Some insights into short-term variability of total gaseous mercury in urban air

The concentrations of total gaseous mercury (Hg) were determined at hourly intervals along with relevant environmental parameters that include both meteorological plus criteria pollutant data during two field campaigns (September 1997 and May/June 1998). The mean concentrations of Hg for the two study periods were computed as 3.94 and 3.43 ng m(-3), respectively. By separating the data into daytime and nighttime periods, we further analyzed diurnal variation patterns for both seasons. Using our Hg data sets, we were able to recognize two contrasting diurnal variation patterns of Hg between two different seasons that can be characterized as: (1) the occurrences of peak Hg concentration during daytime (fall) and (2) slight reductions in daytime Hg concentration relative to nighttime (summer). To study the systematic differences in diurnal patterns between two different seasons, we analyzed Hg data in terms of different statistical approaches such as correlation (and linear regression) and factor analysis. Results of these analyses consistently indicated that different mechanisms were responsible for controlling the daytime distribution patterns of Hg. When the relationship between Hg and concurrently determined O-3 is considered, its reaction with ozone is unlikely to limit Hg levels as the dominant sink mechanism (within the ranges of ozone concentrations found during this study, regardless of season). It is on the other hand suspected that the variation of boundary layer conditions between day/night periods may have been important in introducing the relative reduction in daytime Hg levels during summer. To further provide a general account of short-term variations in Hg distribution data, it is desirable to describe other unknown sink mechanisms. (C) 2000 Elsevier Science Ltd. All rights reserved.