Measurement report: Long-term real-time characterisation of the submicronic aerosol and its atmospheric dynamic in a Mediterranean coastal city: Tracking the polluted events at the Marseille-Longchamp supersite

. A supersite was recently implemented in Marseille to conduct intensive and advanced measurement studies for ambient aerosols. A Time-of-Flight Aerosol Chemical Speciation Monitor (ToF-ACSM) was deployed to investigate the chemical composition of submicronic aerosol over a 14-month period (1 February 2017 - 13 April 2018). Parallel 15 measurements were performed with an Aethalometer, an ultrafine particle monitor and a suite of instruments to monitor regulated pollutants (PM 2.5 , PM 10 , NO x , O 3 and SO 2 ). The averaged PM 1 chemical composition over the period was dominated by organics (49.7%) and black carbon (17.1%) while sulfate accounted for 14.6%, nitrate for 10.2%, ammonium for 7.9% and chloride for 0.5% only. Wintertime was found to be the season contributing the most to the annual PM 1 mass concentration (30%), followed by autumn (26%), summer (24%) and spring (20%). During this season, OA and BC concentrations were 20 found to contribute to 32% and 31% of their annual concentrations, respectively, as a combined result of heavy urban traffic, high emissions from residential heating, open combustion of green wastes and low planetary boundary layer (PBL) height. In summer, sulfate contribution to PM 1 increased with an average and a maximum contribution to the PM 1 of 24% and 66%. This is partly land breeze conditions were observed during local pollution episodes, while high level of oxygenated OA and inorganic nitrate were associated to medium/long-range transported particles. In conclusion this supersite showed a high potential for the study of seasonality and pollution episodes phenomenology in 35 Marseille over multiple geographic scales. The present paper highlights the significant contribution of regional transport of pollutants to the local air pollution that must be considered by local authorities in deploying effective PM abatement strategies.

The Mediterranean city of Marseille, as a highly urbanised area, exposed to a variety of anthropic (traffic, residential heating, shipping, industries) and biogenic (terrestrial vegetation, marine aerosols) sources is a challenging area for fine particle studies. The ESCOMPTE experiment (2001) demonstrated that the topography and the air mass circulation, characterized by local and 70 mesoscale winds, drives the pollution levels in the city (Cachier et al., 2005;Drobinski et al., 2007;Mestayer et al., 2005;Puygrenier et al., 2005). High level of atmospheric pollutants such as fine particles, have often been observed in Marseille, where mortality rate and cardiovascular hospital admissions are significantly elevated, even higher than in Paris whose population is 6 times higher (Pascal et al., 2013).
Fine particles have been previously characterized in Marseille during an intensive field campaign (3 weeks in summer 2008), 75 (El Haddad et al., 2011a, 2011b, 2013. The seasonal variations and sources of aerosol have been well documented through the offline analysis of daily filter samples collected over 1 year (Bozzetti et al., 2017;Salameh et al., 2015Salameh et al., , 2018 but this methodology only gives poor temporal resolution compared to online instruments and therefore cannot capture the fast changes in concentration and chemical composition. In this context a new atmospheric urban background supersite dedicated to the long-term and real time chemical and physical 80 characterization of submicron aerosol was recently implemented in Marseille. This supersite gathers state of art instruments for the measurement of aerosols (chemical composition and size distribution) and a suite of instruments for the monitoring of regulated pollutants (PM2.5, PM10, NOx, O3 and SO2). The goal of the present paper is to characterize the long-term phenomenology of submicron aerosol in a coastal city. The seasonal variations, diurnal profiles, and geographical origins of PM1 are presented with a focus on local and long-range pollution episodes when PM exceedance days occur. Also, the chemical 85 characteristics of shipping/industrial emissions in summer are explored through backtrajectories analysis.
the second largest harbour of the Mediterranean Sea and the first French harbour, with almost 4000 ship berthing in the several basins of Marseille for the year 2017. At 40 km North-West of the city is located the large industrial complex of Fos-sur-mer with petroleum refining, shipbuilding, steel facilities, and coke production plants (El Haddad et al., 2011b;Salameh et al., 2018). The region is well-known for active photochemistry inducing high ozone concentrations (Flaounas et al., 2009) during 100 summer periods (Figure 1), and frequent secondary organic aerosol formation events (El Haddad et al., 2013). Air mass circulation is complex in Marseille area (Drobinski et al., 2007;El Haddad et al., 2013;Flaounas et al., 2009) and is held by the surrounding topography. The city is bordered by Mediterranean Sea from the southwest and enclosed from the north, east and south by mountain ranges up to 700 m a.s.l. The synoptic air masses come from the Rhone valley, the Atlantic and Mediterranean Sea (Drobinski et al., 2007). Moreover, Marseille air quality is often affected by the Mistral wind and sea/land 105 breeze cycles. The Mistral is a strong wind blowing from the North-West (310°-360°) along the lower Rhône River valley toward the Mediterranean Sea. The South-Westerly sea breeze (190°-270°) and North-Easterly land breeze (5°-90°) are local winds prevailing during weak Mistral wind ( Figure S1). Land breeze circulation is often associated with high pollution levels over Marseille due to the low pollutants dispersion. In the early morning of summer days, Marseille is directly downwind of the industrial area and the harbour basins. As the temperature of the land surface rises, sea breeze sets in and the polluted air 110 masses from the industrial area are transported over the Mediterranean Sea before reaching the city. Averaged SO2 concentrations display a decreasing trend thanks to EU legislation on emission control and lower fuel sulphur content (limited to 50 ppm in 2005 and then to 10 ppm since 2009). Another explanation could be the decline of maritime transport during several years since the 2009 financial crisis. But from 2013, however, SO2 concentrations seem to increase again and could be linked to the enhancement of maritime activity at Marseille harbour. Indeed the French goods maritime 120 transport rose again between 2012 and 2017 (+5.9%), together with the passenger maritime transport that has continuously been enhanced for the past 10 years (+31.7%) (French Commissioner-General for Sustainable Development, 2019). For particulate matter (PM10, PM2.5), annual concentrations show very slight decreases in the last 11 years but remain above the WHO recommendations. This is consistent with the global decrease in the EU-28 since 2000 (27 and 28% reduction for PM10 and PM2.5, respectively) and the PM2.5 target value was exceeded in 21 countries (European Environment Agency, 2019). In 125 the last 2 years about 12 and 28 days with exceedance concentrations were recorded respectively for PM10 and PM2.5, mainly in winter periods.

ACSM sampling and data corrections
Ambient submicron particles (NR-PM1) were measured continuously from 1 February 2017 to 13 April 2018 using a time-of-130 flight aerosol chemical speciation monitor (ToF-ACSM, Aerodyne Research Inc., USA). The instrument provides quantitative assessment of non-refractory species as organics, nitrate, sulfate, ammonium and chloride in the size range 40-1000 nm. The aerosol is sampled at the main inlet at a flow rate of 3 L min -1 and dried using a Nafion dryer system (Perma Pure, New Jersey, USA) to keep the relative humidity (RH) below 40%. A subsample flow of 0.085 lpm passes through a critical orifice and enters an aerodynamic lens that focuses the particles into a narrow beam, these are then flash-vaporized upon impaction on a 135 heated tungsten plate at 600°C. The resulting vapours are ionized using 70 eV electron impact (EI) ionization. The time-offlight mass spectrometer (ETOF, TOFWERK, Thun, Switzerland) provides mass spectra at a mass-to-charge resolution of M/ΔM=600. The data were acquired at a time resolution of 15 min using Igro-DAQ v.2.1.4 software and by Tofware v.2.5.13 written in Igor Pro (Wave Metric inc., Lake Oswego, Oregon, USA). Further description and detail of the instrument are presented by Fröhlich et al. (2013Fröhlich et al. ( , 2015 and Timonen et al. (2016). Calibrations of ionization efficiency (IE) of nitrate and 140 relative ionization efficiency (RIE) of ammonium and sulfate were repeated 3 times over the 14-month measurement period.
The calculated values are summarized in table S1. Table S2 lists the detection limits, calculated as three time the noise level, for the 5 quantified species. The collection efficiencies values (CE) were corrected using algorithms described by Middlebrook et al. (2012), the time-dependant CE are shown in Figure S2. For this dataset CE is assessed as 0.47±0.05 which is comparable to values typically found for ambient aerosol (0.5, Middlebrook et al., 2012). An overall uncertainty of ± 30% is associated to 145 the total mass concentrations. It includes the uncertainties on the IE, RIE and CE values (Bahreini et al., 2009).
The organic aerosol (OA) mass was corrected to account for measurement interferences. According to Pieber et al. (2016) ammonium nitrate induces an overestimation of OA at m/z 44. A correction is introduced in the fragmentation table by measuring the relationship between measured CO2 + and the NH4NO3 mass measured during ToF-ACSM calibrations (see equation S1). Our dataset showed very little contribution of NH4NO3 on the organic m/z 44 with value ranging from 0.1-0.5% 150 (table S3).
The ion fragments at m/z 30 and m/z 46 assigned to nitrate (NO + and NO2 + ) may contain interferences from organic species like CH2O + at m/z 30 and CH2O2 + at m/z 46. These interferences lead to an overestimation of UMR nitrate and can falsely suggest the presence of organic nitrate in high OA/NO3environments. Here the m/z 30 and m/z 46 signals have been corrected for these interferences by using correlated organic signals respectively at m/z 29 from CHO + and m/z 45 from CHO2 + (equation 155 S2), as recommended by Fry et al. (2018). These peaks were the closest organic signals to the nitrate peaks with organic interferences.

Collocated instruments
ACMS measurements were combined with several on-line collocated instruments. A dual spot 7-wavelenght AE33 Aethalometer (Magee Scientific) (Drinovec et al., 2015) equipped with a PM2.5 cut-off inlet was used to measure equivalent 160 black carbon (BC) concentrations at a 1 min time resolution. Equivalent black carbon concentrations were calculated from the absorption coefficient at 880 nm with the default mass absorption cross section (MAC) implemented in the AE33 (7.77 m² g -1 ). The submicron aerosol number size distribution was investigated with the model 3031 ultrafine particle monitor (TSI Inc., Minnesota, USA) for the whole study period. This instrument provides measurements from 20 to 1000 nm, with six channels of size resolution. The aerosol number size distribution in the range 10.25-1084 nm was further explored over 45 channels 165 using a Scanning Mobility Particle Sizer system (SMPS, L-DMA, CPC5403, GRIMM) for two periods: from 23 June to 12 August 2017 (summer period) and from 6 November 2017 to 11 January 2018 (winter period).
Off-line measurements were carried out to collect particles (24h PM1) onto 150 mm-diameter quartz fiber filters (Pall Gellman, TISSUQUARTZ) (8h00 to 8h00 UT) using a high volume sampler (Digitel DA-80) operating at a flowrate of 30 m 3 .h -1 . 45 filters were discontinuously sampled from 01 March to 01 May 2017 (22 filters) and from 01 July to 23 September 2017 (23 170 filters). These filters were analysed in order to determine the major anions and cations using ion chromatography (Sciare et al., 2008), and elemental/organic content using Sunset OC/EC analyser (EUSAAR2 thermal protocol) according to Cavalli et al. (2010).
Continuous measurements of SO2, NOx and O3, and PMx were carried out by the air quality monitoring station. A M100E UV fluorescence analyser, a M200E chemiluminescence analyser (Teledyne API, California, USA) and a Serinus 10 ozone 175 analyser (Ecotech, Australia) were deployed for the SO2, NOx and O3 measurements, respectively. A Continuous Betaattenuation continuous particulate monitor (BAM 1020, Met One Instruments Inc., Oregon, USA) was used to measure the mass concentrations of PM2.5 and PM10 and an optical particle counter (FIDAS 200, PALAS, Germany) was dedicated to the measurement of PM1, PM2.5 and PM10 since February 2018. All the time resolutions of this instrumental panel were synchronized to 15 min. 180

Meteorological data and backtrajectories analysis
The site is equipped with an anemometer sonic 3D to provide temperature and wind measurements (directions and speeds) according to 3 orthogonal axes. Precipitations and relative humidity parameters were taken from the Vaudrans meteorological station located 6 km away for the MRS-LC site (43°18′26″N; 5°28′28″E). Non-parametric wind regression (NWR; Henry et al., 2009) and sustained wind incidence method-2 (SWIM-2; Olson et al., 2012) algorithms were used to generate pollution 185 roses. NWR and SWIM-2 analyses were performed using the ZeFir toolkit (Petit et al., 2017a). To investigate air mass origin during specific pollution episodes, 72h-backtrajectories were calculated every hour from the PC-based version of HYSPLIT (Draxler et al., 1999) (Ashbaugh et al., 1985;Petit et al., 2017b). For this study, the CWT domain was set in the range of (40-46° N; -5-10° E) with the grid cell size of 0.05°×0. 05°. 195 3 Results and discussion

Cross-validation of PM1 chemical species concentrations
The PM1 mass concentrations measured by the ToF-ACSM were compared with 24h-PM1 filter measurements ( Figure S3).
The ACSM concentrations were daily averaged and compared with respective offline measurements from 1 March to 23 September 2017 (n=46). A good agreement is found for ammonium and sulfate (R 2 =0.71 and slope of 0.84, and R 2 =0.76 and 200 slope of 0.89, respectively). For nitrate the results are less consistent (R 2 =0.69 and slope of 1.22). This higher slope can be attributed to the volatilization of nitrate from the filters during hot periods (Ripoll et al., 2015;Schaap et al., 2004a). The NH4 measured/NH4 predicted ratio was also investigated from 1 February 2017 to 13 April 2018 ( Figure S4) as an indirect proxy for particle acidity (Zhang et al., 2007a) and/or presence of high organic nitrate concentrations (Petit et al., 2017b). The NH4 predicted represents the theoretical ammonium concentration needed to neutralize the inorganic species concentrations (NO3 -, SO4 2-, Cl -205 ). For most of the cities inorganic species tend to be fully neutralized and a limited amount of acidic particles can only be observed when the site is located close to an emission source (e.g. industries, harbour, fire event…). Here the slope value close to 1 (0.95) reflects a full neutralization for all anions and suggests that there is enough ammonia in the gas phase to neutralize these species despite the nearby harbour and industrial complex.
The OC to organics comparison (filters vs ACSM concentrations) showed a good correlation with R 2 =0.79 with a slope 210 (corresponding to the OM-to-OC ratio) of 1.9. This value is slightly higher than the recommended values for urban areas (1.6±0.2, Petit et al., 2015;Stavroulas et al., 2019;Turpin and Lim, 2001). It is possible that the chosen sampling periods for the comparison (spring and summer 2017) bias high the OM-to-OC value as the photochemical activity and thus atmospheric aging are expected to increase. Ratio up to 2.2 have been observed when a significant fraction of particulate matter is made of aged aerosol (Aiken et al., 2008;Minguillón et al., 2011). 215 BC measurements from AE33 are compared with EC offline filters and an excellent agreement is found with R 2 =0.87. The slope of 1.52 relates the difference in measurements properties between EC (thermic) and BC (absorption). The difference between the measured BC and EC could be attributed to the variability of the mass absorption coefficient (MAC) value used to convert the absorbance to BC mass concentrations in the AE33 instrument. This value could be influenced by light-absorbing OC like brown carbon from biomass burning. Here the slope is in agreement with values from other studies: 1.14 to 2.13 220 through one year at Fresno supersite (Park et al., 2006); 1.62 and 1.92 using EUSAAR2 and NIOSH870 protocols, respectively, between January 2015 and July 2016 at the Environment-Climate Observatory of Lecce (Merico et al., 2019).
Finally the sum of ACSM species and BC mass concentrations were compared to the estimated mass using SMPS volume during the two deployment periods of this instrument (from 23 June to 12 August 2017 and from 6 November 2017 to 11 January 2018). For the SMPS mass conversion, a density (dcalc) was estimated taking into account the chemical composition 225 of PM1 with respective densities of 1.2 g cm -3 for organic matter, 1.75 g cm -3 for nitrate, sulfate and ammonium, 1.52 g cm -3 for chloride (Cross et al., 2007). The organic aerosol density can increase whether there are high contributions of carboxylic/dicarboxylic acids (1.46±0.23 g cm -3 ) and/or polycyclic aromatic hydrocarbons (1.28±0.12 g cm -3 ). In contrast this density would decrease with high contributions of n-alkanes (0.79±0.01 g cm -3 ) and/or n-alkanoic acids (0.89±0.07 g cm -3 ) (Turpin and Lim, 2001). A default density of 1.77 g cm -3 is applied for BC as recommended by Poulain et al. (2014). The 230 calculated density is based on the following equation (Salcedo et al., 2006): where the time-dependant mass fractions are based on ACSM and AE33 measurements. The density was found to range 235 between 1.24 and 1.77, with an average value of 1.43. The scatter plot of ACSM+BC concentrations vs. PM1 concentrations from SMPS in Figure 2a shows strong correlation (R 2 =0.81) and slope close to unity (slope=1.02 and intercept=0.46 for an orthogonal regression). Reconstituted mass (ACSM+AE33) was also compared to PM1 mass measurements from FIDAS for the last 3 months of the database (19 February to 13 April 2018) in Figure S3. Again satisfactory results are displayed with R 2 =0.89 and slope=0.96 (here the intercept is automatically set to 0) and show the consistency of the different measurements. 240 The linear regression analysis of reconstituted PM1 vs PM2.5 yielded a slope of 0.88-0.89 for a confident interval of 99% and R 2 =0.77 ( Figure 2b) so we can assume that PM2.5 concentrations are mainly composed of submicron particles.

Seasonal patterns
Time series and seasonal contributions of ACSM components, Black Carbon and daily-PM2.5 are shown in Figure 3. A 245 summary of the seasonal statistics (average and standard deviation) is reported in table 1. The averaged concentration was 9.9 µg m -3 and 12.0 µg m -3 for PM1 and PM2.5, respectively. The mass concentrations measured by the BAM 1020 show a mass exceedance for this period at Marseille according to WHO recommendation (10 µg m -3 ). The highest concentrations for PM1 were measured during winter with 11.9 µg m -3 and the lowest in spring with 8.1 µg m -3 . Several peaks reached 50 µg m -3 especially during cold periods in winter and autumn. The averaged PM1 chemical composition over the period is dominated 250 by organics, (49.7%), and BC (17.1%) while sulfate accounts for 14.6%, nitrate for 10.2%, ammonium for 7.9% and chloride for 0.5% only. Past long-term studies on daily PM2.5 filters in MRS-LCP described similar yearly trend for those species https://doi.org/10.5194/acp-2020-1015 Preprint. Discussion started: 22 October 2020 c Author(s) 2020. CC BY 4.0 License. (Bozzetti et al., 2017;Salameh et al., 2015): 44-46% for OM, 13% for BC (as EC measurements were performed during these studies, the value was multiplied by the BC/EC slope found in the present work to trace back to BC), 11% for sulfate, 8% for nitrate, 7% for ammonium and less than 1% for chloride over the August 2011 to July 2012 period. This submicron aerosol 255 composition is comparable to other Mediterranean coastal cities, where a noticeable proportion of sulfate is generally observed (Minguillón et al., 2015;Salameh et al., 2015;Stavroulas et al., 2019), whereas most urban sites in northern/central Europe show higher nitrate contribution (Lianou et al., 2011;Petit et al., 2015;Young et al., 2015). This is especially true in summer, where the highest contribution of sulfate to the total mass of PM1 is observed (24.1%). The present study confirms what had been previously reported during a short time summer measurement campaign with a C-ToF-AMS (El Haddad et al., 2013). El 260 Haddad et al. (2013) noticed that elevated sulfate periods corresponded to air masses transported from the Mediterranean Sea.
The origin of sulfate aerosol is further examined in section 3.3.2.
The average carbonaceous fraction (OA+BC) contributes to 66.8% of the PM1 mass. By converting BC to EC with the previous slope of 1.52 (see section 3.1.) the carbonaceous fraction would be 64.8% which is high compared to the OM+EC contributions 265 of PM2.5 from several urban sites in Europe (26 to 47%; Putaud et al., 2010). The OA dominance in every season is a common feature observed for European urban areas (Zhang et al., 2007b). Highest averaged values of 6.2 µg m -3 and 5.1 µg m -3 are observed during winter and autumn months respectively ( Figure S5). This trend was also observed in the 2011-2012 period but higher concentrations were found, with 12.1 µg m -3 in winter and 11.1 µg m -3 in autumn (Bozzetti et al., 2017). The year 2011 showed particularly large amount of days with high level of PM2.5 in the autumn-winter period (43 days exceeding WHO 270 PM2.5 threshold value; table 2) which may explain this difference. BC seasonal cycle has a similar pattern with highest values of 2.1 µg m -3 in winter and 1.9 µg m -3 in autumn. These high carbonaceous concentrations during the cold months are expected considering i) a reduction of the planetary boundary layer (PBL) height resulting in the accumulation of pollutants at ground level compared to other seasons and ii) increasing emissions from residential heating and open combustion of green wastes (Bozzetti et al., 2017). The deconvolution of BC into two contributions, fossil fuel and wood burning (respectively BCFF and 275 BCWB) was carried out using the aethalometer model (Sandradewi et al., 2008). As a first step the procedure recommended by Zotter et al. (2017) (i.e. to use the 470 and 950 nm wavelengths with an Angström exponent of 1.68 and 0.9 for pure wood burning and fossil fuel respectively) was applied. Using the suggested values led to unrealistic high BCWB contributions in the summer (18%) when biomass burning is expected to be negligible during the hot period. It is hypothesised, as previously suggested by Titos et al. (2017), that a fraction of BCFF was wrongly attributed to wood burning as a consequence of a failure 280 of the model to reconstruct sources when the biomass burning fraction is very low. This potential bias was investigated on fossil fuel-derived PM1 from a urban traffic site (station "Kaddouz", location: 43°34'49.8" N;5°37'49.3" E) during summer time. This kerbside site is located at the portal of a tunnel in the surrounding area of Marseille. In order to inspect the different combinations of Angström exponent for fossil fuel and wood burning (αFF and αWB, respectively) a sensitivity test was performed by scanning combination changes in a αFF range of 0.9-1.1 and a αWB range 1.6-2 with a step size of 0.01.The set of 285 combinations was evaluated and optimized based on the BCWB diurnal cycles, which were categorized according to a k-means in France (ADEME, 2020). Additionally, several wood-fired heating plants located in Aubagne, Gardanne and Aix-en-Provence (15, 18 and 25 km from Marseille, respectively) might contribute to biomass burning emissions. Besides wood heating, the open burning is expected to contribute to the increasing OA+BCWB concentrations, particularly in autumn. This includes the agricultural burning, since one third of the department surface area is dedicated to agricultural activities, but also the green waste burning. While this latter practice is prohibited, it is suspected to still be used nowadays. Summer BCWB in 300 MRS-LCP represents the lowest contribution with 7%, which is in the same range than in Fos-sur-Mer (2%; Bonvalot et al., 2019) and another big Mediterranean coastal city like Athens (6%; Diapouli et al., 2017).
The BCFF is still the main contributor to ambient black carbon (72-93% of total BC) with values ranging from 1.1 to 1.7 µg m -3 throughout the seasons. The elevated BC concentrations could be explained by the proximity of the monitoring station to the city center and thus to local urban emissions. Vehicular emissions highly contribute to primary carbonaceous fraction in 305 Marseille as demonstrated by El Haddad et al. (2011b) and can be a significant source for OA and BCFF. Even if BCFF is assumed to be an excellent marker of vehicular emissions (Herich et al., 2011), in Marseille other sources as oil-fired boilers, industrial and shipping activities could contribute to total BCFF. Nitrate exhibits maximum contribution in winter and early spring (13.2% and 14.0%, with 1.58 and 1.13 µg m -3 respectively) and minimum contribution during summer (2.6% with 0.24 µg m -3 ).This feature has been already observed in other European 310 cities (Minguillón et al., 2015;Petit et al., 2015;Reyes-Villegas et al., 2016;Young et al., 2015). Bozzetti et al. (2017) observed lower fraction in spring possibly because of the volatilization of nitrate from the filters surface in warm conditions. Several episodes were reported mostly in late winter/early spring (described by elevated nitrate dispersion for February and March in Figure S5) and concentrations of other species also increased during these nitrate events conducting to highly polluted days.
These episodes are investigated in section 3.3.1. Nitrate was further categorised into inorganic and organic fraction using the 315 method described by Farmer et al. (2010). The detailed methodology is presented in SI.
Seasonal mass fraction of NO3,Org and NO3,Inorg are displayed in Figure 4. NO3,Org average concentration ranges from 0.09 µg m -3 in summer to 0.26 µg m -3 in winter. The resulting average NO3,Org fraction for the whole dataset is 20±7%. The error here is determined from error propagation calculations described by Farmer et al. (2010) and is detailed in SI. This fraction is in the range of reported values for European urban sites (28%, Mohr et al., 2012;24%, Saarikoski et al., 2012). The highest NO3,Org 320 contribution happens to be in summer (38%) when the total nitrate concentration is at its lowest level. NO3,Org is produced from volatile organic compounds (VOCs) oxidations by nitrate radicals and photochemical oxidation with NOx. Biogenic VOCs emissions are high in summer in Mediterranean area (Parra et al., 2004;Steinbrecher et al., 2009) and provide an important potential source for organic nitrate formation.
NO3,Org takes into account only the nitrate functionality of organic nitrates. Assuming a molecular weight between 200 and 300 325 g mol -1 for particle organic nitrates, contribution of organic nitrates to total OA can be estimated (Xu et al., 2015). Results presented in Table S4 give a contribution to total OA of about 6-10% in summer, 11-17% in autumn, 14-20% in winter and 18-28% in spring. Highest contributions in springtime can be due to increased level of biogenic VOCs coupled with favourable meteorological conditions for partitioning. These results suggest that despite the low contribution of NO3,Org to total nitrate, organic nitrates can be an significant fraction of OA in Marseille. 330 The NO3,Inorg average seasonal concentrations range from 0.15 µg m -3 in summer to 1.31 µg m -3 in winter. Here NO3,Inorg was assumed to be mostly ammonium nitrate particles. Ammonium nitrate is semi volatile and its gas/particle partitioning is affected by temperature and relative humidity changes (Stelson and Seinfeld, 1982), which lead to enhanced particle partitioning during winter and higher evaporation during summer (Huffman et al., 2009). This is supported by Figure S5 where lowest temperature and highest relative humidity values are found in winter and early spring, while the inverse trend is 335 illustrated for summer months.
Winter and springtime ammonium concentrations can be driven by the greater availability of ammonia from agricultural activity and waste management. Moreover Suarez-Bertoa et al. (2015) mentioned that urban traffic emissions of ammonia have increased in Europe (+378%) over the last decades leading to possible enhanced ammonium concentrations. High concentrations of inorganic nitrate are related to elevated levels of NOx in winter. This is described in Figure 1 where the 340 hourly maximum concentrations of NOx emissions mostly appear in winter seasons.
It is well known that ultrafine particles (UFPs; diameter <100nm) do not affect the mass concentrations but might be relevant for health-related issues, and their long-term variability is explored in this study. The total average UFPs number concentration (20-100 nm) measured with a TSI 3031 monitor over the full study was 7765 cm -3 . These values are similar to those observed 345 in Barcelona but are higher than those found in Prague, Madrid and Rome (Borsós et al., 2012;Brines et al., 2015). In Marseille, UFPs represented at 85% of total submicron particle number, in agreement with previous observations in urban environment (Rodríguez and Cuevas, 2007;Wehner and Wiedensohler, 2003). UFPs average concentration were slightly higher in winter and autumn with average values of 8600 cm -3 and 8100 cm -3 , respectively, while an average value of 7500 cm -3 was found in spring and summer. The seasonal variation followed the general patterns observed for BC and OA concentrations over the 350 year, suggesting that most particles in number arise from sources of combustion. However, when considering size distribution measurements down to 10-nm, SMPS data revealed the occurrence of sharp UFPs events more frequently in the summer than in winter, with 10 events exceeding 50 000 particles cm -3 in summer against only 1 in winter (Figure 3). To get insights into the sources and processes contributing to ultrafine particles in urban ambient air, number concentrations and BC were investigated using the methodology developed by Rodríguez and Cuevas (2007). This methodology has been extensively 355 applied in urban environment to apportion the number concentration of primary and secondary sources (del Águila et al., 2018;González et al., 2011;Hama et al., 2017a;Hama et al., 2017b;Reche et al., 2011;Rodríguez and Cuevas, 2007;Tobías et al., 2018). To refine the method, BCFF was used instead of total BC to better apportion primary traffic emissions. The total measured UFPs number concentration (N) can be splitted in two components: where N1 accounts for fresh primary emissions of vehicle exhaust, directly emitted in the particle phase or nucleating immediately after emission (Arnold et al., 2006;Burtscher, 2005;Kittelson, 1998) and N2 accounts mostly for secondary 365 particles formed in the atmosphere during the dilution and cooling of the exhaust emissions. S1 is the slope estimated using best-fit line to the points aligned in the lower edge of N vs BCFF scatter plot (see Figure S7 for details). S1 was calculated for each season by fitting data below the 10 th percentile of N/BCFF. The N1 and N2 fractions were derived from the TSI 3031 measurements and average number concentrations during each season are reported in table 1 and Figure 4.
The N2 fraction was predominant with number concentrations 1.02 to 1.7 times higher than N1. The N1 seasonal trend slightly 370 varied through the measurement period with a contribution between 37 and 50%. These emissions appear to be dominant in winter and autumn, with average concentrations of ≈ 4200 and 3800 cm -3 . An explanation for these results could be either the higher traffic rate or the low temperature influencing the soot particle formation during combustion in these seasons. In contrast, the N2 average particle numbers were higher in spring and summer as the concentrations reached its highest value (≈ 4300 cm -3 and 4800 cm -3 ). This behaviour will be further addressed in the following section. 375 Figure 5 shows the average diurnal profiles of OA, NH4 + , NO3 -, SO4 2-, BCFF and BCWB across the different seasons. Distinctive diurnal patterns are found for carbonaceous aerosols. First, a clear traffic-related diurnal profile is observed for BCFF with a morning peak and an evening peak starting at the typical rush hours (04:00 UTC and 16:00 UTC respectively). The amplitude of the traffic-related diurnal cycle seemed to be affected by meteorological conditions since it varied within the year, with 380 maximum min to max amplitude of 0.9 during winter and minimum min to max amplitude of 0.5 during summer. The BCWB had a significantly different diurnal cycle with a typical increase starting at 17:00 UTC and a maximum level of 1.1 µg m -3 at night-time.

Diurnal profiles
The mean to max amplitude increased steadily from spring, autumn to winter (0.1, 0.2 and 0.4, respectively), following the increased heating demand. Under specific meteorological conditions (no rain, low wind speed, low boundary layer) this source 385 led to the highest levels of PM1 episodes in Marseille. This is discussed in detail in Sect. 3.3.1. In winter and autumn, OA diurnal cycle mainly resulted from the superposition of both traffic-related and wood burning-related cycles with maxima of 4.9 µg m -3 and 7.9 µg m -3 for the morning and evening peak, respectively. In spring and summer an additional local maximum appeared during mid-day (3.8 µg m -3 ). While this peak may partly be related to a distinct local source like cooking emissions (Bozzetti et al., 2017), formation of secondary organic aerosol is also expected, as observed by El Haddad et al. (2013). 390 The diurnal variations of NO3,Org are represented for each season in Figure 5. Concentrations increased after sunset and were higher at night, likely due to the oxidation of VOCs by the nitrate radical (Kiendler-Scharr et al., 2016). This trend was more pronounced in summertime with enhanced biogenic VOCs emissions. Summer NO3,Org profile was investigated for June 2017 when concentrations were above the detection limits. Slight daily enhancements of NO3,Org were found in the morning (around 08:00 UTC) and are attributed to photooxidation of VOCs in the presence of high nitrogen monoxide NO concentrations (the 395 diurnal profiles are also reported in Figure 5 and show maxima at 07:00 UTC) as mentioned by Xu et al. (2015). Similarly, NO3,Inorg diurnal pattern suggests fast formation mechanism from local NOx emissions. Concentrations are higher during nighttime when the condensation to the particle phase is favoured by meteorological conditions (low temperature and high relative humidity).

400
Sulfate and ammonium exhibit very similar profiles, except for winter and early spring when ammonium is mainly associated to nitrate. The quite flat diurnal profiles of ammonium sulfate during autumn, winter, and spring might be due to the regional character of this component (regional transport) (Seinfeld and Pandis, 2016).
In summer however, ammonium sulfate diurnal profile shows a clear increase during the daylight period with maxima reached at noon, possibly due to local photochemical production of sulfate from its precursor SO2 emitted by nearby shipping and 405 industrial activities and advected to the city as the sea breeze sets in. Some studies suggest fast SO2 to sulfate conversion in the exhaust plumes (Healy et al., 2009;Lack et al., 2009). This local sulfate fraction will be discussed in detail in Section 3.3.2.
Average daily profiles for N1 and N2 are represented in Figure 5 and show some discrepancies. Particle number N1 exhibits maxima during morning and evening traffic rush hours similarly to NO concentrations, when UFPs are mainly associated to vehicle exhaust emissions. The N2 daily profiles display different diurnal patterns: in winter N2 follows the traffic bimodal 410 profile with 1-2 hours shift (morning peak at starting at 06:00 UTC and evening at 17:00 UTC) which can highlight the fast homogeneous dilution/cooling (favoured by the low temperature and high relative humidity conditions) and mixing of the vehicle exhaust in the ambient air (Casati et al., 2007;Charron and Harrison, 2003). In summer, N2 exhibits a daily evolution with a broad maximum during daylight. This trend closely follows the daily evolution of temperature, ozone and SO4 2concentration suggesting a photo-oxidative process of gaseous precursors (Woo et al., 2001), as SO2, combined with more 415 dilution of pollutants when the boundary layer increases (Reche et al., 2011). For the transitional seasons (autumn and spring) the patterns reveal some mixing between the cars exhaust cooling and photochemistry states and N2 can be attributed to the two processes.
In order to investigate the UFPs apportionment below 20 nm the methodology was applied to SMPS data during the available summer period for a range between 10 and 20 nm. The N2 (10-20 nm) number concentration, corresponding to 90% of the total 420 number in this range, showed the same trend of SO2 diurnal evolution ( Figure 5). The SO2 is considered a gas tracer for industrial and shipping activity and it was advected to the monitoring station from the morning. Furthermore, both SO2 and N2 (10-20 nm) concentrations increased suggesting the nucleation of sulfuric acid particles (Burtscher, 2005;González et al., 2011). Overall the UFPs investigation demonstrates that secondary particle formation is an important contributor to particle number in Marseille and besides road traffic, there is some high influence of industrial/shipping mixed sources. 425 Table 2 Figure 6 shows the time series of PM1 chemical composition and the meteorological parameters (i.e. temperature, relative 440 humidity, wind direction, PBL height and precipitations) during these 2 events. An assessment of transport dynamics of aerosols was carried out using the BC/SO4 2ratio as suggested by Petit et al. (2015). The sulfate fraction (SO4 2-) is considered a good tracer for long-range transport, whereas BC refers to local influence from the city and the surrounding area. The use of this local/long-range proxy requires the assumption of minor local source of SO2 and no direct SO4 2formation in plumes at a local scale during the 2 episodes. Therefore the industrial and shipping contributions to global pollution have to be low for 445 these periods. This is true except for a short period on the 23 rd of December around 12:00 UTC when a temporary increase of SO4 2decreases the BC/SO4 2ratio. This short event possibly originated from industrial and/or shipping emissions as SO2 and UFPs concentrations also increased. Additionally, aged air masses might contain BC particles (Laborde et al., 2013) and this interference must be considered during the BC/SO4 2ratio analysis.
Carbonaceous concentrations increased mostly in the evening by a factor of 4 to 5 in the space of a few minutes. The temporal evolution of BC contributions (Figure 6a) clearly indicated a predominance of BCWB during these nights with a contribution reaching sometimes 100%. At night PM size distribution ranged between 70 and 200 nm, typical of wood burning emissions 455 (Coudray et al., 2009). High BC/SO4 2ratio (average of 6.12 with values up to 20) and decreasing PBL height (470 m) suggested a strong local influence. Figure 7 displays NWR analysis plot for OA and BCWB and show evidence of high concentrations under North-East land breeze. This wind analysis is also performed on the fraction of 2 4 2 + organic ion (f60) which is a pure tracer of levoglucosan fragmentation (Alfarra et al., 2007) and thus of biomass burning emissions. The results displayed a similar hotspot than OA and BCWB concentrations. 460 This configuration is very frequent in Marseille, as land breeze prevailed for around 25% of time, as shown in Figure S8.
Under such meteorological conditions, the average annual PM1 concentration was 14.3 µg m -3 instead of 8.49 µg m -3 for the remaining period. These winds transport to the city anthropogenic emissions from the surrounding suburban residential areas of Marseille, in the north-easterly direction. In winter, these areas are a few degrees colder than the city of Marseille, resulting in an increased use of firewood as an auxiliary heating source. After the 25th December, the PM1 levels dropped down as the 465 wind direction shifted from North-East to East/South-East. The PM1 decrease was also favoured by the vertical dilution that occurred with increased PBL height combined with precipitations.
The February long-range event took place during a period of four consecutive PM2.5 exceedance days from 22 nd to 25 th February 2017 (Figure 6b). The average PM1 concentration was 31.2 µg m -3 and the aerosol chemical composition was stable with OA 470 contribution of 41%, followed by NO3 -(25%), BC (12%), NH4 + (11%), and SO4 2-(10%). The mass concentrations were linked to particles distributed between 200 and 1000 nm, which is in line with size distribution in the accumulation mode of ammonium nitrate, ammonium sulfate and oxidized organic species (Canagaratna et al., 2007). This event is characterised by a low BC/SO4 2ratio (average of 1.13) suggesting advection of secondary pollution, dominated by OA and ammonium nitrate.
The BC/SO4 2ratio can reach values of 4-5 when regional background concentrations are associated with occasional increase 475 of local BC emissions, contributing to enhanced particle level. The geographical origin of some species was then inspected.
As sustained winds come from the same direction for most of the event, wind data spikes with high standard deviation must be down-weighted to conduct a reasonable wind analysis. Instead of NWR, SWIM-2 analyses are used as they allow to use a weighting term. Figure 7 shows SWIM-2 analysis plot for NO3,Inorg (assuming here as ammonium nitrate). The main dominant wind sector seemed to be from North to South-West direction for this species with highest concentration from North/North-480 West at high wind speed and South/South-West at low wind speed. This result likely shows a combined local pollution with medium/long-range transport of secondary species during this time. In order to avoid local influences and highlight the longrange emissions, wind regression analysis are conducted on NO3,Inorg normalized by BC. The resulting plot clearly shows higher values for the North/North-West sector attributed to strong Mistral blow (wind speed >1.7 m s -1 ) conditions. SWIM-2 analysis was also conducted on the 2 + fraction (f44) which is, at high level, specific of oxygenated organic aerosol (OOA). Again 485 the mistral blow conditions are well distinguished suggesting a high fraction of secondary organic aerosol during the event.
North/North-Westerly winds bring into the city polluted air masses (Figure 6b and Figure 7) from central Europe that might pass also over the Pô Valley (North of Italy) as can be observed from the 72h-backtrajectories (n=32; displayed every 3 hours) displayed in Figure S9. The Pô Valley is well known for its high levels of inorganic (Schaap et al., 2004b;Squizzato et al., 2013;Diémoz et al., 2019) and secondary organic aerosol (OOA) (Saarikoski et al., 2012). The plain is enclosed by the Alpine 490 chain and the Apennines limiting the dispersion of pollutants and thus leading to frequent pollution events. During the winter/spring period low temperature and high humidity may favour ammonium nitrate particles formation (Schaap et al., 2004b). Moreover, intense agricultural spraying occurring in early spring might enhance NPF. Once the air masses cross the Alps they are then channelled along the Rhone Valley corridor toward the Mediterranean Sea. The low altitude of the backtrajectories (<500m), when passing by the Rhone Valley, led to the accumulation of pollutants along the air masses 495 trajectories. A strong East/South-Est wind combined with higher temperature and higher PBL height led to the dilution of atmospheric pollutants and the return to normal conditions. The backtrajectories of a similar event occurring from the 14 th to the 17 th of March are represented in Figure S9. The chemical composition was similar to that reported in the previous event (OA = 44%, NO3 -= 26%, NH4 + = 11%, BC = 9%, SO4 2-= 9%).
Air masses were transported from the North direction through North-Estern part of France, Switzerland and North-Western 500 part of Italy. Polluted air masses crossed continental regions known to be hotspots of ammonia emissions (the French Champagne-Ardennes region, the Swiss plateau, the Pô valley, as documented by Viatte et al., 2019), explaining partly the enhanced ammonium nitrate contribution during the winter/spring long-range events.
In term of frequency of occurrence, 40% of exceedance days account for local origin (6 days) and 60% for long-range transport 505 influence (9 days). Globally, the combined observations of phenomenology, chemical composition and meteorology allow to accurately analyse events with exceedance mass concentrations for fine particles at MRS-LCP supersite, highlighting variable situations in an urban area affected by different air mass origins.

Sulfate origin and Shipping/industrial plumes in summer
The summertime aerosol contributes little to the exceedance of PM2.5 air quality thresholds (Figure 1). This season accounts 510 for only 10% of the exceedance days since 2008 and none of these occurred in 2017-2018. Still, the high solar radiation and temperature combined to the dry condition are expected to enhance the formation of secondary pollutants and their accumulation. High ozone levels often exceeding the WHO threshold are indeed recorded in Marseille (Figure 1). These conditions also affect the aerosol composition as it results in an enhanced secondary sulfate fraction (24% of PM1 on average and a maximum contribution of 66%). As shown by the NWR plot in Figure 7, high concentrations of sulfate are associated to 515 a rather broad range of wind sectors and speeds, as expected for potentially aged and processes aerosol. While the geographic area where high sulfate concentrations are observed extend between the South-West and the North-East sectors, there is still a predominance from the South-West sector. In this sector, prevalence of high concentrations of SO2, the major precursor of sulfate aerosol, is also clearly observed. These SO2 concentrations are associated with high UFPs plumes. Figure 8 shows the averaged UFPs number concentrations and the seasonal frequency occurrence of peaks concentration according to SO2 520 concentrations classes. UFPs number increases with the enhanced SO2 levels, especially in the summer which involves more than 55% of the highest SO2 concentrations (>20µg/m 3 ). Local sources of SO2 and UFPs include not only the large petrochemical and industrial area of Fos-Berre, located 40 km northwest, but also the shipping traffic related to these activities in the gulf of Fos. Fresher emission of SO2 can originate from the port of Marseille, located 3 km away from the site. During summertime, the ships traffic increases by 25% (4319 against 3262 for the 2017-2018 period) partly because of the enhanced 525 numbers of passenger ferries and travel cruises during the holiday season. Figure 9 shows the diurnal trend of cumulative number of ship movements for the summer 2017. This number was differentiated according to the type of movement and the basin location (the exact geographic positions are reported in the Supplement in Figure S1). As can be seen by the diurnal profile of the ship movements in the harbour, two maxima were observed: the first at 05:00 UTC due to ship arrivals and the second at 17:00 UTC due to ship departures. The highest SO2 530 peak might be related to the ship arrivals increase; however, El Haddad et al. (2011aHaddad et al. ( , 2013 assigned similar morning SO2 plumes to industrial activity from Fos-sur-mer in summer. Thus, the first SO2 increase could arise from a combination of contribution from ships arrival and industrial air masses, whereas the second peak could be mostly linked to the late afternoon boats departure. SO4 2concentrations increased during the day and could be also partly affiliated to the direct influence of shipping/industrial activity to the monitoring site. 535 To further investigate the large SO4 2concentrations in summer and their origin at a broader scale, a cluster analysis was carried out on the air masses reaching the site. From the analysis performed on the 72h-backtrajectories, 3 distinct clusters were assessed from the total spatial variance (TSV) variation. Cluster 1 (Mediterranean origin) is related to air masses that circulate through western Mediterranean basin before arriving at MRS-LCP site. Cluster 2 (sea breeze) corresponds to an initial low mistral blowing along the Rhone Valley seaward that returns to land the next day. Cluster 3, similarly to cluster 2, is 540 representative of Mistral wind from the Rhone Valley but with higher speeds, as recorded at MRS-LCP (average of 1.01 m s -1 against 0.60 and 0.72 m s -1 for cluster 1 and 2, respectively). Mean calculated trajectories are displayed in Figure S10. Cluster 3 associated to low pollutant levels won't be investigated in this study as its frequency is low (19%). In comparison, Mediterranean and sea breezes air masses account for 43% and 38%, respectively. While the clustering analysis clearly identifies the Mediterranean long range trajectories, Figure S11 shows that they still get mixed with the sea breeze when they 545 approach the shore (as indicated by the wind sector 190°-270° characteristic of the sea breeze, and the by the sharp SO2 peaks included in the Mediterranean regime periods in pink). Still, discernible differences in the particle distribution and in the chemical composition are observed, as shown by Figure 10a and Figure 10b. Sea breeze cluster had a pronounced nucleation mode and Aitken mode (>10.25 nm according to SMPS measurements) (Figure 10a). Mediterranean origin cluster has a broaden size distribution with combined Aitken and accumulation modes. Mallet et al. (2019) found similar results at 550 Lampedusa site for air masses passing over the western Mediterranean. Particles from the sea breeze cluster might be smaller than Mediterranean origin cluster with the likelihood of fresher emission from the nearby shipping/industrial sources.
Moreover the Box plots of SO4 2-, SO2 and N2(10-20 nm) concentrations related to cluster 1 and 2 reveal that slightly higher SO2 (3.1 vs 2.6 µg m -3 ) and N2(10-20 nm) (7855 vs 5740 cm -3 ) concentrations are encountered with cluster 2, while higher SO4 2concentrations are observed with cluster 1 (3.2 µg m -3 against 1.9 µg m -3 for sea breeze). 555 In an effort to further investigate the characteristics of the sulfate constituents, the total SO4 2was first tentatively deconvolved into ammonium sulfate, organosulfate and MSA (methanesulfonic acid) following the methodology developed by Chen et al. (2019) and based on HSO3 + (f81) and H2SO4 + (f98) ion fractions from AMS measurements. The resulting fH2SO4 + vs fHSO3 + data points are displayed in Figure 12 and are color-coded according to the different air masses (Sea breeze and Mediterranean origin). The results show that ammonium sulfate was the dominant source for ToF-ACSM sulfate signals during this period. 560 This was confirmed by the NH4 measured/NH4 predicted ratio close to 1 ( Figure S4), showing that SO4 2was always fully neutralized by NH4 + at MRS-LCP site. It is hypothesised that even local emissions have enough time to mix with ammonia from urban environment to fully neutralize before reaching the station: considering a minimum distance between the harbour and MRS-LCP of 3 km and an average sea breeze wind speed between 1.05 m s -1 and 2.7 m s -1 (measured from a harbour meteo station under the same sea breeze blowing direction) it takes at least around 20-45 minutes for a plume to reach the station. This agrees 565 with Celik et al. (2020) who observed that shipping plumes older than 40 min were mostly neutralized with the ambient NH3 from a cleaner environment.
In order to trace the geographical origin of SO4 2from MRS-LCP at a regional scale, CWT analysis were performed and the results are shown in Figure 11. A weighting function has been implemented to avoid artefacts linked to high concentrations with low number of trajectories passing through a particular cell. The aim is to avoid local sulfate influence from the transport 570 model. Following Waked et al. (2014) recommendations a discrete function based on back trajectory density (log10(n+1)) was applied (using the ZeFir tool on Igor). CWT on SO4 2from cluster 1 (Figure 11a) pointed out the combined influence of low Mistral advection from the Rhone Valley and the switch into South-Westerly thermic breeze, in agreement with the expectations. CWT for cluster 2 (Figure 11b) exhibited long-range transport, with SO4 2concentrations associated to the south and western Mediterranean air mass circulation. In the basis of this, several sources can be related to sulfate emissions in the 575 Mediterranean Sea such as shipping activity, marine biogenic or crustal origin (Becagli et al., 2012).

Summary and conclusions
The chemical composition of submicron aerosols was monitored in real time between 1 February 2017 and 13 April 2018 at an urban background site of the Mediterranean city of Marseille. Measurements were carried out with a ToF-ACSM associated with a suite of collocated instruments including an aethalometer, an ultrafine particle monitor, a SMPS and monitors for 580 regulated pollutants (PM, NOx, O3, SO2).
The reconstituted PM1 mass (ACSM measurements + BC) was cross validated through several comparisons with external parameters. ACSM+BC concentrations were found to be in good agreement with estimated mass concentrations from SMPS (R²=0.81 and slope=1.02) and PM1 concentrations from FIDAS (R 2 =0.89 and slope=0.96).
OA was the most abundant specie of submicron aerosol, with an annual average of 49.7% and the carbonaceous fraction was 585 dominant for every season (66.8%) and especially during cold months. BC contributes largely to this fraction (17.1% of total submicron aerosol) and is mainly dominated by fossil fuel emissions as determined with the AE-33 aethalometer model. BC from wood burning emissions showed higher contribution during winter and very low contribution in summer, as expected.
The organic nitrate contribution was evaluated using the NO2 + /NO + ratio method and gave reasonable results in separating NO3,Org and NO3,Inorg concentrations. Some uncertainties still remain as RON was set to a fixed value and could slightly vary 590 according to the VOC precursors which lead to particle organic nitrate formation. Also, low nitrate signal provide instable Robs and could enhance the uncertainty of the estimation. The NO3,Org fraction was 20±7% for the total nitrate (representing 10.2% on average for the PM1 concentrations) during the entire period and did not significantly contribute to enhanced polluted events with high PM1 concentrations. However organic nitrate contribution to total OA could be estimated to significant values with maximum during springtime (18-28%). 595 Particle number concentration was successfully segregated into two components (N1 and N2) by using the minimum slope found in the N vs BCFF plot, with BCFF accounting for primary particles. N1 was attributed to fresh primary traffic emissions and N2 to secondary particles. The secondary N2 fraction was predominant with number concentrations 1.02 to 1.7 times higher than N1. While N1 showed clear maxima during morning and evening traffic rush hours, N2 was either attributed to dilution/cooling and mixing of vehicular exhausts in the atmosphere during cold seasons or to photooxidation products of 600 gaseous precursors during hot seasons. These results revealed the importance of secondary particle formation and contrasted seasonal sources of UFPs number. PM1 pollution events were determined according to the daily concentrations exceeded WHO recommendations. To illustrate their differences in chemical composition, meteorological dynamic and geographical origins two events (23-24 December 2017 and 22-25 February 2018) were carefully examined. BC/SO4 2ratio, non-parametric wind regressions and back 605 trajectories provided important information to discriminate local and long-range transport contributions. The local contribution during exceedance days is attributed to an increase of biomass burning emissions with domestic heating and green wastes burning, cumulated with intense traffic. In those situations, the OA and BC concentrations strongly increase at night in the space of few minutes when the nocturnal land breeze set up and the boundary layer height decreases. The long-range pollution case, led to high increase of secondary aerosol, more precisely ammonium nitrate and oxygenated OA, transported from the 610 central Europe and notably Pô Valley to the city. The investigation of these two episodes highlighted that local influences are mainly responsible of continuous background pollution and its mix with long-range transport events can trigger to situation with high exceedance levels of fine particles. This elevated pollution occurred mostly in winter and early spring with favourable wind conditions. Even if no exceedance day was found during the summer season, the chemical composition is slightly different, with higher sulfate contributions and intense UFPs plumes (mostly between 10 and 20 nm) associated with SO2 are 615 advected on site. Air mass clustering has been performed to explain the observed differences in aerosol composition and highlighted the presence of local and regional emissions of shipping activity, mixed with industrial plumes in Marseille. This configuration is expected in a coastal city with a consequent harbour and in the vicinity of an industrial area. First sea breeze blowing results in a local advection of SO2, UFPs and SO4 2from the industrial/shipping plumes. Then regional Mediterranean air masses can bring higher sulfate concentrations and larger particles from aged shipping plumes, probably mixed with other 620 sources such as marine biogenic or crustal. For both cases it has been shown that sulfate was completely and rapidly neutralized suggesting that ammonium sulfate predominates in the Mediterranean Sea.
In conclusion, the supersite MRS-LCP successfully recorded long-term observations and seasonality of fine particles. The long-term real-time monitoring in MRS-LCP showed a great potential and will supply direct information to public authorities and citizens. It may provide better understanding of pollution episodes and more effective control of local mitigation within 625 the context of French Atmosphere Protection Plan.
Data availability. Data are available upon request to the contact author Benjamin Chazeau (benjamin.chazeau@univ-amu.fr).
Author contributions. NM designed the research. BC, GG and BM contributed to the measurements. BC performed the 630 analysis and wrote the paper. NM, BD, BT and HW reviewed and all authors commented on the paper.

Competing interests.
The authors declare they have no conflict of interest.