Biological and Environmental Engineering, Universitat Autònoma de Barcelona, 08193 Bellaterra, Barcelona, Spain Correspondence: Alba Badia (alba.badia@uab.cat) Received: 6 February 2023 – Discussion started: 8 March 2023 Revised: 30 July 2023 – Accepted: 1 August 2023 – Published: 28 September 2023 Abstract. Tropospheric ozone (O3 ) is an important surface pollutant in urban areas, and it has complex for- mation mechanisms that depend on the atmospheric chemistry and on meteorological factors. The severe re- ductions observed in anthropogenic emissions during the COVID-19 pandemic can further our understanding of the photochemical mechanisms leading to O3 formation and provide guidance for policies aimed at reducing air pollution. In this study, we use the Weather Research and Forecasting model with Chemistry (WRF-Chem) coupled with the urban canopy building effect parameterization and building energy model (BEP + BEM) to investigate changes in the ozone chemistry over the metropolitan area of Barcelona (AMB) and its atmospheric plume moving northwards, which is responsible for the highest number of hourly O3 exceedances in Spain. The trajectories of the air masses from the AMB to the Pyrenees are studied with the Lagrangian FLEXible PARTi- cle dispersion model with WRF (FLEXPART-WRF). The aim is to investigate the response of ozone chemistry to reduction in precursor emissions (NOx – nitrogen oxides; VOCs – volatile organic compounds). The results show that, with the reduction in emissions, (1) the ozone chemistry tends to enter the NOx -limited or transition regimes, but highly polluted urban areas are still in the VOC-limited regime; (2) the reduced O3 production is overwhelmed by reduced nitric oxide (NO) titration, resulting in a net increase in the O3 concentration (up to 20 %) in the evening; (3) the increase in the maximum O3 level (up to 6 %) during the highest emission-reduction period could be attributed to an enhancement in the atmospheric oxidants hydroxyl and nitrate radical (OH and NO3 ) given their strong link with O3 loss or production chemistry; (4) the daily maximum levels of ozone and odd oxygen species (Ox ) generally decreased (4 %) in May – a period with intense radiation which favours ozone production – with the reduced atmospheric OH and NO3 oxidants, indicating an improvement in the air quality; and (5) ozone precursor concentration changes in the urban plume of Barcelona contribute significantly to the level of pollution along the 150 km south-to-north valley in the Pyrenees. Our results indicate that O3 abatement strategies cannot rely only on NOx emission control but must include a significant reduction in anthropogenic sources of VOCs. In addition, our results show that mitigation strategies intended to reduce O3 should be de- signed according to the local meteorology, air transport, particular ozone regimes, and oxidation capacity of the atmosphere of the urban area. Published by Copernicus Publications on behalf of the European Geosciences Union. 10752 A. Badia et al.: Modelling the impacts of emission changes on O3 chemistry 1 Introduction fective in reducing O3 if NOx -sensitive chemistry dominates. In urban areas, this photochemistry illustrates the difficulties Tropospheric ozone (O3 ) is a radiatively active gas that acts involved in developing policies to reduce O3 in polluted re- as an oxidizing agent and a surface pollutant in urban areas, gions (Sillman, 2003). Thus, a more profound understand- where it is a major component of photochemical smog and ing of the sensitivity of local ozone formation to changes in causes a number of respiratory health effects (Sillman, 2003; NOx and VOC levels is essential for developing effective air Anenberg et al., 2010; Fleming et al., 2018; Sillmann et al., quality policies. Other important processes for removal of O3 2021). In 2019, it was estimated that 365 000 respiratory are reactions of halogen species (Badia et al., 2021a), which mortalities worldwide were due to anthropogenic O3 , rep- remove 30–35 Tg (11 %–15 %) of tropospheric ozone, and resenting an increase of ∼ 16 % with respect to 2010 (GBD dry deposition, which accounts for about 20 % of the O3 lost 2019 Risk Factors Collaborators, 2020). The World Health from the troposphere (Wild, 2007). Organization (WHO) provides guidelines to protect humans In addition to photochemical reactions, the concentration from exposure by reducing the levels of key air pollutants, in- of ozone is sensitive to meteorological variables such as so- cluding ozone. The WHO released revised air quality guide- lar radiation and wind speed and direction (Neiburger, 1969). lines in 2021, keeping the O3 levels to 100 µg m−3 for an 8 h Ozone production is intensified on warm, sunny days when daily average and recommending 60 µg m−3 for an 8 h daily the air is stagnant. Therefore, the increase in the frequency, average during the 6 consecutive months with the highest severity, and duration of heat waves during recent decades average ozone concentrations (World Health Organization, has increased the need to understand the influence of mete- 2021). orological drivers and anthropogenic factors on ground-level Ozone is photochemically produced through non-linear ozone pollution. This is becoming more important for urban chemical processes mainly involving reactions of nitrogen cities in the Mediterranean area with high summer tempera- oxides (NOx = NO2 + NO) and volatile organic compounds tures, and heat waves are projected to become more severe (VOCs); this results from both anthropogenic and biogenic in the future due to anthropogenic climate change (Pyrgou sources in the presence of sunlight (Monks et al., 2015; et al., 2018; Zittis et al., 2015). Crutzen, 1974; Derwent et al., 1996). This chemistry oc- Indeed, local ozone levels depend not only on local pro- curs in two photochemical regimes, NOx -sensitive and VOC- duction and loss mechanisms that are sensitive to meteoro- sensitive (Sillman et al., 1990; Sillman, 1999). In the NOx - logical factors, but also on the transport of ozone and its sensitive regime (low NOx and high VOCs), ozone produc- precursors. Previous studies have shown that ambient ozone tion is controlled (or limited) by the concentration of NOx . concentrations are strongly influenced by transport of re- Therefore, the O3 levels increase with increasing NOx , and gional ozone and its precursors, while local precursor emis- there are only small changes with increases in the VOC sions play limited roles in ozone formation (Romero-Alvarez levels. In this regime, O3 reacts mainly with hydrogenated et al., 2022; Kleanthous et al., 2014). Cristofanelli and Bona- species to form hydrogen peroxide (H2 O2 ), which is re- soni (2009) showed that the background tropospheric ozone moved by wet and dry deposition. In the VOC-sensitive concentration in the Mediterranean area and southern Europe regime (high NOx ), ozone levels increase with increased is affected mainly by three transport processes: (1) regional VOC concentrations and decrease with increased NOx con- and long-range transport of pollutants, (2) downwards trans- centrations. This last regime is typical of urban areas, where port from the stratosphere, and (3) transport of dust from the a reduction in NOx emissions enhances ozone levels locally Sahara. due to the higher levels of oxidants (mainly hydroxyl rad- The lockdown period provided a significant reduction in icals, OH) and reactions with VOCs. Peak concentrations ozone precursors, and it represents an excellent opportunity of ozone usually occur during the midday hours, when the to further our understanding of the photochemical reactions sunlight is most intense. However, during the afternoon and involved in ozone chemistry. Estimated average emission re- evening, high ozone concentrations are observed in remote ductions in Spain during the most severe lockdown period areas due to higher biogenic VOC emissions, less titration by were reported, with road and air traffic reductions reaching NO, and transport of O3 and its precursors from their sources. 80 %–90 % (Guevara et al., 2021). During this period, Gue- At night and next to a source with high emissions of NO (e.g. vara et al. (2021) estimated the average emission reductions power plants), ozone is lost through the process of NOx titra- at the EU-30 level to be −33 % for NOx . Consequently, re- tion and forms NO2 , which is subsequently converted to ni- strictive mobility measures that included important reduc- tric acid (HNO3 ) and removed from the atmosphere by wet tions in traffic had many positive environmental impacts, and dry deposition (Monks et al., 2015). In summary, O3 lev- and improvements in air quality were reported globally (Liu els can be reduced only if there are reductions in the amounts et al., 2020; Venter et al., 2020; Sharma et al., 2020). of the precursors NOx , VOCs, and carbon monoxide (CO). Previous modelling studies analysed the changes in air Reductions in VOC emissions would be an effective path- quality, emissions, and chemical regimes seen on global and way to reducing ozone in a high-NOx area (VOC-sensitive). regional scales during the COVID-19 lockdown (Miyazaki On the other hand, reductions in NOx emissions would be ef- et al., 2021; von Schneidemesser et al., 2021; Roozitalab Atmos. Chem. Phys., 23, 10751–10774, 2023 https://doi.org/10.5194/acp-23-10751-2023 A. Badia et al.: Modelling the impacts of emission changes on O3 chemistry 10753 et al., 2022; Sicard et al., 2020; Badia et al., 2021b). A O3 averages (−0.4 % yr−1 ) at rural sites and an increase at global modelling study by Miyazaki et al. (2021) showed urban and suburban stations (0.6 % and 0.4 %, respectively) that the global total tropospheric ozone burden declined by during 2000–2010 for the Mediterranean area. These changes 6 Tg (∼ 2 %) during May–June 2020, mainly due to emission resulted from the mitigation policies for NOx and VOC emis- reductions in Asia and North and South America. The mod- sions in the EU that led to an increase in O3 levels in urban elling study of Venter et al. (2020) found that, after taking areas due to a reduction in titration by NO. Therefore, there is into account the meteorological variations, lockdown mea- an urgent need to expand our knowledge of ozone chemistry sures reduced the levels of NO2 (mainly due to the reduction to help decision-makers choose better mitigation strategies. in transportation) and the levels of particulate matter (PM) In the present study, we use the Weather Research and by approximately 60 % and 31 %, respectively, in 34 coun- Forecasting model with Chemistry (WRF-Chem) coupled tries, with a general increase in O3 of 4 % (−2 % to 10 %). with the urban canopy building effect parameterization and Sicard et al. (2020) described ozone increases in cities (17 % building energy model (BEP + BEM) to investigate the re- in Europe, 36 % in Wuhan) resulting from lower titration of sponse of NOx –VOC–O3 chemistry to changes in precursor O3 by NO due to the strong reduction in NOx emissions emissions in the metropolitan area of Barcelona (AMB); this from road transport. Most of this literature was focused only furthers our understanding of ozone formation mechanisms on the lockdown period, and the de-escalation period, which and transport to rural areas, and it enables the design of effec- had different ozone chemistry, was not analysed. In addition, tive air quality mitigation strategies. We compare the ozone these studies did not discuss changes in the chemical regimes levels for two periods with very different anthropogenic and arising from different land uses and the transport of pollu- biogenic activity levels, i.e. (1) March and April 2020, when tants due to lockdown measures from cities to rural areas. ozone precursors were at their lowest levels because of re- Only a few studies have reported enhanced atmospheric peated lockdowns due to COVID, and (2) May 2020, when oxidation capacity (AOC), which describes the removal rate the AMB was in a relaxation plan but the ozone levels were of primary pollutants and the formation of secondary species typically highest because of the bloom of biogenic emissions together with the associated O3 increases during the COVID- under the intense sunlight conditions. In addition, ozone pro- 19 lockdown due to increases in the major oxidants OH, duction regimes for different land uses of the AMB are anal- hydroperoxy radical (HO2 ), and nitrate radical (NO3 ) (Zhu ysed to determine changes in the NOx/VOC ratio in differ- et al., 2021; Wang et al., 2021, 2022). The predominant oxi- ent areas of the city that are affected by different biogeo- dant for AOC during the daytime is OH, since NO3 radicals chemical effects (biogenic VOC emissions, dry deposition, photolyse rapidly during the day, which is responsible for anthropogenic emissions of ozone precursors). Here, we also the oxidation and removal of most natural and anthropogenic discuss changes in the AOC oxidants and propose that the trace gases (Elshorbany et al., 2009; Saiz-Lopez et al., 2017). AOC should be considered when designing air quality con- On the other hand, during the night, the concentration of OH trol policies. Changes in ozone circulation from the AMB to is significantly reduced, and the AOC is then controlled by the Pyrenees are also discussed for specific days character- NO3 together with O3 , which is also an important oxidant ized by high ozone levels. (Elshorbany et al., 2009; Saiz-Lopez et al., 2017). During The case study is described in Sect. 2. The air quality the lockdown, the significant decreases in NO2 concentra- model, including the model set-up, validation, and simula- tions increased the OH levels, which led to the formation of tions, is described in Sect. 3, followed by the results (Sect. 4). harmful oxidants such as O3 (Zhu et al., 2021; Wang et al., A discussion of the ozone chemistry that includes ozone sen- 2021, 2022). Therefore, the AOC is an indicator used to de- sitivity, AOC, and transport from the city to rural areas is sign control policies for secondary species. presented in Sect. 5. In recent decades, EU emission mitigation policies have been successful in decreasing emissions of key air pollutants such as SO2 , NOx , non-methane VOCs – NMVOCs, and PM (Sicard et al., 2021; Agency et al., 2019). However, the cur- 2 Case study rent levels for the secondary air pollutant O3 in cities con- tinue to exceed the EU standards and the WHO air qual- The AMB serves as our case study. It is located in Catalo- ity guidelines (Guerreiro et al., 2014; Agency et al., 2019). nia (Spain) in the north-eastern part of the Iberian Penin- Indeed, local ozone pollution mitigation efforts are gener- sula (Fig. 1). This region is characterized by a Mediterranean ally inefficient, mainly because (1) ozone formation depends climate, with dry and hot summers and clear skies. Due to on non-linear chemical interactions; (2) with a lifetime of a complex orographic territory with different altitudes and several weeks, ozone levels are strongly influenced by long- several peripheral mountain ranges and depressions, it is not distance transport, which is associated with specific weather easy to generalize the climatic features of the Catalan lands conditions and the hemispheric background; and (3) its pre- for the whole territory (Martín-Vide et al., 2010). The AMB, cursors are emitted mostly far from the sites of ozone ex- with more than 3 million people, is the most populated urban ceedances. Sicard et al. (2013) found a decrease in annual area on the Mediterranean coast. https://doi.org/10.5194/acp-23-10751-2023 Atmos. Chem. Phys., 23, 10751–10774, 2023 10754 A. Badia et al.: Modelling the impacts of emission changes on O3 chemistry Figure 1. (a) Location and (b) main topographic features of the study area. Base maps in panel (a) were taken from © Google Earth. The locations of air pollution monitoring stations (Xarxa de Vigilància i Previsió de la Contaminació Atmosfèrica, XVPCA) along the S–N axis (Barcelona–Vic Plain–Pyrenean range) are shown in panel (a) (right). The city of Barcelona annually reports some of the high- ceedances for hourly O3 levels (180 µg m−3 ) when the sea est air pollution levels in Europe, and the most problematic breeze transports the ozone precursors from the AMB inland pollutants are NO2 , PM2.5 , and PM10 (Rivas et al., 2014). In to this rural plain (Querol et al., 2017; Massagué et al., 2019; particular, in 2019, the NO2 annual mean levels in the high- Jaén et al., 2021). During the late spring and summer seasons, traffic urban air pollution ground monitoring stations (Eix- the combination of daily upslope winds and sea breezes may ample and Gràcia-Sant Gervasi) exceeded the WHO guide- cause the intrusion of polluted air masses up to 160 km in- line (40 µg m−3 ) (Rico et al., 2019). In the same year, the land. Thus, the air masses from a polluted area (such as the mean values for PM2.5 and PM10 exceeded the WHO guide- AMB) can be transported northwards and injected at high al- line (20 and 10 µg m−3 , respectively) at all urban stations titudes (2000–3000 m a.s.l.) by the Pyrenean mountain range in the city (Rico et al., 2019). Exceeding these air quality (Querol et al., 2017; Massagué et al., 2019). reference levels is associated with significant risks to public Ozone levels were higher compared to previous years in health (World Health Organization, 2021; Rivas et al., 2014). Ciutadella, the background station in Barcelona, during the The year 2020 was the first year in which the NO2 values period March–June 2020 (see Fig. S2). In Tona (a rural sta- in Barcelona remained within the WHO limits (Rico et al., tion located in the Vic Plain, situated 45–70 km north of 2020) due to the significant reduction in traffic emissions Barcelona, and surrounded by high mountains) and in Par- that resulted from the Spanish government’s emergency rule dines (also a rural site situated in the Pyrenees; see Fig. 1), and its lockdown restrictions (see Fig. S1 in the Supplement). high ozone concentrations were registered and clearly ex- Note that 2020 had more rainfall than previous years, which ceeded the WHO limit of 60 µg m−3 for peak seasons. Ozone has important implications for the removal of pollutants from levels were high in these two rural stations during March– the air. June, with higher values for 2015–2019 compared to 2020 Another air quality problem was found for the Vic Plain due to the reduced emissions in the AMB. Note that, for some (see Fig. 1), which records the highest number of ex- weeks, the O3 levels were higher in 2020 than in 2015–2019 Atmos. Chem. Phys., 23, 10751–10774, 2023 https://doi.org/10.5194/acp-23-10751-2023 A. Badia et al.: Modelling the impacts of emission changes on O3 chemistry 10755 (turning from north to south). This was caused by a strong anticyclone located in northern Europe (over 1030 hPa) that generated an undetermined situation with no effects of high or low pressures in the peninsula. See Table 1 for a summary of the meteorological conditions for these periods. In addition, we select two days in the lockdown period (3 and 6 April) and two days in the relaxation period (22 and 26 May) on which high ozone concentrations were reg- istered (see Table S1 in the Supplement), and there is a clear influence of the air masses from the AMB on rural areas far from the city (discussed in Sect. 5.3) for studying the changes in the O3 circulation from Barcelona (Ciutadella) to the Pyrenees (Pardines), including the Vic Plain (Tona) and Montseny. The mean surface temperatures, pressures at the surface level, and accumulated precipitation among other meteorological variables for these days are presented for the Figure 2. Model domains. D1: Iberian Peninsula 9 km × 9 km; D2: four sites (Ciutadella, Montseny, Tona, and Pardines) in Ta- Catalonia horizontal resolution 3 km × 3 km. ble S2. According to the meteorological data from the Servei Meteorològic de Catalunya (2020), there was no precipita- tion for the days selected except in the Pyrenees (2.9 mm) on due to meteorological conditions that increased the levels of 22 May, the wind intensity was low, and the surface temper- ozone precursors (see Figs. S1 and S2). atures were significantly high for this time of the year during In this study, we selected two periods to discuss the the two days in May. changes in O3 chemistry: (1) a full lockdown period (30 March–12 April, weeks 3–4 in Table 1), in which we 3 Air quality model found the highest mobility reduction; and (2) a relaxation period (18–30 May, weeks 10–11 in Table 1), when restric- We used the regional chemistry transport model WRF-Chem tions started to relax and O3 formation increased due to the (Grell et al., 2005) version 4.1, a highly flexible community warm temperatures of the late spring (see Table 1). The first model for atmospheric research in which aerosol–radiation– period (30 March–12 April) was characterized by meteoro- cloud feedback processes are considered. The WRF-Chem logical dynamism in the Iberian Peninsula. The period be- model is widely used for simulations of air pollution episodes gan with a high-pressure centre in the North Atlantic Ocean (Georgiou et al., 2018; Yegorova et al., 2011) and, in particu- that generated a cold and dry continental north-eastern flux lar, the air quality over the AMB has been analysed in Badia over Europe. General precipitation was registered in Catalo- et al. (2021b). nia during the first days (30 March–4 April), caused by a low- pressure centre that moved from the Atlantic to the Mediter- 3.1 Model set-up ranean. These conditions were optimal for generating precip- itation in the Mediterranean area of the peninsula. The next The WRF-Chem model is configured with two domains days were defined by high pressures over Europe, which gen- covering the Iberian Peninsula (D1: 9 km × 9 km) and the erated a more stable meteorology with isolated precipitation Catalonia region (D2: 3 km × 3 km) with 45 vertical layers and a warm air mass crossing Europe from the south. During up to 100 hPa (Fig. 2). The meteorological and chemical this period, no important fluxes from any directions were de- initial and lateral boundary conditions (IC and BCs) were tected. At the surface level, the Azores anticyclone became determined using the ERA5 global model data (Hersbach stronger and reached pressures above 1030 hPa over the At- et al., 2020) and the Whole Atmosphere Community Climate lantic. The second period (18–30 May) had more stability Model (WACCM) (Gettelman et al., 2019), respectively. The due to a strong anti-cyclonic ridge covering the south-west HERMESv3 pre-processor tool (Guevara et al., 2019) was and centre of Europe, resulting in typical summer weather. used to create the anthropogenic emission files from the Winds from the north and north-west were detected north- CAMS-REG-APv3.1 database (Granier et al., 2019). This east of Catalonia and in the Ebro Valley, respectively, due to emission inventory is based on data from 2016. Biogenic the Pyrenees’ natural barrier, which modifies the trajectories emissions are computed online from the Model of Emissions of superficial winds. Low pressures were found over Italy, of Gases and Aerosols from Nature v2 (MEGAN; Guenther which intensified the winds from the north over the east of et al., 2012). For the gas-phase chemical scheme, we used the peninsula. The last days of this period (24–30 May) reg- the Regional Acid Deposition Model (RADM2, Stockwell istered weak precipitation in the Pyrenees and the area north- et al., 1990), which accounts for 63 chemical species, 21 east of Catalonia, with weak winds from different directions photolytic reactions, and 136 gas-phase reactions. NMVOC https://doi.org/10.5194/acp-23-10751-2023 Atmos. Chem. Phys., 23, 10751–10774, 2023 10756 A. Badia et al.: Modelling the impacts of emission changes on O3 chemistry Table 1. Meteorological conditions prevailing during the COVID-19 mobility restrictions (2020). The meteorological data were extracted from the following source: Servei Meteorològic de Catalunya (SMC). The circulation weather types (CWTs) were calculated by using the surface pressures of the NCEP and NCAR reanalysis dataset. CWTs: cyclonic 56 (C), anti-cyclonic (A), pure advectives (N, S, E, W, NE, SE, NW, SW), hybrid cyclone advectives (CN, CNE, CE, CSE, CS, CSW, CW, CNW), and hybrid anticyclone advectives (AN, ANE, AE, ASE, AS, ASW, AW, ANW). The mobility reduction over the AMB is provided in the fourth column (source: Google mobility reports (https://www.google.com/covid19/mobility/, last access: 1 February 2023)). F0, F1, F2, and F3 are the different phases of the de-escalation period, F0 being the first phase after lockdown and F3 being the last phase before all restrictions were eliminated. Psfc denotes surface pressure. Precipitation Mobility accumulated Week Dates (2020) Stage reduction Psfc (hPa) T (◦ C) (mm) CWTs 1 16–22 Mar Lockdown 75.4 % 1020.4 12.9 24.1 SE 2 23–29 Mar Lockdown 82.2 % 1015.8 11.4 20.4 E 3 30 Mar–5 Apr Full lockdown 84.4 % 1015.9 11.6 25.3 C 4 6–12 Apr Full lockdown 84.4 % 1023.0 15.0 0 A 5 13–19 Apr Lockdown 82.6 % 1014.5 15.1 109.1 U 6 20–26 Apr Lockdown 79,8 % 1011.1 15.7 126.7 C 7 27 Apr–3 May Lockdown 74 % 1013.8 18.5 2.3 N 8 4–10 May F0 60 % 1014.0 19.3 6.9 U 9 11–17 May F0 58,2 % 1010.0 17.6 13.7 C 10 18–24 May F0–F1 50,6 % 1020.4 21.8 0.0 U 11 25–30 May F1 43,2 % 1021.0 20.8 4.6 E 12 1–7 Jun F1–F2 41,4 % 1009.5 19.8 21.9 U 13 8–14 Jun F2 39 % 1012.0 19.2 24.5 C 14 15–20 Jun F2–F3 34,8 % 1018.1 20.5 16.4 U oxidation in RADM2 only explicitly treats ethane, ethene, urban canopy scheme to compute the heat and momentum and isoprene species, and all other NMVOCs are classified fluxes in the urban areas (see Segura et al., 2021, for more as grouped species based on OH reactivity and molecular details on the use of the LCZ and urban morphology). We weight. Thus, the RADM2 gas-phase chemical mechanism performed a spin-up of 1 month. Table 2 describes the main grouped the VOCs into 14 species, such as alkane, alkene, configuration of the model. aromatic, and formaldehyde. In WRF-Chem, RADM2 is coupled to the Modal Aerosol Dynamics Model for Europe (MADE)/Secondary Organic Aerosol Model (SORGAM) (Ackermann et al., 1998; Schell et al., 2001). RADM2 has 3.2 Description of the simulation cases been broadly used in studies of air quality over Europe (Im et al., 2015; Tuccella et al., 2011; Badia et al., 2021b), and To better understand the impacts of emission reduction mea- its model biases for NO2 and O3 are in line with other air sures on air quality, the WRF-Chem model was utilized to quality modelling studies over Europe (Im et al., 2015; Mar calculate the changes in O3 chemistry during the COVID et al., 2016). In particular, the RADM2 chemical mechanism lockdown period. We ran two simulations: (1) Business As has been used in Badia et al. (2021b) over the AMB. Usual (BAU) and (2) COVID for the period of March– Here, we used a multi-layer urban canopy scheme, June 2020 (see Table 1). The COVID run used the emis- BEP + BEM (Salamanca et al., 2011)), to represent the urban sion changes (see Fig. 3) provided by the Barcelona Super- areas in our domain; this takes into account the energy con- computing Center (Guevara et al., 2021), which were previ- sumed by buildings and the anthropogenic heat, which has ously used in other studies (von Schneidemesser et al., 2021; previously been validated for the area under study (Ribeiro Brancher, 2021). These emission changes varied per day, et al., 2021; Segura et al., 2021). The local climate zone country, and sector. Figure 3 displays the emission changes (LCZ) classification (Stewart and Oke, 2012) is used for the used in this study, and the highest changes were found for AMB, which associates specific values of the thermal, radia- the road transport (up to 80 %) and aviation (up to 90 %) sec- tive, and geometric parameters of the buildings and ground tors. Inputs for the other emissions (biogenic, dust, sea spray) with 11 urban classes, which are used by the BEP + BEM and meteorology used in WRF-Chem were set to be consis- tent. As a result, the differences in pollutant concentrations Atmos. Chem. Phys., 23, 10751–10774, 2023 https://doi.org/10.5194/acp-23-10751-2023 A. Badia et al.: Modelling the impacts of emission changes on O3 chemistry 10757 Table 2. Model details and experiment configuration. Chemistry Chemical mechanism RADM2 (Stockwell et al., 1990) Aerosol scheme MADE/SORGAM aerosol scheme (Ackermann et al., 1998; Schell et al., 2001) Photolysis scheme Fast-J (Wild et al., 2000) Dry deposition Wesely (2007) Wet deposition Grell and Dévényi (2002) Anthropogenic emissions CAMS-REG-APv3.1 database (Granier et al., 2019) Biogenic emissions MEGAN (Guenther et al., 2012) Physics Urban canopy scheme BEP + BEM (Salamanca et al., 2011) PBL scheme BouLac (Bougeault and Lacarrere, 1989) Resolution and initial conditions Horizontal resolution D1: 9 km × 9 km, D2: 3 km × 3 km Vertical layers 45 Top of the atmosphere 100 hPa Chemical initial condition WACCM (Gettelman et al., 2019) Meteorological initial condition ERA5 (Hersbach et al., 2020) Chemistry spin-up 1 month calculated by WRF-Chem were attributed to changes in the and an RMSE of 1.4 ◦ C. The rural stations presented a higher anthropogenic emissions. bias inside the AMB (0.9 ◦ C), which resulted from erroneous descriptions of the land use at the model resolution level 3.3 Model validation (3 km). Loose performance was found for the relative hu- midity and the wind speed, with average RMSEs of 12.2 % Several meteorological and air quality stations were used and 2.5 m s−1 , respectively. The model underestimated the herein to evaluate the model (COVID simulation) for the relative humidity in the urban and rural areas of the AMB lockdown (30 March to 12 April 2020) and relaxation (18 by 2.0 % and 5.2 %, respectively, while it overestimated the to 30 May 2020) periods. The same model configuration has humidity in the rural areas of the CAT (1.4 %). In the case been evaluated previously over the AMB for the meteorol- of the wind speed, the model overestimated the wind flow ogy (Ribeiro et al., 2021; Segura et al., 2021) and the chem- over the entire domain (1.5 m s−1 on average in CAT), espe- istry, without any reduction in anthropogenic emissions (Ba- cially in the rural areas of CAT (1.9 m s−1 ). For the second dia et al., 2021b), for different periods. study period, a similar performance was obtained for the air temperatures inside the AMB, with a slight increase in the 3.3.1 Meteorology RMSEs (1.5 and 1.6 ◦ C for the urban and rural areas) and a decrease in the correlations between modelled and observed The meteorological data used to validate our model were data (0.90 and 0.92, respectively). Unlike the first period, the from the Xarxa d’Estacions Meteorològiques Automàtiques model overestimated the relative humidities at all the stations (XEMA). Stations within this network are classified as ur- in the second period, except for the rural stations in the AMB ban and rural according to the land use of the model. Herein, (−3.8 %). The model provided lower overestimates for the we used data for the wind speed (WS), temperature (T ), and wind speeds during the second period (1.0 m s−1 on average relative humidity (RH). Tables S3 and S4 in the Supplement in CAT), although the correlations decreased in the second present statistical evaluations of hourly data for the AMB and period. the Catalonia (CAT) region for the lockdown (30 March to 12 April) and relaxation (18 to 30 May) periods, respectively. The validations of the WRF-Chem simulation (COVID 3.3.2 Air quality run) revealed that the model generally reproduced the air temperatures of the two simulation periods well, but the per- Air quality data from the Xarxa de Vigilància i Previsió de formance was poorer in representing the relative humidity la Contaminació Atmosfèrica (XVPCA) monitoring stations and the wind speed. For the first period, the simulated air were used here. Stations in this network were classified into temperatures for the urban stations showed low positive bi- different groups: urban background, urban traffic, suburban ases of 0.4 and 0.3 ◦ C for the AMB and CAT, respectively, background, and rural. Here, we used the data for O3 and https://doi.org/10.5194/acp-23-10751-2023 Atmos. Chem. Phys., 23, 10751–10774, 2023 10758 A. Badia et al.: Modelling the impacts of emission changes on O3 chemistry tions in NOx emissions, e.g. from road transport. Low values for the modelled NOx levels underestimated ozone loss via NO titration, which resulted in high nighttime surface ozone concentrations. The lifetime of surface NOx (a few hours) is shorter than that of O3 (days or weeks); thus, the surface NOx concentrations are very sensitive to emissions. We should also mention that there might be large uncertainties for the calculations of emission factors, as discussed in Doumbia et al. (2021). Underestimates of traffic NOx emissions over Europe have been mentioned previously in several air qual- ity modelling studies (von Schneidemesser et al., 2021; Karl et al., 2017). Although the model exhibits biases for NO2 and O3 , the modelled air quality changes presented in the next section are in line with other studies such as Querol et al. (2021) that present a comparison between data from years 2015–2019 and the lockdown and relaxation periods for the year 2020 in the city of Barcelona. 4 Results 4.1 Air quality changes During the first period (30 March to 12 April) we see a gen- eral reduction in NO2 concentrations of the COVID simula- Figure 3. Emission reduction percentage (%) for each sector. Note tion with respect to the BAU all over the Catalonia region that the other stationary combustion sector has a different reduction at the surface level, with high reductions found during the level for each pollutant. The periods analysed here are written at the bottom of the figure. evening peaks (19:00–21:00 UTC) and over the AMB (−2 to −18 µg m−3 , −10 % to −70 %) (see Fig. S3). The highest reductions were found around the airport due to a reduction in air traffic emissions (see Fig. 3). The surface concentra- NO2 . Tables S5 and S6 in the Supplement present statisti- tions of VOCs were slightly lower during the morning peak, cal evaluations of hourly data for the AMB and CAT during with reductions up to −2 µg m−3 (−10 %), as can be seen the lockdown (30 March to 12 April) and relaxation (18 to in Fig. S4. Similarly, there was also a reduction during the 30 May) periods, respectively. The modelled concentrations evening peak of up to −1.5 µg m−3 (−12 %). Note that, dur- were converted to unit micrograms per cubic metre by using ing the lockdown, the VOC emissions increased up to 8 % the temperatures and pressures from the model. in the stationary combustion sector (see Fig. 3). Changes in Overall, the model (COVID simulation) showed reason- emissions that showed a significant decrease in NO2 con- able agreement with the observations for NO2 and O3 con- centrations and a slight decrease in VOC concentrations en- centrations during both periods. The best performance in the hanced O3 levels over the AMB. This is consistent with the lockdown period was observed over the urban background observations, where there was a decrease in NO2 concen- (R between 0.43 and 0.45 for NO2 and between 0.70 and trations (40 %–80 %) and an increase in O3 levels (up to 0.73 for O3 ), while low R values were found over the rural 10 %) between 2020 and 2015–2019 during the lockdown areas (0.32 for NO2 and 0.42 for O3 ). The performance for (see Figs. S1–S2). The reduction in O3 concentrations that the relaxation period was not as good as that for the lock- normally results from lower levels of precursors was can- down period, with R values between 0.24 and 0.40 for NO2 celled out by a reduction in NO titration, resulting in a net and between 0.42 and 0.62 for O3 . However, there were neg- increase in O3 levels. During the evening peaks (19:00– ative and positive biases in both periods for NO2 (normal- 21:00 UTC), we found the highest increases in O3 of the ized mean bias, NMB, between −0.15 and −0.66) and O3 COVID simulation compared to the BAU (1 to 18 µg m−3 , (NMB between 0.13 and 0.28), respectively. Similar biases 1 % to 20 %). However, when surface O3 concentrations typ- were seen in another study (von Schneidemesser et al., 2021). ically reach their peak in the middle to late afternoon (13:00– Part of our model bias was attributed to the (1) boundary 15:00 UTC), the increases in O3 levels were much lower conditions used for this study (WACCM model) that added a (up to 6 µg m−3 , 6 %) than those for the evening peak (see bias to the O3 background levels (Giordano et al., 2015) and Fig. S5). Outside the AMB, the concentrations did not dif- (2) the current emission inventory being too coarse to accu- fer significantly (< 2 µg m−3 , < 2 %) between the BAU and rately represent the spatial distributions and temporal varia- COVID simulations. Differences in the Ox (NO2 + O3 ) val- Atmos. Chem. Phys., 23, 10751–10774, 2023 https://doi.org/10.5194/acp-23-10751-2023 A. Badia et al.: Modelling the impacts of emission changes on O3 chemistry 10759 Table 3. Average NOx / VOC ratio and ozone concentrations from 30 March to 12 April (only weekdays) and from 18 to 30 May (only weekdays) in the morning (06:00–08:00 UTC), afternoon (13:00–15:00 UTC), and evening (19:00–21:00 UTC). Italic and bold italic font indicate VOC and NOx regimes, respectively. The relative changes in ozone concentrations (%) are shown in parentheses and were calculated as ((COVID − BAU) / BAU)×100. March–April Land use Morning Afternoon Evening BAU COVID BAU COVID BAU COVID [NOx / VOC] Forest 0.102 0.083 0.073 0.059 0.143 0.111 Natural open 0.158 0.098 0.109 0.077 0.194 0.122 Agriculture 0.156 0.100 0.113 0.077 0.195 0.101 Water 0.25 0.201 0.495 0.393 0.707 0.535 Compact urban 0.166 0.139 0.171 0.137 0.236 0.184 Open urban 0.125 0.088 0.108 0.081 0.182 0.135 Industrial 0.149 0.120 0.135 0.100 0.215 0.151 O3 (ppb) Forest 40.3 40.7 (1.0 %) 51.7 51.7 (0.0 %) 42.3 42.9 (1.3 %) Natural open 42.8 43.9 (2.5 %) 51.7 51.9 (0.3 %) 45.4 46.4 (2.1 %) Agriculture 37.3 38.5 (3.4 %) 51 51.2 (0.4 %) 40.7 42.2 (3.7 %) Water 37 38.7 (4.6 %) 48.6 49.4 (1.6 %) 42.1 44.2 (5.0 %) Compact urban 35.8 36.9 (2.9 %) 52 52.3 (0.6 %) 40.4 42.2 (4.3 %) Open urban 39.4 40.4 (2.4 %) 52.6 52.8 (0.3 %) 42.6 43.6 (2.4 %) Industrial 36.8 37.7 (2.3 %) 52.2 52.5 (0.5 %) 40.1 41.8 (4.2 %) May Land use Morning Afternoon Evening BAU COVID BAU COVID BAU COVID [NOx / VOC] Forest 0.076 0.066 0.053 0.043 0.139 0.115 Natural open 0.122 0.087 0.078 0.052 0.194 0.129 Agriculture 0.122 0.090 0.077 0.046 0.192 0.118 Water 0.349 0.279 0.792 0.663 1.327 1.028 Compact urban 0.137 0.121 0.156 0.125 0.261 0.210 Open urban 0.101 0.078 0.099 0.072 0.191 0.150 Industrial 0.125 0.104 0.107 0.075 0.230 0.165 O3 (ppb) Forest 36.9 36.7 (−0.6 %) 51.3 50.5 (−1.6 %) 37.7 37.6 (−0.1 %) Natural open 37.0 37.3 (0.6 %) 48.9 48.3 (−1.3 %) 40.3 40.6 (0.8 %) Agriculture 33.3 33.8 (1.4 %) 50.5 49.9 (−1.2 %) 36.3 37.3 (3.0 %) Water 29.7 31.2 (4.9 %) 42.5 43.4 (2.2 %) 32.0 35 (9.4 %) Compact urban 33.6 33.9 (1.0 %) 50.8 50.4 (−0.8 %) 35.8 37.2 (3.8 %) Open urban 36.7 36.9 (0.5 %) 51.3 50.7 (−1.1 %) 38.2 38.6 (1.1 %) Industrial 34.3 34.6 (0.9 %) 52.1 51.6 (−1.1 %) 36.3 37.6 (3.5 %) ues were calculated to aid our interpretation of the O3 con- tions in the NO2 (−2 to −15 µg m−3 , −10 % to −65 %) and centrations by diminishing the effect of O3 titration by NO VOC (up to −2 µg m−3 , −16 %) levels, with high reductions in highly polluted areas (see Fig. S6). The overall changes found during the evening peaks (see Figs. S3 and S4). Ozone between the BAU and COVID in the Ox concentrations re- levels decreased (by up to 3.5 µg m−3 ; see Fig. S5) in most of mained practically constant due to a balance between the in- Catalonia due to significant reductions in most of the emis- creases in O3 levels and the decreases in NO2 levels. This sion sectors (see Fig. 3) during the COVID simulation, which has important policy implications because one air pollutant decreased the high ozone productivity normally seen for this problem is being replaced by another, which is an undesir- time of the year. However, we still found enhanced O3 levels able consequence due to the ground-level ozone effects on around the Barcelona airport in the evenings; the reductions human health, vegetation, and ecosystems. A similar result in emission levels were still significant (more than 80 %; see was seen by Grange et al. (2021), who found that Ox con- Fig. 3) and inhibited titration of the O3 by NO. Note that, centrations only changed very slightly due to the lockdowns in this case, the Ox concentrations decreased nearly every- across most European urban areas. where in the Catalonia area and up to −4 µg m−3 over the The differences between the BAU and COVID simulations AMB (see Fig. S6) for the COVID simulation, resulting in for the second period (18 to 30 May) showed overall reduc- overall improvements in the air quality. https://doi.org/10.5194/acp-23-10751-2023 Atmos. Chem. Phys., 23, 10751–10774, 2023 10760 A. Badia et al.: Modelling the impacts of emission changes on O3 chemistry Figure 4. Modelled O3 concentrations (top panels) for 30 March to 12 April (only weekdays) and for 18 to 30 May (only weekdays) for both simulations, BAU (left panels) and COVID (right panels), over the AMB area during the morning (06:00–08:00 UTC). Each dot of the top row corresponds to the O3 concentration difference (ppb) of one grid cell of the AMB at the surface level. The dots in the lower row represent the land use for each model grid cell. Figure 5. Same as Fig. 4 during the afternoon (13:00–15:00 UTC). 5 Discussion this complex relationship in Figs. 4 (06:00 to 08:00 UTC), 5 (13:00 to 15:00 UTC), and 6 (19:00 to 21:00 UTC), which 5.1 O3 sensitivity to precursors and land use show the differences in surface NOx , VOC, and O3 concen- trations in the BAU and COVID simulations during the first Variations in the levels of the O3 precursors (NOx and VOCs) period (30 March to 12 April, only weekdays) and the second had large effects on O3 production. In addition, land use is period (18 to 30 May, only weekdays). To complement this the key to understanding how industrial, open urban, com- information, we calculated the average NOx /VOC ratios for pact urban, water, agriculture, natural open, and forest land these two periods in Table 3. Values of NOx and VOC con- uses influenced the O3 regimes (see Fig. S12 and Table S9 centrations and relative changes for each land use are shown for more details on the land use classification). We represent in Tables S7 and S8. In this study, we establish a relationship Atmos. Chem. Phys., 23, 10751–10774, 2023 https://doi.org/10.5194/acp-23-10751-2023 A. Badia et al.: Modelling the impacts of emission changes on O3 chemistry 10761 Figure 6. Same as Fig. 4 during the evening (19:00–21:00 UTC). between surface VOC : NOx and O3 concentrations and, sub- while in the afternoon, these grids are in the transition or sequently, derive the line separating the two different photo- NOx -sensitive regimes. Areas close to highly polluted areas chemical regimes by the local O3 maximum (see Fig. S7). (including urban areas such as compact urban, industrial, and The local O3 maximum occurs when VOC : NOx ≈ 8 coin- water) are in VOC-sensitive or transitional regimes all day, cides with the ratio defined in National Research Council especially during the evening due to high traffic emissions. (1991). In Figs. 4–6, we indicate the NOx -limited regime Note that “water points” are located around the harbour and with a dark solid line separating VOC : NOx > 8, which is the airport and are typically high NOx sources. In terms of typical for locations located downwind of urban and sub- ozone levels (BAU simulation), high values during the morn- urban areas, and the VOC-limited regime (VOC : NOx < 8), ing (40–43 ppb) and evening (46–47 ppb) hours are found in which is typical for highly polluted urban areas (National suburban areas (forest and natural open) because there is less Research Council, 1991). We also indicate the transitional NO (because of less traffic) and thus less ozone degradation. regime with two dotted lines (VOC : NOx > 4/1 and VOC : In the afternoon, high O3 levels are found everywhere in the NOx < 15/1) showing where ozone becomes less sensitive to AMB (49–55 ppb), especially in urban areas (industrial, open NOx changes and increases with increasing VOC levels, as urban, and compact urban). identified in National Research Council (1991). With the significant decrease in NOx (20 %–40 %) and a Overall, without any reduction in emissions (BAU simu- slight decrease in VOC (1 %–10 %) levels during the lock- lation), this analysis indicates that in urban forests far from down (see Sect. 4.1), the averaged O3 levels by land use in- anthropogenic sources and influenced by high biogenic VOC creased for the COVID run (1 %–5 %), and chemical forma- emissions, the photochemical regime of O3 formation is tion tends to enter the NOx -sensitive regime, especially in NOx -sensitive in the mornings and afternoons. With the in- the morning and afternoon hours during April–March. In the land sea breeze fronts seen in the late afternoons and early case of green areas (forest, natural open, and agriculture), we evenings (Massagué et al., 2019), pollutants are transported see clear transitions towards a NOx -sensitivity regime. How- from their sources (urban areas) to other areas (rural areas). ever, despite the cuts in emissions, most of the grids close to Consequently, we found a transition to a VOC-limited regime highly polluted areas (compact urban, industrial, and water) in green areas (forest, natural open, and agriculture) in the were still in the VOC-sensitive or transitional regimes all day, evenings. Some of the grid points classified as naturally open especially in the evenings (high traffic emissions), for which and agriculture are close to the Barcelona airport (high NOx we found the highest ozone increases (2.4 %–5 %). Similar sources), and the transport of pollutants (driven by the wind results were found when we compared both runs (BAU and speed and direction) has a significant impact on their regime. COVID) for the period in May in terms of the changes in the However, we see that most of the grids classified as naturally chemicals for each land use. However, we found that, during open are NOx -limited all day. In the case of grids classified May, the maximum ozone levels decreased in the COVID as agriculture, we see that, in the morning and evening, most run during the afternoons (up to 1.6 % in green areas), which of these grids are in the transition or VOC-limited regimes, was attributed to reductions in the anthropogenic emissions https://doi.org/10.5194/acp-23-10751-2023 Atmos. Chem. Phys., 23, 10751–10774, 2023 10762 A. Badia et al.: Modelling the impacts of emission changes on O3 chemistry Figure 7. Morning and afternoon averaged surface OH changes over the metropolitan area of Barcelona (AMB) and the Catalonia region (CAT) during 30 March to 12 April (only weekdays) and 18 to 30 May (only weekdays), with absolute values (ppt) and relative changes (%) shown. Relative changes (%) were calculated as ((COVID − BAU) / BAU) × 100. that decreased the ozone precursor levels (24 % to 40 % for O3 levels during the evening hours, where there was no pho- NOx and 24 % to 40 % for VOCs) and consequently ozone tolytic reaction with NO2 in urban areas with high population production. For green areas far from anthropogenic sources densities. We found that air quality policies based solely on (forests), the ozone levels were also reduced in the mornings transport reduction (as illustrated by the COVID lockdowns, and evenings during this period. which reduced NOx levels) actually intensified O3 levels The lockdown measures inhibited NO titration of O3 , over urban areas, indicating the need for a protocol with stri- mainly due to changes in the local NOx emissions result- dent control measures to reduce NOx emissions without sig- ing from road transport. This resulted in an increase in the nificantly reducing anthropogenic VOCs to control O3 levels. Atmos. Chem. Phys., 23, 10751–10774, 2023 https://doi.org/10.5194/acp-23-10751-2023 A. Badia et al.: Modelling the impacts of emission changes on O3 chemistry 10763 However, high ozone production during May was reduced due to reduced levels of the precursors, and consequently there were reductions in the maximum ozone levels for that period. 5.2 Impacts on the atmospheric oxidation capacity In addition to understanding changes in the levels of O3 pre- cursors, it is important to determine how emission changes affect the AOC, because this plays an important role in the loss and production rates of primary and secondary pollu- tants. The AOC is a good indicator of photochemical reac- tions, leading to the formation of secondary pollutants such as O3 (Wang et al., 2022). We saw increases in the con- centrations of the oxidants OH and NO3 during the period March–April, mainly over the AMB, as shown in Figs. 7 and 8, where the left-hand-side panels indicate absolute concen- trations (COVID − BAU) and the right-hand-side panels in- dicate relative changes in comparison to the BAU simulations ((COVID − BAU) / BAU × 100). OH is the dominant tropo- spheric daytime oxidant, and it increases considerably (up to 0.12 ppt, +45 %) because of significant reductions in NO2 levels, since NO2 is the primary OH sink (Elshorbany et al., 2009). The increase in these free radical concentrations could be the leading cause of the diurnal O3 increases (see Fig. S5) given that VOC and CO oxidations by OH are the initial re- actions for ozone formation. In addition, the NO3 radical, which is a primary nighttime oxidant, also increases in areas close to the airport and harbour (4 ppt, 210 %). This increase can be explained by reductions in the VOC and NO2 lev- els, which are important sinks for NO3 radicals (Elshorbany et al., 2009; Saiz-Lopez et al., 2017). During the period in May, we also found increases in the oxidant radicals (OH, NO3 ) and also O3 (see Fig. S5) of the COVID simulation with respect to the BAU in areas where substantial NOx emission reduction took place, such as the airport and the harbour. In these areas, OH levels increase up to 0.3 ppt (55 %) in the afternoon, and NO3 increases up to Figure 8. Evening-averaged surface NO3 changes over AMB and 4 ppt (230 %) in the evening. However, other areas showed CAT during 30 March to 12 April (only weekdays) and 18 to general decreases in AOC radicals (−0.1 ppt for OH and 30 May (only weekdays), with absolute values (ppt) and rela- −2 ppt for NO3 ), resulting in decreases in the O3 levels. Note tive changes (%) shown. Relative changes (%) were calculated as ((COVID − BAU) / BAU) × 100. that, for both periods, the decrease in shipping emissions (a source of NOx ) led to increases in the levels of both radicals along the ship tracks. limited regimes during this period (see Sect. 5.1). In terms of Our results indicate that changes in the anthropogenic air quality policy, it is important to understand the interplay emissions (mainly NOx and VOCs) lead to significant between these free radicals and O3 chemistry so that mitiga- changes in the OH and NO3 radical levels, which in the case tion strategies are not counterproductive. of emission reduction, such as that experienced during the COVID lockdown, leads to enhanced oxidation efficiency in the urban atmosphere of the AMB and O3 enhancements. 5.3 Pollution transport from urban to rural However, during the period in May, when O3 formation in- creased due to warm temperatures and an increase in bio- In addition to studying the mechanisms for ozone formation genic emissions, there was a decrease in the AOC (except in in the AMB, we explored how ozone is transported to ru- the airport and harbour areas) for the COVID run. The eleva- ral areas to determine the influence of urban pollution. Rural tions of AOC occurred because these areas were still VOC- areas far from the city, such as the Vic Plain and the Pyre- https://doi.org/10.5194/acp-23-10751-2023 Atmos. Chem. Phys., 23, 10751–10774, 2023 10764 A. Badia et al.: Modelling the impacts of emission changes on O3 chemistry Figure 9. Simulated air parcel trajectories at the footprint layer (0–300 m a.g.l., top panels) and interlayer (300–2000 m a.g.l., bottom panels) for 3 and 6 April at 16 h at the four sites (from left to right): Barcelona, Montseny, Tona (Vic Plain), and Pardines. The location of each site is shown with a green cross. Atmos. Chem. Phys., 23, 10751–10774, 2023 https://doi.org/10.5194/acp-23-10751-2023 A. Badia et al.: Modelling the impacts of emission changes on O3 chemistry 10765 Figure 10. Same as Fig. 9 for 22 and 26 May. The location of each site is shown with a green cross. https://doi.org/10.5194/acp-23-10751-2023 Atmos. Chem. Phys., 23, 10751–10774, 2023 10766 A. Badia et al.: Modelling the impacts of emission changes on O3 chemistry Figure 11. O3 changes between the COVID and BAU simulations along the atmospheric plume from the AMB to the Pyrenees for 3 April. The modelled PBLH is shown with a green line. nees mountain range, are frequently affected by the atmo- Figures 9 and 10 show the trajectories of the air masses arriv- spheric plume transported northwards from the AMB (Mas- ing at the monitoring stations on the selected days: these were sagué et al., 2019). Indeed, ozone exceedances over these modelled with the Lagrangian FLEXible PARTicle disper- places occur when there are high levels of NO2 (mainly sion model with WRF (FLEXPART-WRF) (Brioude et al., due to road traffic) over the AMB (see Sect. 2). The ur- 2013). This version of the Lagrangian model works with ban plume is driven inland by south-easterly and southerly the WRF mesoscale meteorological model, with the same combined sea–valley–mountain breeze winds channelled by parametrization as the WRF-Chem model (see Sect. 3.1). north-south valleys, and it crosses the Catalan coastal and The transport model has been run in backwards mode, which pre-coastal ranges to an intra-mountain plain. means that what is represented in each plot is the residence To understand how emission reductions in the AMB can time, at each grid cell of the map, for the air masses arriving affect the ozone levels in these rural areas, we analysed the at each site. Twenty-four-hour back-trajectories were calcu- differences in ozone concentrations between the COVID and lated for each day at a release time of 16 h and with a grid BAU runs for 4 specific days (3 and 6 April from the lock- cell size of 0.03 × 0.03 ◦ . Figures 9 and 10 show that the air down period and 22 and 26 May from the relaxation period) masses on 3 April and 22 May were transported from the during which the air masses flowed from the AMB north- AMB to rural areas such as Montseny and the Vic Plain, and wards to the Pyrenees (as shown in Figs. 9 and 10) and gen- we can see an influence from the bottom layers (0–300 m) erated high levels of O3 pollution in the rural areas (see Ta- and the upper layers (300–2000 m) at the different sites. The ble S1 and Figs. S8–S11). On 3 and 6 April there was no pre- air masses on 6 April were channelled from the AMB north- cipitation, with slightly high temperatures and low to mod- wards to Montseny, the Vic Plain, and the Pyrenees. The erate wind intensity, and on 22 and 26 May there was warm, air masses on 26 May were also transported from the AMB sunny weather with anti-cyclonic conditions (see Table S2 northwards to Montseny, the Vic Plain, and the Pyrenees, but for more information about the meteorological conditions). the air masses that arrived at the surfaces of these locations Atmos. Chem. Phys., 23, 10751–10774, 2023 https://doi.org/10.5194/acp-23-10751-2023 A. Badia et al.: Modelling the impacts of emission changes on O3 chemistry 10767 Figure 12. Same as Fig. 11 for 6 April. had strong local components and larger influences from the of Barcelona to the mountain area of Montseny. On that day, upper layers. this increase was not seen for the sites located farther north. On 3 April, we found significant differences between the Note that ozone decreases (1–2 ppb) were seen in the free COVID and BAU runs during the afternoon and especially in troposphere, especially at night. the evening, when O3 differences increased up to 2–3 ppb We found significant decreases in ozone levels in the and higher enhancements were found over the Vic Plain COVID simulation (up to 3 ppb from the surface up to (from the surface up to 2000 m), as shown in Fig. 11. High 2000 m) from the AMB north to the mountain area of ground-level ozone concentrations over the Vic Plain could Montseny on 22 and 26 May (Figs. 13 and 14, respec- have been affected by vertical recirculation of the air mass tively), especially during the afternoons, when the PBLH with O3 reservoirs from the upper layers (see Fig. S7) to was highest and solar radiation led to enhancement of the the lower troposphere (Querol et al., 2017; Massagué et al., sea breeze front, which provided favourable conditions for 2019). Later at night, ozone accumulated on the surface fol- regional transport (Massagué et al., 2019). The decreases in lowing a decrease in the planetary boundary layer height ozone precursor emissions (COVID simulation) resulted in (PBLH), and it was removed by deposition and titration. less ozone production from the AMB plume and production However, the reduction in daytime NOx emissions in the of new O3 , and consequently the ozone concentrations de- COVID simulations resulted in less titration capacity and creased (discussed in Sects. 5.1 and 5.2). This decrease was consequently an increase in the ozone levels (1–2 ppb) that also seen for the evening hours. Note that, at night, when remained at the surface layer along the different sites on ozone accumulated on the surface following the decrease in the S–N trajectory connecting the AMB, the Vic Plain, and the PBLH, there was a slight increase in its levels due to lim- the Pyrenees. For 6 April (Fig. 12), differences between the ited titration by NO. However, there were still reductions in COVID and BAU runs were only found during the evenings the ozone levels from 500 to 2000 m at night. and later in the nights, and the ozone levels increased in the Our results showed that reduced ozone precursor levels in- COVID run (up to 2 ppb at the surface level) from the city creased the ozone levels during the evenings and nights in https://doi.org/10.5194/acp-23-10751-2023 Atmos. Chem. Phys., 23, 10751–10774, 2023 10768 A. Badia et al.: Modelling the impacts of emission changes on O3 chemistry Figure 13. Same as Fig. 11 for 22 May. the COVID run due to reductions in the ozone titration pro- as would the recirculated O3 and thus the intensity of surface cess during the period of April–March, not only in urban ar- O3 fumigation from high-O3 reservoir layers in other areas. eas, but also in rural areas such as Montseny, the Vic Plain, and the Pyrenees. During May, the emission reductions in the COVID simulation decreased the maximum ozone levels 6 Conclusions (in comparison with the BAU run) in the afternoon from the city of Barcelona to the mountain areas of Montseny, the Vic Improving air quality is a top priority in urban areas and re- Plain, and the Pyrenees, thereby improving the air quality in quires a better understanding of how the O3 levels respond to all these areas. The improvement in the air quality is con- changes in the emission levels of the precursors, of the ozone sistent with the decrease in surface Ox concentrations seen formation regimes, and of the atmospheric oxidation capacity during the period of May over most of the Catalonia region and the associated O3 formation. Furthermore, urban emis- (see Sect. 4.1). sions affect the O3 levels in rural areas outside the cities. In A comparison of these two periods, April–March and May, this study, we used the WRF-Chem air quality model to anal- showed that the mitigation strategies designed to reduce the yse the air quality changes occurring over the metropolitan high ozone levels were more efficient in May, when ozone area of Barcelona and other (rural) areas affected by trans- formation was high (high biogenic emissions coinciding with port of the atmospheric plume from the AMB during mobil- anti-cyclonic conditions). Thus, given the importance of me- ity restrictions. teorology in air pollution events occurring over urban and The large reduction in NOx levels (up to 60 %) seen dur- rural areas, new mitigation strategies are needed to improve ing the lockdown period combined with a slight change in the air quality and would result in significant O3 reductions: VOC levels (up to 10 %) led to increased O3 concentrations the local O3 coming from the AMB plume would be reduced, (up to 20 % in the evening). The significant increase found in the evening was mainly due to reduced O3 titration by NO, which prevailed over the lower O3 production level caused by Atmos. Chem. Phys., 23, 10751–10774, 2023 https://doi.org/10.5194/acp-23-10751-2023 A. Badia et al.: Modelling the impacts of emission changes on O3 chemistry 10769 Figure 14. Same as Fig. 11 for 26 May. decreases in the levels of the O3 precursors. The lockdown During a period in May exhibiting high ozone production occurred during April–March, when ozone photochemical (high biogenic emissions and intense sunlight), decreases in production was still not at the highest level. In addition, our the levels of the ozone precursors NO2 and VOCs conse- results showed a significant increase in the atmospheric ox- quently decreased the maximum ozone (up to 4 µg m−3 ) and idation capacity (AOC) indicated by the enhanced oxidant Ox (up to 4 µg m−3 ) concentrations seen over Catalonia. This (OH and NO3 ) levels, which was consistent with the slight was consistent with the unchanged or decreased AOCs given increases seen in the maximum O3 concentrations during the their strong link with O3 production. For several days, this lockdown. We also found that, for several days, these in- decrease in the ozone level was seen farther north in the rural creases were seen farther north in rural areas such as the Vic areas. In this period, we also found ozone enhancements in Plain, which produced the most annual exceedances in Spain. the evening, mainly due to reduced O3 titration by NO. Large enhancements over these areas were the result of (1) a Furthermore, this analysis suggests that there was a ten- higher regional O3 background level, (2) vertical recircula- dency for both periods to move towards a NOx -sensitive tion of the air masses that transport high concentrations of O3 regime. However, some areas (open urban, compact urban, from the upper levels to the lower levels, and (3) the contri- industrial, and water) were still under VOC-limited or transi- butions of the AMB pollution plume travelling along the S–N tion regimes despite the remarkable NOx reduction. valley connecting the AMB, the Vic Plain, and the Pyrenees. We propose that measures intended to reduce ozone pre- High ozone levels seriously affect human health and the en- cursor emissions (NOx and VOC emissions) while maintain- vironment. In addition, the consistent differences seen in Ox ing stable O3 formation levels in the AMB will result in im- (NO2 + O3 ) concentrations during the period April–March portant reductions in O3 levels in both urban and rural areas, have important policy implications, i.e. that effective miti- especially in the spring and summer, when the ozone pro- gation strategies designed to reduce air pollutants and their ductivity is highest. The current policies based on reducing health effects should include reductions in both Ox and VOC transportation-related emissions alone could unintentionally levels to avoid increases in ozone levels. increase the AOCs in and around large cities, thereby in- https://doi.org/10.5194/acp-23-10751-2023 Atmos. Chem. Phys., 23, 10751–10774, 2023 10770 A. Badia et al.: Modelling the impacts of emission changes on O3 chemistry creasing the ozone levels. There is still a lack of research ties; the Sostenipra research group (grant no. 2021 SGR 00734) on relationships between AOC and ozone pollution. We also of the Department of Research and Universities of the Generalitat find that air quality policies must be designed in accordance de Catalunya; and the MCIN AEI/10.13039/501100011033 project with the VOC/NOx ratio, which dictates the O3 sensitivity of the Spanish Ministry of Science and Innovation. The authors specific to a city. Furthermore, the significant effect of NO acknowledge the computer resources at PICASSO and the tech- nical support provided by the Universidad de Málaga (grant no. titration demonstrates the importance of defining mitigation RES-AECT-2020-2-0004). Furthermore, the authors wish to thank strategies focused on VOC reductions. We also propose that the XVPCA for providing pollutant measurements from the vari- more measurements of VOC levels are required to constrain ous stations that they manage. The authors are also grateful to the the models representing these chemical processes, given the Barcelona Supercomputing Center for the use of HERMESv3_GR. complexity of the relationship between these pollutants. In Moreover, we wish to thank the Copernicus Atmosphere Monitor- addition, the difficulty of models in simulating urban ozone ing Service for the emission inventory. All of the numerical analy- precursors such as NOx and VOC levels and, consequently, ses were performed with the HTCondor cluster hosted by the Port their link to the ozone chemistry should be addressed in fu- d’Informació Científica (PIC). The authors acknowledge Qinyi Li ture work to use the models as effective tools for assessing and Xavier Querol for their constructive suggestions and feedback future studies aimed to reduce air quality. during this study. Code availability. The WRF-Chem model code is available Financial support. This research has been supported by the Euro- from http://www2.mmm.ucar.edu/wrf/users/download/get_sources. pean Research Council H2020 Research and Innovation programme html (WRF, 2023), with the specific code used in this study avail- (URBAG (grant no. 818002)). able from the authors upon request (alba.badia@uab.cat). Review statement. This paper was edited by John Orlando and Data availability. The WRF-Chem simulation out- reviewed by two anonymous referees. puts are openly available via the Zenodo repository at https://doi.org/10.5281/zenodo.8319390 (Badia, 2023). References Supplement. The supplement related to this article is available Ackermann, I. J., Hass, H., Memmesheimer, M., Ebel, A., online at: https://doi.org/10.5194/acp-23-10751-2023-supplement. Binkowski, F. S., and Shankar, U.: Modal aerosol dynam- ics model for Europe: development and first applications, At- mos. Environ., 32, 2981–2999, https://doi.org/10.1016/S1352- Author contributions. AB carried out all the model simulations 2310(98)00006-5, 1998. and data analyses, led the interpretation of the results, and prepared Agency, E. E., Guerreiro, C., Colette, A., Leeuw, F., and the manuscript with contributions from all the co-authors. GV con- González Ortiz, A.: Air quality in Europe: 2018 report, Publi- tributed to the interpretation of the results and provided extensive cations Office, https://doi.org/10.2800/777411, 2019. comments on the manuscript. RC ran the FLEXPART model and Anenberg, S. C., Horowitz, L. W., Tong, D. Q., and West, J. J.: analysed the output. SV analysed the synoptic situation of the do- An estimate of the global burden of anthropogenic ozone and main. VV and RS validated the meteorology of the model with ob- fine particulate matter on premature human mortality using at- servations. mospheric modeling, Environ. Health Persp., 118, 1189–1195, https://doi.org/10.1289/ehp.0901220, 2010. Badia, A., Iglesias-Suarez, F., Fernandez, R. P., Cuevas, C. A., Kin- Competing interests. The contact author has declared that none nison, D. E., Lamarque, J.-F., Griffiths, P. T., Tarasick, D. W., of the authors has any competing interests. Liu, J., and Saiz-Lopez, A.: The role of natural halogens in global tropospheric ozone chemistry and budget under differ- ent 21st century climate scenarios, J. Geophys. Res-Atmos., 126, e2021JD034859, https://doi.org/10.1029/2021JD034859, 2021a. Disclaimer. Publisher’s note: Copernicus Publications remains Badia, A., Langemeyer, J., Codina, X., Gilabert, J., Guilera, N., Vi- neutral with regard to jurisdictional claims in published maps and dal, V., Segura, R., Vives, M., and Villalba, G.: A take-home mes- institutional affiliations. sage from COVID-19 on urban air pollution reduction through mobility limitations and teleworking, npj Urban Sustainability, 6, 12, https://doi.org/10.1038/s42949-021-00037-7, 2021b. Acknowledgements. This work has been made possible by sup- Badia, A.: WRF-Chem output- concentrations for the two port from the European Research Council (ERC) Consolidator simulations: Business as Usual and COVID for the Grant project “Integrated System Analysis of Urban Vegetation and two periods, March–April and May [data set], Zenodo, Agriculture” (URBAG, grant no. 818002); the “Maria de Maeztu” https://doi.org/10.5281/zenodo.8319390, 2023. programme for Units of Excellence (grant no. CEX2019-000940- Bougeault, P. and Lacarrere, P.: Parameterization of Orography- M) of the Spanish Ministry of Science, Innovation and Universi- Induced Turbulence in a Mesobeta–Scale Model, Mon. Atmos. Chem. Phys., 23, 10751–10774, 2023 https://doi.org/10.5194/acp-23-10751-2023 A. Badia et al.: Modelling the impacts of emission changes on O3 chemistry 10771 Weather Rev., 117, 1872–1890, https://doi.org/10.1175/1520- son, I. R., DuVivier, A. K., Hodzic, A., and Randel, W. J.: 0493(1989)117<1872:POOITI>2.0.CO;2, 1989. The Whole Atmosphere Community Climate Model Version Brancher, M.: Increased ozone pollution alongside reduced nitrogen 6 (WACCM6), J. Geophys. Res-Atmos., 124, 12380–12403, dioxide concentrations during Vienna’s first COVID-19 lock- https://doi.org/10.1029/2019JD030943, 2019. down: Significance for air quality management, Environ. Pol- Giordano, L., Brunner, D., Flemming, J., Hogrefe, C., Im, U., Bian- lut, 284, 117153, https://doi.org/10.1016/j.envpol.2021.117153, coni, R., Badia, A., Balzarini, A., Baró, R., Chemel, C., Curci, G., 2021. Forkel, R., Jiménez-Guerrero, P., Hirtl, M., Hodzic, A., Honzak, Brioude, J., Arnold, D., Stohl, A., Cassiani, M., Morton, D., Seib- L., Jorba, O., Knote, C., Kuenen, J., Makar, P., Manders-Groot, ert, P., Angevine, W., Evan, S., Dingwell, A., Fast, J. D., Easter, A., Neal, L., Pérez, J., Pirovano, G., Pouliot, G., San José, R., R. C., Pisso, I., Burkhart, J., and Wotawa, G.: The Lagrangian Savage, N., Schröder, W., Sokhi, R., Syrakov, D., Torian, A., Tuc- particle dispersion model FLEXPART-WRF version 3.1, Geosci. cella, P., Werhahn, J., Wolke, R., Yahya, K., Žabkar, R., Zhang, Model Dev., 6, 1889–1904, https://doi.org/10.5194/gmd-6-1889- Y., and Galmarini, S.: Assessment of the MACC reanalysis and 2013, 2013. its influence as chemical boundary conditions for regional air Cristofanelli, P. and Bonasoni, P.: Background ozone in the quality modeling in AQMEII-2, Atmos. Environ., 115, 371–388, southern Europe and Mediterranean area: Influence of https://doi.org/10.1016/j.atmosenv.2015.02.034, 2015. the transport processes, Environ. Pollut., 157, 1399–1406, Grange, S. K., Lee, J. D., Drysdale, W. S., Lewis, A. C., Hueglin, https://doi.org/10.1016/j.envpol.2008.09.017, special Issue C., Emmenegger, L., and Carslaw, D. C.: COVID-19 lock- Section: Ozone and Mediterranean Ecology: Plants, People, downs highlight a risk of increasing ozone pollution in Eu- Problems, 2009. ropean urban areas, Atmos. Chem. Phys., 21, 4169–4185, Crutzen, P. J.: Photochemical reactions initiated by and influenc- https://doi.org/10.5194/acp-21-4169-2021, 2021. ing ozone in unpolluted tropospheric air, Tellus, 26, 47–57, Granier, C., Darras, S., Denier van der Gon, H., Doubalova, J., https://doi.org/10.1111/j.2153-3490.1974.tb01951.x, 1974. Elguindi, N., Galle, B., Gauss, M., Guevara, M., Jalkanen, J.- Derwent, R., Jenkin, M., and Saunders, S.: Photochemical ozone P., Kuenen, J., Liousse, C., Quack, B., Simpson, D., and Sin- creation potentials for a large number of reactive hydrocar- delarova, K.: The Copernicus Atmosphere Monitoring Service bons under European conditions, Atmos. Environ., 30, 181–199, global and regional emissions (April 2019 version), Tech. Rep., https://doi.org/10.1016/1352-2310(95)00303-G, 1996. https://doi.org/10.24380/D0BN-KX16, 2019. Doumbia, T., Granier, C., Elguindi, N., Bouarar, I., Darras, S., Grell, G. A. and Dévényi, D.: A generalized approach to pa- Brasseur, G., Gaubert, B., Liu, Y., Shi, X., Stavrakou, T., Tilmes, rameterizing convection combining ensemble and data as- S., Lacey, F., Deroubaix, A., and Wang, T.: Changes in global air similation techniques, Geophys. Res. Lett., 29, 38-1–38-4, pollutant emissions during the COVID-19 pandemic: a dataset https://doi.org/10.1029/2002GL015311, 2002. for atmospheric modeling, Earth Syst. Sci. Data, 13, 4191–4206, Grell, G. A., Peckham, S. E., Schmitz, R., McKeen, S. A., Frost, G., https://doi.org/10.5194/essd-13-4191-2021, 2021. Skamarock, W. C., and Eder, B.: Fully coupled “online” chem- Elshorbany, Y. F., Kurtenbach, R., Wiesen, P., Lissi, E., Rubio, M., istry within the WRF model, Atmos. Environ., 39, 6957–6975, Villena, G., Gramsch, E., Rickard, A. R., Pilling, M. J., and Kl- https://doi.org/10.1016/j.atmosenv.2005.04.027, 2005. effmann, J.: Oxidation capacity of the city air of Santiago, Chile, Guenther, A. B., Jiang, X., Heald, C. L., Sakulyanontvittaya, Atmos. Chem. Phys., 9, 2257–2273, https://doi.org/10.5194/acp- T., Duhl, T., Emmons, L. K., and Wang, X.: The Model of 9-2257-2009, 2009. Emissions of Gases and Aerosols from Nature version 2.1 Fleming, Z. L., Doherty, R. M., von Schneidemesser, E., Malley, (MEGAN2.1): an extended and updated framework for mod- C. S., Cooper, O. R., Pinto, J. P., Colette, A., Xu, X., Simp- eling biogenic emissions, Geosci. Model Dev., 5, 1471–1492, son, D., Schultz, M. G., Lefohn, A. S., Hamad, S., Moolla, https://doi.org/10.5194/gmd-5-1471-2012, 2012. R., Solberg, S., and Feng, Z.: Tropospheric Ozone Assessment Guerreiro, C. B., Foltescu, V., and de Leeuw, F.: Air quality Report: Present-day ozone distribution and trends relevant to status and trends in Europe, Atmos. Environ., 98, 376–384, human health, Elementa: Science of the Anthropocene, 6, 35, https://doi.org/10.1016/j.atmosenv.2014.09.017, 2014. https://doi.org/10.1525/elementa.273, 12, 2018. Guevara, M., Tena, C., Porquet, M., Jorba, O., and Pérez GBD 2019 Risk Factors Collaborators: Global burden of 87 risk García-Pando, C.: HERMESv3, a stand-alone multi-scale at- factors in 204 countries and territories, 1990–2019: a systematic mospheric emission modelling framework – Part 1: global analysis for the Global Burden of Disease Study 2019, Lancet, and regional module, Geosci. Model Dev., 12, 1885–1907, 396, 1223–1249, https://doi.org/10.1016/S0140-6736(20)30752- https://doi.org/10.5194/gmd-12-1885-2019, 2019. 2, 2020. Guevara, M., Jorba, O., Soret, A., Petetin, H., Bowdalo, D., Ser- Georgiou, G. K., Christoudias, T., Proestos, Y., Kushta, J., Hadjini- radell, K., Tena, C., Denier van der Gon, H., Kuenen, J., Peuch, colaou, P., and Lelieveld, J.: Air quality modelling in the summer V.-H., and Pérez García-Pando, C.: Time-resolved emission re- over the eastern Mediterranean using WRF-Chem: chemistry and ductions for atmospheric chemistry modelling in Europe during aerosol mechanism intercomparison, Atmos. Chem. Phys., 18, the COVID-19 lockdowns, Atmos. Chem. Phys., 21, 773–797, 1555–1571, https://doi.org/10.5194/acp-18-1555-2018, 2018. https://doi.org/10.5194/acp-21-773-2021, 2021. Gettelman, A., Mills, M. J., Kinnison, D. E., Garcia, R. R., Hersbach, H., Bell, B., Berrisford, P., Hirahara, S., Horányi, A., Smith, A. K., Marsh, D. R., Tilmes, S., Vitt, F., Bardeen, Muñoz-Sabater, J., Nicolas, J., Peubey, C., Radu, R., Schepers, C. G., McInerny, J., Liu, H.-L., Solomon, S. C., Polvani, L. M., D., Simmons, A., Soci, C., Abdalla, S., Abellan, X., Balsamo, Emmons, L. K., Lamarque, J.-F., Richter, J. H., Glanville, G., Bechtold, P., Biavati, G., Bidlot, J., Bonavita, M., De Chiara, A. S., Bacmeister, J. T., Phillips, A. S., Neale, R. B., Simp- G., Dahlgren, P., Dee, D., Diamantakis, M., Dragani, R., Flem- https://doi.org/10.5194/acp-23-10751-2023 Atmos. Chem. Phys., 23, 10751–10774, 2023 10772 A. Badia et al.: Modelling the impacts of emission changes on O3 chemistry ming, J., Forbes, R., Fuentes, M., Geer, A., Haimberger, L., Monks, P. S., Archibald, A. T., Colette, A., Cooper, O., Coyle, M., Healy, S., Hogan, R. J., Hólm, E., Janisková, M., Keeley, S., Derwent, R., Fowler, D., Granier, C., Law, K. S., Mills, G. E., Laloyaux, P., Lopez, P., Lupu, C., Radnoti, G., de Rosnay, P., Stevenson, D. S., Tarasova, O., Thouret, V., von Schneidemesser, Rozum, I., Vamborg, F., Villaume, S., and Thépaut, J.-N.: The E., Sommariva, R., Wild, O., and Williams, M. L.: Tropospheric ERA5 global reanalysis, Q. J. Roy. Meteor. Soc., 146, 1999– ozone and its precursors from the urban to the global scale from 2049, https://doi.org/10.1002/qj.3803, 2020. air quality to short-lived climate forcer, Atmos. Chem. Phys., 15, Im, U., Bianconi, R., Solazzo, E., Kioutsioukis, I., Badia, A., 8889–8973, https://doi.org/10.5194/acp-15-8889-2015, 2015. Balzarini, A., Baró, R., Bellasio, R., Brunner, D., Chemel, C., National Research Council, N.: Rethinking the Ozone Problem Curci, G., Denier van der Gon, H., Flemming, J., Forkel, R., in Urban and Regional Air Pollution, The National Academies Giordano, L., Jiménez-Guerrero, P., Hirtl, M., Hodzic, A., Hon- Press, Washington, DC, https://doi.org/10.17226/1889, 1991. zak, L., Jorba, O., Knote, C., Makar, P. A., Manders-Groot, A., Neiburger, M.: the role of meteorology in the study and con- Neal, L., Pérez, J. L., Pirovano, G., Pouliot, G., San Jose, R., trol of air pollution, B. Am. Meteorol. Soc., 50, 957–966, Savage, N., Schroder, W., Sokhi, R. S., Syrakov, D., Torian, https://doi.org/10.1175/1520-0477-50.12.957, 1969. A., Tuccella, P., Wang, K., Werhahn, J., Wolke, R., Zabkar, R., Pyrgou, A., Hadjinicolaou, P., and Santamouris, M.: Enhanced near- Zhang, Y., Zhang, J., Hogrefe, C., and Galmarini, S.: Evalua- surface ozone under heatwave conditions in a Mediterranean tion of operational online-coupled regional air quality models island, Sci. Rep., 8, 9191, https://doi.org/10.1038/s41598-018- over Europe and North America in the context of AQMEII phase 27590-z, 2018. 2. Part II: Particulate matter, Atmos. Environ., 115, 421–441, Querol, X., Gangoiti, G., Mantilla, E., Alastuey, A., Minguillón, M. https://doi.org/10.1016/j.atmosenv.2014.08.072, 2015. C., Amato, F., Reche, C., Viana, M., Moreno, T., Karanasiou, Jaén, C., Udina, M., and Bech, J.: Analysis of two heat wave driven A., Rivas, I., Pérez, N., Ripoll, A., Brines, M., Ealo, M., Pan- ozone episodes in Barcelona and surrounding region: Meteo- dolfi, M., Lee, H.-K., Eun, H.-R., Park, Y.-H., Escudero, M., Bed- rological and photochemical modeling, Atmos. Environ., 246, dows, D., Harrison, R. M., Bertrand, A., Marchand, N., Lyasota, 118037, https://doi.org/10.1016/j.atmosenv.2020.118037, 2021. A., Codina, B., Olid, M., Udina, M., Jiménez-Esteve, B., Soler, Karl, T., Graus, M., Striednig, M., Lamprecht, C., Hammerle, M. R., Alonso, L., Millán, M., and Ahn, K.-H.: Phenomenology A., Wohlfahrt, G., Held, A., von der Heyden, L., Deventer, of high-ozone episodes in NE Spain, Atmos. Chem. Phys., 17, M. J., Krismer, A., Haun, C., Feichter, R., and Lee, J.: Ur- 2817–2838, https://doi.org/10.5194/acp-17-2817-2017, 2017. ban eddy covariance measurements reveal significant miss- Querol, X., Massagué, J., Alastuey, A., Moreno, T., Gangoiti, G., ing NOx emissions in Central Europe, Sci. Rep., 7, 2536, Mantilla, E., Duéguez, J. J., Escudero, M., Monfort, E., Pérez https://doi.org/10.1038/s41598-017-02699-9, 2017. García-Pando, C., Petetin, H., Jorba, O., Vázquez, V., de la Rosa, Kleanthous, S., Vrekoussis, M., Mihalopoulos, N., Kalabokas, J., Campos, A., Muñóz, M., Monge, S., Hervás, M., Javato, R., P., and Lelieveld, J.: On the temporal and spatial variation and Cornide, M. J.: Lessons from the COVID-19 air pollution of ozone in Cyprus, Sci. Total Environ., 476–477, 677–687, decrease in Spain: Now what?, Sci. Total Environ., 779, 146380, https://doi.org/10.1016/j.scitotenv.2013.12.101, 2014. https://doi.org/10.1016/j.scitotenv.2021.146380, 2021. Liu, F., Page, A., Strode, S. A., Yoshida, Y., Choi, S., Zheng, Ribeiro, I., Martilli, A., Falls, M., Zonato, A., and Vil- B., Lamsal, L. N., Li, C., Krotkov, N. A., Eskes, H., van der lalba, G.: Highly resolved WRF-BEP/BEM simulations over A, R., Veefkind, P., Levelt, P. F., Hauser, O. P., and Joiner, Barcelona urban area with LCZ, Atmos. Environ., 248, 105220, J.: Abrupt decline in tropospheric nitrogen dioxide over China https://doi.org/10.1016/j.atmosres.2020.105220, 2021. after the outbreak of COVID-19, Sci. Adv., 6, eabc2992, Rico, M., Font, L., Arimon, J., Marí, M., Gómez, A., and E., R.: https://doi.org/10.1126/sciadv.abc2992, 2020. Informe qualitat de l’aire de Barcelona, Tech. rep., Agència Mar, K. A., Ojha, N., Pozzer, A., and Butler, T. M.: Ozone air qual- de Salut Pública de Barcelona, https://www.aspb.cat/documents/ ity simulations with WRF-Chem (v3.5.1) over Europe: model qualitat-aire-2019 (last access: 20 September 2023), 2019. evaluation and chemical mechanism comparison, Geosci. Model Rico, M., Font, L., Arimon, J., Gómez, A., and E., R.: Informe Dev., 9, 3699–3728, https://doi.org/10.5194/gmd-9-3699-2016, qualitat de l’aire de Barcelona, Tech. rep., Agència de Salut 2016. Pública de Barcelona, 69 pp., https://www.aspb.cat/documents/ Martín-Vide, J., Brunet, M., Prohom, M., and Rius, A.: Segon in- qualitat-aire-2020 (last access: 20 September 2023), 2020. forme sobre el canvi climàtic a Catalunya. Capítol 2. Els climes Rivas, I., Viana, M., Moreno, T., Pandolfi, M., Amato, F., Reche, de Catalunya. Present i tendències recents, Tech. Rep., Generali- C., Bouso, L., Àlvarez Pedrerol, M., Alastuey, A., Sunyer, J., tat de Catalunya, ISBN 978-84-9965-027-2, 2010. and Querol, X.: Child exposure to indoor and outdoor air pollu- Massagué, J., Carnerero, C., Escudero, M., Baldasano, J. M., tants in schools in Barcelona, Spain, Environ. Int., 69, 200–212, Alastuey, A., and Querol, X.: 2005–2017 ozone trends and poten- https://doi.org/10.1016/j.envint.2014.04.009, 2014. tial benefits of local measures as deduced from air quality mea- Romero-Alvarez, J., Lupaşcu, A., Lowe, D., Badia, A., Archer- surements in the north of the Barcelona metropolitan area, At- Nicholls, S., Dorling, S., Reeves, C. E., and Butler, T.: Sources mos. Chem. Phys., 19, 7445–7465, https://doi.org/10.5194/acp- of surface O3 in the UK: tagging O3 within WRF-Chem, Atmos. 19-7445-2019, 2019. Chem. Phys., 22, 13797–13815, https://doi.org/10.5194/acp-22- Miyazaki, K., Bowman, K., Sekiya, T., Takigawa, M., Neu, 13797-2022, 2022. J. L., Sudo, K., Osterman, G., and Eskes, H.: Global tropo- Roozitalab, B., Carmichael, G. R., Guttikunda, S. K., and spheric ozone responses to reduced NOx emissions linked to Abdi-Oskouei, M.: Elucidating the impacts of COVID- the COVID-19 worldwide lockdowns, Sci. Adv., 7, eabf7460, 19 lockdown on air quality and ozone chemical char- https://doi.org/10.1126/sciadv.abf7460, 2021. Atmos. Chem. Phys., 23, 10751–10774, 2023 https://doi.org/10.5194/acp-23-10751-2023 A. Badia et al.: Modelling the impacts of emission changes on O3 chemistry 10773 acteristics in India, Environ. Sci.-Atmos., 2, 1183–1207, bined impacts of climate and air pollution on human health https://doi.org/10.1039/D2EA00023G, 2022. and agricultural productivity, Environ. Res. Lett., 16, 093004, Saiz-Lopez, A., Borge, R., Notario, A., Adame, J. A., de la https://doi.org/10.1088/1748-9326/ac1df8, 2021. Paz, D., Querol, X., Artíñano, B., Gómez-Moreno, F. J., and Stewart, I. D. and Oke, T. R.: Local Climate Zones for Urban Cuevas, C. A.: Unexpected increase in the oxidation capacity Temperature Studies, B. Am. Meteorol. Soc., 93, 1879–1900, of the urban atmosphere of Madrid, Spain, Sci. Rep., 7, 45956, https://doi.org/10.1175/bams-d-11-00019.1, 2012. https://doi.org/10.1038/srep45956, 2017. Stockwell, W. R., Middleton, P., Chang, J. S., and Tang, X.: The sec- Salamanca, F., Martilli, A., Tewari, M., and Chen, F.: A Study ond generation regional acid deposition model chemical mecha- of the Urban Boundary Layer Using Different Urban Pa- nism for regional air quality modeling, J. Geophys. Res-Atmos., rameterizations and High-Resolution Urban Canopy Parame- 95, 16343–16367, https://doi.org/10.1029/JD095iD10p16343, ters with WRF, J. Appl. Meteorol. Clim., 50, 1107–1128, 1990. https://doi.org/10.1175/2010jamc2538.1, 2011. Tuccella, P., Curci, G., Visconti, G., Bessagnet, B., Menut, L., Schell, B., Ackermann, I. J., Hass, H., Binkowski, F. S., and Park, R. J.: Modeling of gas and aerosol with WRF/Chem and Ebel, A.: Modeling the formation of secondary or- over Europe: Evaluation and sensitivity study, J. Geophys. Res- ganic aerosol within a comprehensive air quality model Atmos, 117, D3, https://doi.org/10.1029/2011JD016302, 2011. system, J. Geophys. Res-Atmos, 106, 28275–28293, Venter, Z. S., Aunan, K., Chowdhury, S., and Lelieveld, https://doi.org/10.1029/2001JD000384, 2001. J.: COVID-19 lockdowns cause global air pollution de- Segura, R., Badia, A., Ventura, S., Gilabert, J., Martilli, A., clines, P. Natl. Acad. Sci. USA, 117, 18984–18990, and Villalba, G.: Sensitivity study of PBL schemes and https://doi.org/10.1073/pnas.2006853117, 2020. soil initialization using the WRF-BEP-BEM model over von Schneidemesser, E., Sibiya, B., Caseiro, A., Butler, T., a Mediterranean coastal city, Urban Climate, 39, 100982, Lawrence, M. G., Leitao, J., Lupascu, A., and Salvador, https://doi.org/10.1016/j.uclim.2021.100982, 2021. P.: Learning from the COVID-19 lockdown in berlin: Ob- Servei Meteorològic de Catalunya: Butlletí climàtic mensual (maig servations and modelling to support understanding poli- del 2020), Tech. rep., Departament de Territori i Sostenibilitat., cies to reduce NO2 , Atmos. Environ. X, 12, 100122, https://static-m.meteo.cat/wordpressweb/wp-content/uploads/ https://doi.org/10.1016/j.aeaoa.2021.100122, 2021. 2021/03/01155108/Butllet-Maig2020_v2.pdf (last access: 20 Wang, H., Huang, C., Tao, W., Gao, Y., Wang, S., Jing, S., Wang, September 2023), 2020. W., Yan, R., Wang, Q., An, J., Tian, J., Hu, Q., Lou, S., Pöschl, Sharma, S., Zhang, M., Anshika, Gao, J., Zhang, H., and U., Cheng, Y., and Su, H.: Seasonality and reduced nitric oxide Kota, S. H.: Effect of restricted emissions during COVID- titration dominated ozone increase during COVID-19 lockdown 19 on air quality in India, Sci. Total Environ., 728, 138878, in eastern China, npj Climate and Atmospheric Science, 5, 24, https://doi.org/10.1016/j.scitotenv.2020.138878, 2020. https://doi.org/10.1038/s41612-022-00249-3, 2022. Sicard, P., De Marco, A., Troussier, F., Renou, C., Vas, N., and Pao- Wang, Y., Zhu, S., Ma, J., Shen, J., Wang, P., Wang, P., letti, E.: Decrease in surface ozone concentrations at Mediter- and Zhang, H.: Enhanced atmospheric oxidation capacity and ranean remote sites and increase in the cities, Atmos. Environ., associated ozone increases during COVID-19 lockdown in 79, 705–715, https://doi.org/10.1016/j.atmosenv.2013.07.042, the Yangtze River Delta, Sci. Total Environ., 768, 144796, 2013. https://doi.org/10.1016/j.scitotenv.2020.144796, 2021. Sicard, P., Marco, A. D., Agathokleous, E., Feng, Z., Xu, Wesely, M.: Parameterization of surface resistances to gaseous dry X., Paoletti, E., Rodriguez, J. J. D., and Calatayud, deposition in regional-scale numerical models, Atmos. Envi- V.: Amplified ozone pollution in cities during the ron., 41, 52–63, https://doi.org/10.1016/j.atmosenv.2007.10.058, COVID-19 lockdown, Sci. Total Environ., 735, 139542, 2007. https://doi.org/10.1016/j.scitotenv.2020.139542, 2020. Wild, O.: Modelling the global tropospheric ozone budget: explor- Sicard, P., Agathokleous, E., Marco, A. D., Paoletti, E., and Ca- ing the variability in current models, Atmos. Chem. Phys., 7, latayud, V.: Urban population exposure to air pollution in Eu- 2643–2660, https://doi.org/10.5194/acp-7-2643-2007, 2007. rope over the last decades, Environmental Sciences Europe, 33, Wild, O., Zhu, X., and Prather, M. J.: Fast-J: Accurate 28, https://doi.org/10.1186/s12302-020-00450-2, 2021. Simulation of In- and Below-Cloud Photolysis in Tropo- Sillman, S.: The relation between ozone, NOx and hydrocarbons spheric Chemical Models, J. Atmos. Chem., 37, 245–282, in urban and polluted rural environments, Atmos. Environ., https://doi.org/10.1023/a:1006415919030, 2000. 33, 1821–1845, https://doi.org/10.1016/S1352-2310(98)00345- World Health Organization: WHO global air quality guidelines: par- 8, 1999. ticulate matter (PM2.5 and PM10 ), ozone, nitrogen dioxide, sul- Sillman, S.: 9.11 – Tropospheric Ozone and Photochemical fur dioxide and carbon monoxide, Tech. rep., World Health Or- Smog, in: Treatise on Geochemistry, edited by: Holland, ganization, 273 pp., ISBN 9789240034228, 2021. H. D. and Turekian, K. K., 407–431, Pergamon, Oxford, WRF: WRF Source Codes and Graphics Software Download Page, https://doi.org/10.1016/B0-08-043751-6/09053-8, 2003. WRF [code], http://www2.mmm.ucar.edu/wrf/users/download/ Sillman, S., Logan, J. A., and Wofsy, S. C.: The sensitivity get_sources.html (last access: 2 June 2022), 2023. of ozone to nitrogen oxides and hydrocarbons in regional Yegorova, E. A., Allen, D. J., Loughner, C. P., Pickering, K. E., and ozone episodes, J. Geophys. Res.-Atmos., 95, 1837–1851, Dickerson, R. R.: Characterization of an eastern U.S. severe air https://doi.org/10.1029/JD095iD02p01837, 1990. pollution episode using WRF/Chem, J. Geophys. Res.-Atmos., Sillmann, J., Aunan, K., Emberson, L., Büker, P., Oort, B. V., 116, D17, https://doi.org/10.1029/2010JD015054, 2011. O’Neill, C., Otero, N., Pandey, D., and Brisebois, A.: Com- https://doi.org/10.5194/acp-23-10751-2023 Atmos. Chem. Phys., 23, 10751–10774, 2023 10774 A. Badia et al.: Modelling the impacts of emission changes on O3 chemistry Zhu, S., Poetzscher, J., Shen, J., Wang, S., Wang, P., and Zhang, Zittis, G., Hadjinicolaou, P., Fnais, M., and Lelieveld, J.: Projected H.: Comprehensive Insights Into Osub3/sub Changes During changes in heat wave characteristics in the eastern Mediterranean the COVID-19 From Osub3/sub Formation Regime and At- and the Middle East, Reg. Environ. Change, 16, 1863–1876, mospheric Oxidation Capacity, Geophys. Res. Lett., 48, 10, https://doi.org/10.1007/s10113-014-0753-2, 2015. https://doi.org/10.1029/2021gl093668, 2021. Atmos. Chem. Phys., 23, 10751–10774, 2023 https://doi.org/10.5194/acp-23-10751-2023