Seasonality of Aerosol Sources Calls for Distinct Air Quality Mitigation Strategies

Abstract

An Aerosol Chemical Speciation Monitor (ACSM) was deployed to investigate the temporal variability of non-refractory particulate matter (NR-PM₁) in the coastal city of Galway, Ireland, from February to July 2016. Source apportionment of the organic aerosol (OA) was performed using the newly developed rolling PMF strategy and was compared with the conventional seasonal PMF. Primary OA (POA) factors apportioned by rolling and seasonal PMF were similar. POA factors of hydrocarbon-like OA (HOA), peat, wood, and coal were associated with domestic heating, and with an increased contribution to the OA mass in winter. Even in summer, sporadic heating events occurred with similar diurnal patterns to that in winter. Two oxygenated OA (OOA) factors were resolved, including more-oxygenated OOA and less-oxygenated OOA (i.e., MO-OOA and LO-OOA, accordingly) which were found to be the dominant OA factors during summer. On average, MO-OOA accounted for 62% of OA and was associated with long-range transport in summer. In summer, compared to rolling PMF, the conventional seasonal PMF over-estimated LO-OOA by nearly 100% while it underestimated MO-OOA by 30%. The results from this study show residential heating and long-range transport alternately dominate the submicron aerosol concentrations in this coastal city, requiring different mitigation strategies in different seasons.

Keywords: aerosol chemical speciation monitor; air pollution; rolling PMF; source apportionment; submicron aerosol.

Figure 8

Polar plot of hourly averaged MO-OOA, LO-OOA, peat (for the rolling PMF), SO₄, and NO₃ as a function of wind speed (radial axis, m s⁻¹) and wind direction during spring, summer, and winter, color-coded by respective concentrations (μg m⁻³). The polar plots were plotted using openair package with R software (version 4.1.2) [45].

Figure 1

Time series of (a) wind speed (ws) and wind direction (wd); (b) relative humidity (RH) and Temperature (T); and (c) the major NR-PM₁ species, i.e., organics (OA), sulfate (SO₄), nitrate (NO₃), ammonium (NH₄), and chloride (Chl) from 5 February 2016 to 31 July 2016. The sampling period covered the season of winter (February), spring (March, April, and May), and summer (June and July). Inset pie charts show the relative contribution of different NRPM₁ species, with values above showing the average concentration over different seasons.

Figure 2

Diurnal cycles of the NR-PM₁ species, i.e., organics (OA), sulfate (SO₄), nitrate (NO₃), ammonium (NH₄), and chloride (Chl), as well as BC over winter, spring, and summer 2016. Note that BC measurements were only available in winter and spring.

Figure 3

Averaged mass spectral profiles of the OA factors resolved by the rolling PMF strategy. Error bar represents one standard deviation.

Figure 4

Time series of the OA factors (mid panel) resolved by the rolling PMF. Also shown is the time series of the temperature (top panel) and the relative fraction of each factor (bottom panel). Two episodic pollution events with one featured with low temperature (EP1) and the other with high temperature (EP2) are highlighted, with the pie charts showing the relative fraction of the factors.

Figure 5

Averaged diurnal pattern of the OA factors from rolling PMF in winter, spring and summer.

Figure 6

Comparison of the OA factor fractions in winter, spring, and summer apportioned by the rolling PMF (a–c) and seasonal PMF (d–f).

Figure 7

Scatter plot between the time series of the OA factors resolved by the seasonal PMF (y-axis) and rolling PMF (x-axis), color coded by date. The correlation r² and slope for the linear fit are also shown for each factor.