Seasonal and diurnal variations of volatile organic compounds (VOCs) in the atmosphere of Hong Kong

Abstract

Ambient VOCs samples were collected at three locations (PolyU campus (PU), Kwun Tong (KT), Hok Tsui (HT)) in Hong Kong during the periods of November 2000–February 2001 and June 2001–August 2001. Also the concentrations of VOCs in Cross Harbor tunnel in Hong Kong were obtained in order to determine the vehicular sources of VOCs. Toluene was the most abundant VOC detected in Hong Kong. At the PU station, which is close to a main road, the concentrations of most VOCs were higher in summer than in winter. However, at the background location HT, the concentrations of all VOCs except tetrachloroethene were higher in winter than in summer. Regional physical dispersion/transportation and mixing depth may be the reasons for higher VOC concentrations in winter at HT. The BTEX (benzene:toluene:ethylbenzene:xylene) ratios of PU and KT during winter period were (1.9:10.1:1.0:1.8) and (1.9:10.4:1.0:1.5), and (0.9:8.3:1.0:2.2) and (0.8:29.6:1.0:1.8) for summer season, respectively. The xylene/ethylbenzene (X/E) ratio was used to assess the relative age of the air parcels in this study. The concentrations of VOCs in the atmosphere in Hong Kong were mainly affected by direct emissions from vehicles, evaporation of fuels, photochemical reactions and few industrial emissions. The BTEX ratio in the tunnel was 2:10.4:1:3.2. The BTEX ratios at PU and KT during the winter period were similar to that in tunnel (except for xylenes). The X/E ratio in the tunnel was higher than that in the ambient air. This indicated that the freshly emitted xylenes in the tunnel decayed at different rates from OH-oxidation in the atmosphere. Good BTEX correlations ( r>0.8) were found at PU and KT in winter (** P<0.01). Vehicular exhaust was the dominant source at PU and KT stations, and less evaporation of fuel or additive occurred at low temperature in winter. Diurnal variations of mean BTEX concentrations at the roadside monitoring station (PU) showed two peaks associated with traffic density and vehicle type.