Four years of highly time resolved measurements of elemental and organic carbon at a rural background site in Central Europe
Abstract
Elemental carbon (EC) and organic carbon (OC) in fine atmospheric aerosols (PM2.3: aerodynamic diameter smaller than 2.5 mu m) have been measured with a semi -automatic instrument during a 4 -year survey at the National Atmospheric Observatory Koletice (NAOK), Czech Republic. Ground based measurements were performed from March 2013 to December 2016 with a field Semi -Continuous OCEC Aerosol Analyzer (Sunset Laboratory Inc., USA).
The variation of EC and OC concentrations and the OC/EC ratio was characterized for different seasons and days of the week. During our survey, higher concentrations of EC and OC were observed in winter (0.83 +/- 0.67 and 3.33 2.28 mu m(-3), respectively), and lower concentrations were recorded in summer (0.34 +/- 0.18 and 2.30 +/- 1.15 mu g m(-3), respectively).
Inversely, the OC/EC ratio with mean value (5.1 +/- 2.6) characteristic to rural background area was higher in summer (7.33 +/- 3.23) in comparison to the other seasons. Since the data contain values below detection and quantification limits of the measuring device (i.e., censored values), statistical methods for censored data have been used in order to compare mean EC and OC concentrations between various seasons.
It was found out that there is a significant difference between summer and the other seasons with the exception of mean OC concentrations at noon. In most cases, there was also a significant difference between winter and the other seasons.
Moreover, it was found out that when dealing with OC concentrations, it is possible to replace censored values by a constant and still obtain reasonable results. In case of EC concentrations, the method based on censored distributions should be preferred when the sample size is small and the proportion of censored values is high.
The diurnal variation of EC and OC is less pronounced in summer. During working days, the EC diurnal pattern displays a morning (between 6:00 and 10:00) and an afternoon/evening (between 18:00 and 22:00) peaks, while for OC, only the afternoon/evening peak is observed.
These seasonal, diurnal and weekly variations of EC and OC concentrations and OC/EC ratio are probably related to variability in terms of emission sources (residential heating, traffic), transport characteristic and meteorological conditions. A weaker correlation between EC and OC in summer (r = 0.56) suggests additional sources and/or transport processes during other seasons.
The elevated OC/EC ratio, the higher correlation between OC and 03, and the temperature and solar radiation during summer confirmed an increasing contribution of OC from secondary organic carbon (SOC) estimated as at least 59 +/- 11% of total carbon in the PM2.5 using the EC tracer method. Backward trajectories of air masses arriving at 100 m AGL calculated in winter and summer show that higher pollution episodes of EC and OC are predominantly associated with continental air masses confined over Central Europe (about 79%), while lower EC and OC levels are mainly associated with episodes of long-range transport of marine air masses.
Interestingly, the results reveal that in winter pollutants emitted during workdays could be accumulated above the region and influence the rural background air quality during some prolonged time of the weekend, especially on Saturday.