Abstract. Elevated ambient concentrations of sub-3 nm particles (nanocluster aerosol, NCA) are generally related to atmospheric new particle formation events, usually linked with gaseous sulfuric acid (H2SO4) produced via photochemical oxidation of sulfur dioxide. According to our measurement results of H2SO4 and NCA concentrations, traffic density, and solar irradiance at an urban traffic site in Helsinki, Finland, the view of aerosol formation in traffic-influenced environments is updated by presenting two separate and independent pathways of traffic affecting the atmospheric NCA concentrations: by acting as a direct nanocluster source and by influencing the production of H2SO4. As traffic density in many areas is generally correlated with solar radiation, it is likely that the influence of traffic-related nanoclusters has been hidden in the diurnal variation and is thus underestimated because new particle formation events also follow the diurnal cycle of sunlight. Urban aerosol formation studies should, therefore, be updated to include the proposed formation mechanisms. The formation of H2SO4 in urban environments is here separated into two routes: primary H2SO4 is formed in hot vehicle exhaust and is converted rapidly to the particle phase; secondary H2SO4 results from the combined effect of emitted gaseous precursors and available solar radiation. A rough estimation demonstrates that ∼85 % of the total NCA and ∼68 % of the total H2SO4 in urban air at noontime at the measurement site are contributed by traffic, indicating the importance of traffic emissions.

Particle emissions from marine traffic affect significantly air quality in coastal areas and the climate. The particle emissions were studied from a 1.4 MW marine engine operating on low-sulfur fuels natural gas (NG; dual-fuel with diesel pilot), marine gas oil (MGO) and marine diesel oil (MDO). The emitted particles were characterized with respect to particle number (PN) emission factors, PN size distribution down to nanometer scale (1.2–414 nm), volatility, electric charge, morphology, and elemental composition. The size distribution of fresh exhaust particles was bimodal for all the fuels, the nucleation mode highly dominating the soot mode. Total PN emission factors were 2.7 × 1015–7.1 × 1015 #/kWh, the emission being the lowest with NG and the highest with MDO. Liquid fuel combustion generated 4–12 times higher soot mode particle emissions than the NG combustion, and the harbor-area-typical lower engine load (40%) caused higher total PN emissions than the higher load (85%). Nonvolatile particles consisted of nanosized fuel, and spherical lubricating oil core mode particles contained, e.g., calcium as well as agglomerated soot mode particles. Our results indicate the PN emissions from marine engines may remain relatively high regardless of fuel sulfur limits, mostly due to the nanosized particle emissions.

Shipping is the main source of anthropogenic particle emissions in large areas of the globe, influencing climate, air quality, and human health in open seas and coast lines. Here, we determined, by laboratory and on-board measurements of ship engine exhaust, fuel-specific particle number (PN) emissions for different fuels and desulfurization applied in shipping. The emission factors were compared to ship exhaust plume observations and, furthermore, exploited in the assessment of global PN emissions from shipping, utilizing the STEAM ship emission model. The results indicate that most particles in the fresh ship engine exhaust are in ultrafine particle size range. Shipping PN emissions are localized, especially close to coastal lines, but significant emissions also exist on open seas and oceans. The global annual PN produced by marine shipping was 1.2 × 1028 (±0.34 × 1028) particles in 2016, thus being of the same magnitude with total anthropogenic PN emissions in continental areas. The reduction potential of PN from shipping strongly depends on the adopted technology mix, and except wide adoption of natural gas or scrubbers, no significant decrease in global PN is expected if heavy fuel oil is mainly replaced by low sulfur residual fuels. The results imply that shipping remains as a significant source of anthropogenic PN emissions that should be considered in future climate and health impact models.

<p><strong>Abstract.</strong> Elevated ambient concentrations of sub-3 nm particles (nanocluster aerosol, NCA) are generally related to atmospheric new particle formation events, usually linked with gaseous sulfuric acid (H<sub>2</sub>SO<sub>4</sub>) produced via photochemical oxidation of sulfur dioxide. According to our measurement results of H<sub>2</sub>SO<sub>4</sub> and NCA concentrations, traffic density, and solar irradiance at an urban traffic site in Helsinki, Finland, the view of aerosol formation in traffic-influenced environments is updated by presenting two separate and independent pathways of traffic affecting the atmospheric NCA concentrations: by acting as a direct nanocluster source, and by influencing the production of H<sub>2</sub>SO<sub>4</sub>. As traffic density is frequently correlated with solar radiation, it is likely that the influence of traffic-related nanoclusters has been hidden in the diurnal variation, and is thus underestimated because new particle formation events also follow the diurnal cycle of sunlight. Urban aerosol formation studies should, therefore, be updated to include the proposed formation mechanisms. Additionally, the directly emitted NCA poses a potentially elevated health risk, highlighting the need for quantifying the source and loadings in populated areas. The formation of H<sub>2</sub>SO<sub>4</sub> in urban environments is here separated in two routes: primary H<sub>2</sub>SO<sub>4</sub> is formed in hot vehicle exhaust and is converted rapidly to particle phase; secondary H<sub>2</sub>SO<sub>4</sub> results from the combined effect of emitted gaseous precursors and available solar radiation.</p>

Figure S1. Time series of traffic density, solar irradiance, precipitation, temperature (T), relative humidity (RH), and atmospheric pressure (p).