National Ambient Air Quality Objectives For Ground-Level Ozone- Summary - Science Assessment Document

4. Environmental Levels

4.1 Monitoring

Ozone levels in the ambient air are measured with continuous analyzers that record one minute readings, which are then averaged over one hour. With the dataset of one hour values, ozone concentrations over other averaging periods can be calculated (e.g. 6 or 8 hours). Routine monitoring of ozone in Canada uses continuous gas analyzers that operate on a UV light absorption principle. Absorption of UV light by ambient air containing ozone can be compared to that of a reference air sample that is ozone-free.

Ambient monitoring of ozone in Canada is carried out by the National Air Pollution Surveillance Network (NAPS Network), a collaboration of federal, provincial and municipal monitoring agencies. There were 153 Canadian ozone monitoring stations reporting data from 1986 - 1993. Of these sites, 112 were located in urban or suburban areas, and 41 in rural locales. In addition, short-term studies have added to the rural ozone database. There are very few long-term data sets for remote - rural locations.

4.2 Background Ozone Concentrations

The term "background ozone" is used in different ways in different circumstances. In the truest sense, background ozone refers to the concentration of ozone in the absence of any anthropogenic input of precursor gases. Given the global transport of air pollutants though, and the almost ubiquitous presence of the precursor gases, it is not possible to identify such a level. Ozone concentrations at remote sites offer the best estimate of current background ozone levels. In these locations, ozone levels are a product of natural sources of precursors, long range transport of ozone and precursors, and contributions from stratospheric ozone. Given the limited monitoring of ozone at remote sites, "clean sites", those distant from major urban or agricultural centres, are often used for estimates of background ozone concentrations. Ozone concentrations at remote or clean sites exhibit two general characteristics: minimal diurnal variation and seasonal peaks occurring in late winter or early spring. The peak season coincides with the time of year in northern mid-latitudes when atmospheric conditions favour the mixing of stratospheric ozone down into the lower atmosphere.

Fifteen sites in Canada are identified as being sufficiently removed from anthropogenic influence to provide reasonable estimates of background ozone concentrations. The ozone concentrations observed at these sites are similar to those reported for other locations in the Northern Hemisphere. Based on these data, reasonable estimates of background ozone for areas of Canada relatively unimpacted by anthropogenic pollution are:

When all months of the year are included, values are slightly lower. The cleanest Canadian sites experience average daily maxima concentrations even lower than those above. Alert, NWT, possibly the only truly remote monitoring site in Canada, experienced an avg. daily 1 hr. maximum of 28 ppb over the 3 year period during which concentrations were monitored. Values tend to vary from year to year as shown in Figure 3 (SAD Figure 5.2).

Figure 3: Yearly variation in mean daily maximum hourly ozone concentrations (May to September) for selected remote and rural sites.

4.3 Geographic and Meteorological Factors Affecting Ground Level Ozone

As discussed above, the concentration of ozone in the ambient air is a function of complex chemical reactions and the balance between the precursors NO_x and VOC. The formation of ozone is maximal over the summer season, when higher temperatures, more intense solar radiation and longer day lengths enhance the photochemistry. Meteorological processes and geographic / topographic features also play significant roles in determining ozone concentrations. The meteorological conditions necessary for the occurrence of high ozone concentrations are well documented; they involve slow moving, anticyclonic (high pressure) weather systems. These systems, characterized by slow wind speeds and sinking of air through the troposphere, are conducive to trapping air pollutants near ground level and preventing their dispersion and dilution. Ozone episodes are therefore generally associated with climatic and meteorological conditions that favour enhanced ozone production and limited dilution/dispersion. Geographic and/or topographic features of a region can exacerbate this situation by affecting either of these processes. For example, the Lower Fraser Valley of British Columbia is known to be one area where the adjacent mountains act to confine air masses, contributing to episodes of high air pollution.

Ozone and its precursors can also be transported over distances that range from hundreds to a few thousand kilometers. Understanding the extent to which ambient ozone levels in an area are the result of local emissions of NO_x and VOC versus long range transport (LRT) of these gases, and of ozone, is a necessary step towards being able to control and reduce ozone concentrations. Studies have shown that the Windsor-Québec Corridor and the Southern Atlantic Region are two regions of Canada where LRT is a major contributor to episodes of high ozone concentrations.

4.4 Ozone Concentrations (1 hr.) at Canadian Sites

Ozone concentrations vary on a number of spatial and temporal scales, primarily due to meteorological variability and the impact this has on the transport of precursor gases and on photochemical processes. Ozone data from sites representative of different land uses across Canada were selected from the NAPS network and analyzed to illustrate seasonal, diurnal and day-of-the-week pattern in ozone levels. These datasets were also used for trend analysis. Much of the analysis was restricted to the period May to September in order to focus on the period of the year when photochemcial ozone production is greatest, and on the time of year when the primary "receptors" (people and vegetation) are most exposed to ambient ozone.

For the Canadian sites, mean ozone concentrations (May-Sept., 1986-1993) ranged from 6.1 ppb (Vancouver - Robson and Hornby) to 44.3 ppb (Ontario - Long Point). As shown in many studies, mean and median ozone concentrations are highest at rural sites and lowest at downtown urban sites. This pattern occurs because rural sites are affected by the transport of ozone and precursor gases from urban areas. Typically, rural areas lack the high NO_x values that prevail in downtown urban areas, which would otherwise scavenge ozone out of the air. Hourly concentrations of ozone in the 0-5 ppb range are not uncommon at urban sites in Vancouver, Toronto or Ottawa for example during the night. Maximum 1-hr. ozone values (May-Sept., 1986 -1993) varied from a minimum of 57 ppb (Vancouver - Robson and Hornby) to a maximum of 213 ppb (Vancouver - Hamilton and Paisley), with most sites recording maximum hourly ozone values over 100 ppb.

There are pronounced seasonal variations in monthly average daily means and daily maximum across Canada (Figure 4 and 5(from SAD Figures 5.3-5.6). The seasonal pattern varies slightly across the country, however, ozone concentrations are most elevated during April through September.

Analyses of the monthly variation in ozone concentration (monthly averages of daily mean and daily maximum 1 hr. values) revealed pronounced seasonal variations in ozone concentrations at individual sites and regions across Canada, as well as variations in the time of year when maximum ozone levels are observed. Ozone concentrations exhibit a clear seasonal cycle, with concentrations rising with the onset of warmer weather in the spring and declining again as the autumn approaches. However, the "summer season" varies across the country, and ozone concentrations clearly peak much earlier in Western Canada (May) than they do in Central Canada (July). Much less seasonal variation in ozone concentrations occurs in the Atlantic Region. Higher ozone concentrations measured during the spring may reflect the impact of ozone transport from the stratosphere. In Western Canada, the tropopause (boundary between the stratosphere and the troposphere) is closest to the ground during the spring. As a result, occasional intrusion of ozone-rich air from the lower stratosphere can occur.

In general, the diurnal cycle of ozone concentrations can be described as unimodal, with lower nighttime concentrations and a mid-day peak. The shape and amplitude of diurnal ozone cycles are strongly influenced by meteorological conditions, site location (relative to local pollution sources) and prevailing levels of precursors. In urban areas, the daily cycle of NO_x levels arising from vehicle emissions has a major impact on the daily cycle of ozone levels. The area within Canada with the most dynamic ozone patterns, with respect to both spatial and temporal variability, is Southwestern Ontario. This results from the combined influence of numerous sources within the region and the LRT of ozone and its precursors from the heavily industrialized Great Lakes Region of Canada and the U.S.

Analysis of summer versus winter data revealed large differences in the amplitude of the diurnal cycle between winter and summer seasons. Daytime ozone concentrations are much higher in summer than in winter. Nighttime ozone levels are comparable during both winter and summer seasons. For most sites, mean winter ozone concentrations are in the range of 15-20 ppb throughout the day. Concentrations tend to be somewhat lower at some Vancouver sites (5-10) and somewhat higher at more remote sites in Northern Ontario and in the Southern Atlantic Region (25-30 ppb) and in Montréal. There is much more variation across the country in daytime ozone concentrations during the summer (Figure 6 (from SAD Figures 5.7-5.10)).

Diurnal cycles of ozone concentrations were also examined for weekday-weekend differences. These differences are presumed to be in large part due to traffic flow patterns in urban areas, ^{which drive prevailing NO}x ^levels. At many sites in major urban areas, the mean maximum 1 hr. ozone on the weekend is 10-20% - and sometimes 20-35% - higher than on weekdays. At non-urban sites that are potentially affected by transport of ozone and its precursors from nearby urban centres, there is likewise an increase, albeit smaller (4-8%), in ozone concentrations on weekends. At sites in the Atlantic and Prairie provinces, the weekend change is small. The strong weekend signal in some major urban areas may be the result of less titration (removal) of regional ozone entering the city on the weekend, rather than a function of local photochemical processes. However, it should be noted that measurements of VOC are not available on the same time scales as for ozone and NO_x. When information on VOC concentrations has been available, some investigators have been able to show that changes in VOC concentrations also influence the diurnal and weekly cycles of ozone.

4.5 Ozone Concentrations For 8 hr Averaging Periods

An analysis of selected sites from the NAPS database was undertaken to investigate patterns in 8 hr. ozone concentrations, and the relationship between 8 hr. and 1 hr. maximum ozone concentrations. The analysis showed that for any given 8 hour daily maximum concentration, there is a very low probability that hourly maxima, 20 ppb or more greater than the 8 hr. daily maximum, will occur. Figure 7 (SAD Figure 5.24) illustrates the relationship between daily maximum 1 hr. ozone concentrations and daily maximum 8 hr. ozone concentrations using data from 41 sites (1992). The relationship has been shown to be quite consistent over time when data from other years have been examined.

Figure 7: Relationship between daily maximum 8 hour average and daily maximum 1 hour average for 1992 (n = 1541).

4.6 Long-term Trends in Ozone Concentrations

Ozone concentrations are highly correlated with meteorological conditions (e.g. hot summers are associated with more frequent ozone episodes). Given the variation in meteorological conditions (on many time scales), and the variability this will induce in ozone concentrations, this "noise" must first be removed from the data if trends related to precursor emission controls are to be identified. A regression model which accounted for meteorological variability was applied to daily maximum 1 hr. ozone data from sites in the Lower Fraser Valley, Windsor-Québec Corridor and the Southern Atlantic Region. While the majority of sites in Ontario showed small, but statistically significant increasing trends, 3 sites showed weak declining trends and one site showed no significant trend. On average, within the whole of Ontario, ozone concentrations seemed to be increasing at a rate of approximately 1.2% per year over the period 1980-1990. For Montréal, two sites showed a statistically significant declining trend of 1.5 and 1.1% per year, while a third site showed no trend over the period 1981-1993. In Vancouver and the LFV, the majority of sites showed statistically significant increasing trends, although three sites showed declining trends. The average trend for the region was +0.45% per year over the period 1985-1992. Similar inconsistencies were noted among sites within the SAR over the period 1985-1992. These analyses illustrate the variability of ozone trends within urban areas and within geographic regions. However, the analysis is limited in that for most sites, less than 10 years of data were utilized in the trend analysis, and because most of the sites were urban. Thus the reported trends may not be indicative of rural or regional ozone behaviour.