Skip booklet index and go to page content

Canadian Smog Science Assessment Highlights and Key Messages

Levels of Smog in the Atmosphere

Recent Levels of Fine Particulate Matter (PM2.5)

As mentioned under “Effects on Human Health”, fine particulate matter (PM2.5) is especially (although not exclusively) harmful to human health and, therefore, is the size fraction emphasized in this discussion.

The spatial pattern of ambient PM2.5 levels across the country in 2004–2006 (Figure 1), as defined by the Canada-wide Standard metric, shows southern Ontario and southern Quebec having the highest concentrations (>25 µg m-3). This area is also part of a larger high concentration airshed that encompasses all of the eastern U.S. The highest levels in Canada over the 2004–2006 period occurred in the Great Lakes region, particularly in southwestern Ontario where densely populated urban sites experienced levels above the Canada-wide Standard. Across southern Quebec and eastern Ontario, PM2.5 concentrations were generally below the Canada-wide Standard target, with the exception of some specific communities influenced by local industries and large urban centres, highlighting the potential for emissions to lead to localized high levels. There are uncertainties in the exact levels and local details of the spatial distribution of PM2.5 (Figure 1) over some areas of the country because of a lack of PM2.5 measurement sites. However, on a broad scale, Figure 1 provides a general picture of the spatial pattern.

To track the occurrence of high PM2.5 concentration events across the country, the number of days per month in which the daily 24-hour average of PM2.5 concentrations exceeded 30 µg m-3 was counted over the period 2001–2005 at sites across Canada. Days with PM2.5 concentrations >30 µg m-3 can occur any month of the year but sites in southern Ontario and southern Quebec experienced the greatest frequency of days >30 µg m-3 in summer followed by winter. At the western sites, the highest frequency of days occurred in the summer associated with the occurrence of forest fires. 

figure 1

Note: Areas in black indicate either an insufficient number of sites or incomplete data for mapping.

Figure 1 Spatial distribution of the 98th percentile 24-hour PM2.5 concentrations (µg m-3) across Canada and the U.S. for the period 2004–2006

The frequency of regional scale episodes, defined as days where 33% of air monitoring sites in a region record 24-hour average PM2.5 concentrations above 30 µg m-3, was also determined for 2001–2005. Regional scale episodes of PM2.5 occurred in both winter and summer. The greatest frequency of regional scale episodes also occurred in Ontario followed by Quebec with high PM2.5 values often persisting for several days. Summer PM2.5 episodes in Ontario and Quebec were often associated with O3 values greater than the Canada-wide Standard target of 65 ppb, whereas winter episodes were solely a result of high levels of PM2.5. Regional episodes were infrequent in the Prairies and the Lower Fraser Valley of British Columbia, and the events that did occur were associated with forest fires. Although some areas may not experience frequent regional episodes, they may still experience days when PM2.5 levels are considered high locally and for which there would be an increase in health effects and a decline in visibility relative to average conditions.

Levels of PM2.5 vary considerably by season within a region as observed through the analysis of PM2.5 daily peaks and averages measured at monitoring sites. In southern Atlantic Canada, the daily peaks and averages in PM2.5 are higher in summer than winter. This seasonal difference is due to more intense sunlight leading to greater sulphate concentrations, and more frequent favourable wind patterns that carry pollutants from sources in the southwest. In winter PM2.5 levels are influenced more by local sources such as residential wood combustion. Across southern Quebec and eastern Ontario daily peaks and averages are higher in winter than summer. In particular, over the southern Great Lakes region, the highest daily averages occur in winter as colder temperatures favour the formation and build up of ammonium nitrate on the particles (see “Composition of Secondary PM2.5”) and meteorological conditions result in less pollutant dispersion.

In Alberta, both PM2.5 peaks and averages are higher in the winter when cold and calm conditions lead to some of the highest concentrations of primary pollutants in Canada. In the Lower Fraser Valley, daily average levels are highest in late summer and early fall with peaks in fall and winter. This is due to favourable meteorological conditions in the late summer and changes in local activities, including wood burning and space heating in the fall and winter. In the interior of British Columbia, levels also peak in the fall and winter where emissions from residential wood combustion play a role along with strong inversions due to cold air pooling in the valleys and/or the trapping of air by the mountains. In Whitehorse, Yukon, daily averages are highest during the summer months, and annual variability at this site appears to be influenced by forest fires. In the winter high PM2.5 levels occur due to woodsmoke.

The spatial pattern of the annual mean of PM2.5 concentrations in Canada is very similar to the spatial pattern of the Canada-wide Standard metric (Figure 1), with the highest concentrations (>8 µg m-3) occurring in southern Ontario and southern Quebec.

Composition of Secondary PM2.5

The major components of PM2.5 as measured at sites across the country are ammonium sulphate, ammonium nitrate and organic matter (Figure 2), although the composition varies by location and by season. The first two components are products of nitrogen oxides (NOX) and sulphur dioxide (SO2) oxidation in the presence of ammonia (NH3). Organic particulate matter can be directly emitted, along with elemental or black carbon, from many combustion sources including fossil fuels, wood and cooking. Organic matter (OM) can also be formed in the atmosphere (secondary formation), largely from reactions involving both anthropogenic and biogenic volatile organic compounds (VOC).

figure 2a

figure 2b

Figure 2 Mass fractions of PM2.5 component species in the warm season (a) and in the cold season (b) sampled in 2003–2006 at selected National Air Pollution Surveillance (NAPS) network sites

In the warm season, PM2.5 at speciation sites in Ontario, Quebec and the Atlantic region contains a relatively large amount of ammonium sulphate (NH4)2SO4 followed by organic matter (OM) (Figure 2a) on average days and during peak episodes. Ammonium sulphate reaches its maximum during this season when photochemistry and transport from Canadian and U.S. emission sources are greatest. The warm season data from the western sites (with the exception of Golden, British Columbia) indicate that contributions from the main chemical constituents tend to be more equal, though organics have greater importance during peak episodes.

In the cold season, ammonium nitrate (NH4NO3) becomes a more important contributor (Figure 2b). In locations where wood burning is common, exemplified by Golden, British Columbia, organic matter is dominant and PM2.5 attains some of the highest levels observed in the country. For the highest concentration days in winter at Ontario and Quebec sites, ammonium nitrate was the primary contributor. Ammonium sulphate remains an important contributor at eastern sites and the primary contributor at Atlantic sites. At sites in British Columbia, the highest PM2.5 days in winter are dominated by organic matter.

Baseline Levels of PM2.5

In the context of this assessment, background levels of PM2.5 are defined as the ambient concentrations resulting from natural emission sources within North America, and from the long-range transport of anthropogenic and natural emissions from outside North America. Background levels are important in air pollution management as the responsible sources cannot be controlled by domestic or continental emission reduction strategies, although international negotiations on anthropogenic emission reductions may provide some benefit. Background levels explain some of the spatial variation in ambient concentrations, though they are influenced by atmospheric conditions and can be highly variable over space and time making them particularly difficult to quantify.

However, since background PM2.5 levels are not directly observable, measurement data have been used to define a baseline level, the PM2.5 level at a given site in the absence of strong local influence. Estimates of baseline annual median PM2.5 levels are available for only a few regions across the country and range from 1–4 µg m-3. Due to an insufficient number of long-term and regionally-representative sites, it is not currently possible to assess the temporal trends of baseline PM2.5. These baseline levels reflect average atmospheric concentrations upon which PM2.5 associated with North American anthropogenic activities is superimposed. Thus, neither background nor baseline is associated with the peak concentrations observed in time or space that are of considerable importance to human exposure. This conceptual difference makes it difficult to incorporate background or baseline values into discussions of the Canada-wide Standard targets which are indicators of high concentrations alone.

Trends in Ambient PM2.5

In general, ambient annual mean and 98th percentile PM2.5 concentrations at urban sites across Canada have declined by approximately 40% from 1985 to 2006 (Figure 3). The largest declines occurred prior to 1996 as a result of SO2 emissions reductions in eastern Canada and the eastern U.S. with little change since then. Some urban locations still experience high ambient SO2 levels due to large point sources and the same applies for PM2.5.

At rural and remote sites, the lack of long-term data precludes analysis of long-term trends for ambient PM2.5 concentrations. However, between 1999 and 2006, SO2 and NOx emissions reductions in eastern Canada and the eastern U.S. resulted in overall reduced ambient concentrations of precursors (i.e., ambient SO2 and nitric acid (HNO3)) and PM2.5 constituents (i.e., ambient particle sulphate, particle nitrate and particle ammonium). 

figure 3

Figure 3 Trend in composite annual mean and 98th percentile PM2.5 mass from dichotomous sampler sites

In light of decreasing SO2 and NOX emissions in eastern North America, emissions and ambient concentrations of NH3 are expected to become more important in determining levels and trends of PM2.5 (see section on “Factors Influencing Levels of Smog Across Canada”). PM2.5 levels in various regions of the country exhibit different sensitivities to ambient NH3 concentrations depending on the season and local chemical regime11, increasing the complexity in reducing ambient PM2.5.

Recent Levels of O3

The spatial pattern of O3 levels across the country in 2004–2006, expressed according to the Canada-wide Standard metric (Figure 4), shows that the highest concentrations in Canada continue to occur over southern Ontario and southern Quebec. This area is within a large high concentration region (>65 ppb), encompassing the entire northeastern U.S. The highest concentrations in Canada occurred in the Great Lakes region, particularly in southwestern Ontario coincident with the highest PM2.5 concentrations. Virtually all sites in the southern Great Lakes region recorded levels above the Canada-wide Standard for O3. In southern Atlantic Canada, O3 levels were considerably lower than southern Ontario and extreme southern Quebec, but there was noticeable spatial variability and some areas just surpassed the Canada-wide Standard. Similarly, levels observed in most of western Canada are lower (40–60 ppb) except around Edmonton, Alberta, and the eastern edge of the Lower Fraser Valley, British Columbia. Air quality in the Yukon and in the Northwest Territories (not shown in Figure 4 due to the limited number of sites) is generally better than in southern Canadian cities because of fewer local industrial sources and smaller, more dispersed populations. While the spatial and temporal coverage of monitoring sites for O3 is greater than for PM2.5 there are still uncertainties in the exact levels and local details of the spatial distribution of O3; however, on a broad scale this map provides a general picture of the spatial pattern across Canada.

To provide an indication of the occurrence of high concentration events across the country, the number of days in which the daily maximum 8-hr average O3 concentration exceeded 65 ppb over the period 2004–2006 was counted at sites across the country. The greatest number of days was recorded in southern Ontario along the north shore of Lake Erie (30–50 days) followed by the rest of southern Ontario and southern Quebec (5–30 days). In contrast, almost all sites in Western and Atlantic Canada had zero to five days with daily maximum 8-hr O3 levels of 65 ppb or greater.

figure 4

Note: Areas in black indicate either an insufficient number of sites or incomplete data for mapping purposes.

Figure 4 Spatial distribution of the three year average (2004–2006) of the fourth highest daily maximum 8-hr O3 average concentration (ppb) across Canada and the U.S.

The greatest frequency of regional scale episodes over the 2000–2005 period, defined as days when 33% of monitoring sites in a given geographic area record daily maximum 8-hour average O3 levels above 65 ppb, occurred in Ontario and Quebec, with high levels persisting for several days and often associated with PM2.5 levels above the Canada-wide Standard target of 30 µg m-3. Regional scale O3 episodes were almost non-existent in the Prairies and in the Lower Fraser Valley, with a couple of notable exceptions associated with forest fires.

Temporal Variations in Ambient O3 Levels

There are pronounced seasonal variations in ambient O3 levels, regionally and at individual sites throughout Canada. Many sites in Canada, especially outside of Ontario and Quebec, record the highest average daily mean and average daily maximum O3 levels in the spring (Figure 5). Remote sites most commonly experience a predominant maximum O3 concentration in the spring due to several mechanisms, including enhanced photochemistry in the troposphere, downward exchange of O3 from the stratosphere and enhanced hemispheric transport. Some of these sites with a spring maximum also have a less-pronounced secondary maximum in the summer. 

Figure 5 Seasonal variation in monthly averages of daily mean (o) and maximum (•) O3 concentrations at selected sites (both urban and rural) across Canada averaged over the period 2001–2005

In southern Ontario, southern Quebec and the eastern Lower Fraser Valley, the highest daily mean and maximum ambient O3 levels occur in the late-spring and/or summer, which generally indicates the impact by anthropogenic emissions and subsequent local and regional scale photochemical production. In some regions closer to the border, long-range transport of pollutants from the U.S. also plays a large role in these peak O3 events. 

On average, ambient O3 levels are low in the winter due to less intense sunlight, which reduces photochemical production of O3, and different meteorological patterns, including a greater frequency of strong, northerly winds. 

Baseline Levels of O3

As earlier described under “Baseline Levels of PM2.5”, background is the ambient concentration resulting from natural emissions within North America and from the long-range transport of anthropogenic and natural emissions from outside North America. On the other hand, baseline O3 is the O3 level measured at a given site in the absence of strong local influence. In Canada, estimated baseline seasonal average O3 levels are 19 ± 10 ppb in Pacific Canada, 28 ± 10 ppb in continental western Canada, 30 ± 9 ppb in continental eastern Canada, and 27 ± 9 ppb in Atlantic Canada. For broad comparison purposes only, because direct comparison of different metrics is not possible, the fourth highest daily maximum 8-hr concentration value at the most remote sites in Canada ranged from 44–53 ppb in 2006.

Baseline O3 levels have been increasing in a number of areas of Canada; namely the Georgia Basin (coast of British Columbia), the Atlantic coast, and continental western Canada. The increasing trend in western Canada is consistent with evidence of increasing transport of O3 into North America from Asia and a general increase in hemispheric O3. In contrast, the data indicate that the baseline has been decreasing in Ontario and Quebec; however, the baseline level in these areas is less important to air quality management considerations because anthropogenic sources within the region and transboundary transport from the U.S. continue to be a dominant contributor to O3. As O3 precursor emissions continue to decline in North America, any increase in levels of baseline O3 could play an increasingly important role in the achievement of ambient air quality targets.

Trends in O3 Levels

Although a downward trend was observed in national O3 levels in the form of the Canada-wide Standard from 2003–2006, no overall significant trend has been observed from 1990–2006. However, trends vary markedly depending on site location (urban, rural, background), period of record analyzed and the metric being examined.

The impact of Canada and U.S. NOX emission reductions12 on O3 can be seen by examining meteorologically-adjusted trends at non-urban sites. Comparing the two periods of 1997–2000 and 2003–2006 (Figure 6) reveals that the four year average daily maximum 8-hr O3 concentrations in the summer declined by 3.2% at sites in Quebec and by 4.1% at sites within an international airshed encompassing Ontario and U.S. locations in the Great Lakes/Upper Ohio region. On the other hand, ambient concentrations increased by 2.1% at sites in Atlantic Canada, remained constant at sites in Alberta, and increased by 5.2% at sites in the Georgia Basin area of the Pacific coast. 

Figure 6 Differences in the meteorologically-adjusted four-year average summer O3 levels from 1997–2000 to 2003–2006 based on daily maximum 8-hr values

At urban sites across Canada, there has been an increasing trend (not meteorologically-adjusted) in median and lower percentile ambient 8-hr O3 concentrations between 1990 and 2006. This upward trend in the annual median and lower percentile ambient O3 observations at urban sites is because there is less O3 scavenging (where O3 is removed when it reacts with NO to form NO2) as a result of reductions in NOX emissions. Over this period, annual mean ambient concentrations of NO, NO2 and VOCs at urban sites decreased by 55%, 34% and 46% respectively, with similar decreases at all sites. O3 levels in urban areas have typically been lower than the surrounding rural areas as a result of O3 scavenging. Thus, with less NO urban O3 can increase depending upon other factors such as the relationship between ambient concentrations of NOX and VOCs. 

Consistent with the meteorologically-adjusted analysis, rural sites in Ontario and Quebec experienced a decreasing trend for all percentile O3 concentrations from 1990–2006. This decrease was most significant in the upper part of the data distribution (i.e., higher percentiles, maximum), in response to regional scale reductions in precursor emissions. This includes a possible accelerating reduction in eastern Canada from 2004–2007 due to NOX emission reductions in eastern Canada and in the northeastern and midwestern U.S.

Projections of Future Levels of PM2.5 and O3

Chemical transport models are amongst current state-of-the-science models capable of simulating atmospheric chemical conditions in response to anticipated changes in emissions relative to a base or reference year. One such model, A Unified Regional Air Quality Modelling System (AURAMS), was used to estimate the levels of PM2.5 and O3 across Canada and the U.S. for the year 2015 relative to the reference year 2002. A comparison of model predictions to observations of annual O3 and PM2.5 for the year 2002 shows they are well correlated geographically, providing a positive indication of model performance. 

The emissions underlying the 2015 AURAMS PM2.5 and O3 projections (Table 1) were based on the implementation of existing Canadian and U.S. legislation, or business as usual (BAU), including the NOX State Implementation Plan (SIP) Call and the Clean Air Interstate Rule (CAIR) as proposed prior to 2008 and now replaced by the Transport Rule.

Table 1 Emissions changes between 2002 and 2015 in Canada and the U.S. included in the 2015 AURAMS business as usual scenario

Table 1 Emissions changes between 2002 and 2015 in Canada and the U.S. included in the 2015 AURAMS business as usual scenario
SubstanceCanada*Eastern U.S.Western U.S.
Primary PM2.5+19.6%-21.1%-23.7%

Note: Emissions from forest fires are excluded and a 0.75 discount factor is assumed for primary PM2.5 emissions from open sources.

*The 2015 BAU scenario was based on an earlier version of the 2015 Canadian and U.S. combined inventory and does not reflect recent updates in emissions projections presented in Chapter Four of the Assessment. As a result, the projected changes in Canadian PM2.5, NOX and SO2 emissions based upon the latest information and presented in Figure 9 of this document are smaller than the projections on this table, and the direction of the Canadian VOC emissions trend is actually reversed (+12% versus -14%). The modelling results in Figures 7 and 8, which correspond to the emissions changes in this table are, nonetheless, indicative of the directional response of PM2.5 and O3 that can be expected across Canada relative to such emissions changes.

In 2015, annual PM2.5 levels (Figure 7) are projected to be 0.2–3 µg m-3 (10–30%) lower than the 2002 levels in most of southern Ontario, responding primarily to emissions reductions in the U.S.; however, only marginal improvements are projected east of Toronto, including Quebec and most of the Atlantic Provinces. Increases of 0.2–3 µg m-3 (15–30%) are projected for a number of localized areas along the Windsor–Quebec City corridor, including the major cities. These estimates are due to increases in primary PM2.5 emissions in the Windsor–Quebec City corridor and other upwind regions. Urban centres in Manitoba, Saskatchewan and British Columbia are also projected to experience PM2.5 increases of 1–3 µg m-3 (20–50%), while the surrounding areas are projected to remain unchanged. These results show that projected PM2.5 levels are sensitive to changes in primary PM2.5 emissions; therefore, the magnitude of the change in ambient PM2.5 needs ongoing re-evaluation based on the latest projections of Canadian primary PM2.5 emissions.

figure 7

Note: Blue regions correspond to areas of projected decreases in PM2.5 concentrations while yellow to red regions correspond to areas of projected increases in PM2.5 concentrations.

Figure 7 Absolute difference in the annual PM2.5 24-hour average between the 2015 BAU simulation and the 2002 reference case

Widespread increases in PM2.5 are projected to occur in the Prairies, with the largest increases exceeding 3.0 µg m-3 (20–40%) in the vicinity of Edmonton, Alberta and southern Saskatchewan. While these are the greatest projected Canadian increases in annual ambient PM2.5 concentrations, they do not result in average annual levels above 10 µg m-3 anywhere in the Prairies except over the main urban centres. As in the Windsor–Quebec City corridor, these changes are due to increases in primary PM2.5 emissions. However, changes in the local chemical regime also appear to be playing a role given that increases in the ambient concentrations of inorganic secondary PM components (sulphate, nitrate and ammonium) are predicted despite an overall decrease in emissions of NOX, SO2 and VOCs over the area.

The 2015 BAU AURAMS simulation did not predict additional excedances of the Canada-wide Standard for PM2.5 in new locations compared to the 2002 reference year. Rather, it showed an increase in the frequency of exceedances in places that were already exhibiting levels above the PM2.5 Canada-wide Standard numerical value of 30 µg m-3.

Relative to 2002, average summer time 8-hour daily maximum O3 levels in rural or remote southern Ontario, southern Quebec and the Atlantic provinces are projected to decrease by approximately 3–10 ppb (or 10–30%) by 2015 (Figure 8). These changes are in response to projected decreases in NOX and VOC emissions (Table 1), especially along the Windsor–Quebec City corridor. Furthermore, these regions are significantly influenced by transboundary transport of O3 and its precursors from the U.S., where NOX and VOC emissions are also projected to decrease.

figure 8

Note: Blue regions correspond to areas of projected decreases in O3 concentrations while yellow to red regions correspond to areas of projected increases in O3 concentrations. Note that the Fort McMurray increase is displaced to the west due to a geographic reporting error in the 2015 projected emissions inventory.

Figure 8 Absolute difference in the average summertime (June–August) 8-hour daily maximum O3 between the 2015 BAU simulation and the 2002 reference case

Ambient O3 concentrations in urban centres in eastern Canada are projected to increase by 5–10 ppb from 2002–2015, particularly in Toronto and Montréal, due to reduced NO titration associated with decreases in local NOX emissions. This is a drawback associated with the complexity of O3 chemistry whereby reductions in precursor emissions do not always translate into anticipated reductions in O3 concentrations. However, these urban areas will benefit directly from the local NOX emission reductions and also from reductions in O3 levels and NOX emissions transported from neighbouring cities. 

In southwestern British Columbia, southern Alberta and the vicinity of Edmonton, O3 levels are projected to decrease at about the same magnitude as the eastern provinces (3–10 ppb). In Saskatchewan and Manitoba, improvements are projected to be more moderate at 1–3 ppb. In contrast, O3 levels in the vicinity of Fort McMurray are projected to increase by 1–15 ppb as a result of oil sands development and associated increases in local NOX and VOC emissions.

Projected changes in ambient O3 levels largely correspond to projected changes in the number of exceedance days of the Canada-wide Standard numeric value for O3 (65 ppb). Despite an overall decrease in the number of exceedance days, it was predicted that the vicinity of Vancouver, most of Alberta, southern Ontario and Quebec, and part of Atlantic Canada would still experience exceedance days under the 2015 AURAMS BAU scenario.

The AURAMS predictions highlight the fact that changes in PM2.5 and O3 through to 2015 are not expected to be uniform across the country, at least partially due to geographic differences in the changes in emissions. It is worth noting that projected increases in primary PM2.5 emissions in Canada have the potential to offset some of the improvements associated with lower levels of transboundary transport. In addition, an increase in PM2.5 levels is projected in large urbanized areas especially in winter, when the formation of nitrate particles dominates. Further analyses are needed to confirm these findings based on the most up-to-date emission projections for Canada.

Additional reductions in emissions, beyond the BAU scenario discussed above, have also been simulated by AURAMS in order to provide an opportunity to consider how PM2.5 and O3 can be expected to respond if additional Canadian measures were implemented. In this scenario, Canadian SO2, NOX, VOC and primary PM2.5 emissions from 15 industrial sectors have been substantially reduced relative to the 2015 BAU changes, by 41%, 23%, 16% and 6%, respectively. Based on absolute emission levels, the main sectors driving these emission reductions were smelting and electricity generation for SO2; upstream oil and gas, oil sands, and electricity generation for NOX; upstream oil and gas and oil sands for VOCs; and wood and electricity generation for primary PM2.5.

With these additional emission reductions, the largest declines in PM2.5 levels would occur in the Prairies and in the Windsor–Quebec City corridor, but only lead to marginal improvements in the Maritimes and the Vancouver area. Annual decreases in PM2.5 levels across the country corresponded largely to decreases in secondary PM2.5 formation primarily driven by reductions in sulphate levels, particularly in the summer. More moderate reductions in PM2.5 levels were projected in the winter as these were exclusively driven by expected reductions in nitrate, which are smaller in comparison to those expected for sulphate.

Widespread declines were projected in summer daily O3 in the Prairies, particularly over large portions of Alberta (>5 ppb) where the largest reductions in NOX and VOC emissions were considered in AURAMS. However, projected O3 declines would be marginal in densely populated areas, such as Vancouver, the Windsor–Quebec City corridor and the Maritimes. 

From a Canada-wide Standards perspective, the additional emission reductions discussed above would be beneficial for reducing ambient PM2.5 and O3 levels, including a potential decline in the number of exceedance days in densely populated areas. Nonetheless, the reductions investigated would be insufficient to bring all of Canada below the numerical Canada-wide Standards threshold for both pollutants.

Overall, the model runs indicate that the seasonality in the chemical formation of particles means that the potential for winter chemistry to dampen or even offset summer improvements resulting from emission reductions needs to be carefully considered. In addition, as SO2 and NOX emissions continue to decline, primary (directly emitted) particles will become relatively more important. Consequently, to realize further reductions in ambient PM2.5, a different approach with more emphasis on primary PM emissions would need to be explored. 

Relative Influence of Key Economic Sectors

Air quality model scenario analyses were performed using an ensemble of chemical transport models similar to and including AURAMS, to examine the influence of changes in emissions from six major Canadian economic sectors on ambient PM2.5 and O3 levels. The sectors studied were: agriculture, marine transportation, oil and gas (including oil sands), refinery and chemical, electricity generation, and residential wood combustion.

Simulated reductions in NH3 emissions from the agricultural sector at both the continental and regional scale (i.e., Lower Fraser Valley) were estimated to have little to no effect on O3 levels, but a moderate effect on average and median PM2.5 levels. There are also indications that locally, the influence of NH3 emissions reductions on short-term episodic PM2.5 can be substantial. Initial studies on the relative influence of the marine transportation sector on PM2.5 and O3 ambient levels in coastal areas and along seaways were indicative of a potentially large impact. Additional studies investigating the role of marine emissions in North America13, released after the Canadian Smog Science Assessment review period, supported the establishment of a Canada–U.S. Emission Control Area sanctioned by the International Marine Organization.

Upstream oil and gas was shown to be the major sector influencing ambient O3 levels in Alberta where this sector’s activities are most prevalent. This includes oil sands activities from which simulated increases in NOX and VOC emissions seemed to substantially increase O3 levels in the surrounding area. From the limited information available for PM2.5, the upstream oil and gas sector also appeared to have a large role in determining ambient PM2.5 levels in Alberta. A relatively lower sensitivity of O3 to NOX and VOCs emissions from the refinery and chemical sector in comparison to the oil and gas sector was detected in Alberta, with influences constrained to the vicinity of the source. The electricity generation sector is also a major influence on both PM2.5 and O3 levels in Alberta and southern Ontario.

The residential wood combustion sector, studied over eastern Canada, did not exhibit any influence on O3 levels since emissions are confined to winter periods and the scenario reviewed did not involve any changes in NOX and VOC emissions. The influence of reductions in primary PM2.5 emissions from this source sector on ambient PM2.5 levels in eastern Canada was also assessed. Initial results point to a significant influence, although the magnitude of the atmospheric response needs to be further investigated.

While the role of other Canadian transportation sources was not investigated as part of the model scenario runs for the Canadian Smog Science Assessment, modelling studies were conducted under the Program of Energy Research and Development (PERD)14. The results showed that in 2002, Canadian mobile transportation emissions contributed to approximately 7% of ambient PM2.5 in Canada. The study also showed that Canadian mobile sources showed the largest influence on ambient PM2.5 levels in the western provinces, while the combined estimated contribution from both Canadian and U.S. mobile sources exhibited a larger influence in eastern Canada. The results highlight the importance of transboundary transport from the U.S. in defining air pollution in eastern Canada.

11. For additional detail on the impacts of agricultural ammonia on local PM2.5, see Li, S.-M., R. Vet, J. Liggio, P. Makar, K. Hayden, R. Staebler, E. Chan, and M. Shaw, 2010. Chapter 7: Sensitivity of particulate matter to NH3 in major agricultural regions of Canada, in: The 2008 Canadian Atmospheric Assessment of Agricultural Ammonia. Environment Canada, Ottawa.

12. In Canada, NOX emissions reductions have been achieved via the On-Road Vehicle and Engine Emission Regulations and annual caps on emissions from fossil-fuelled power plants in central and southern Ontario and southern Quebec. NOX emissions reductions in the U.S. have been achieved under the NOX SIP (State Implementation Plan) Call (NOX Budget Trading Program) and Title IV of the Clean Air Act which requires NOX reductions from certain coal-fired electrical generation units.

13. ECA proposal ( available from U.S. Environmental Protection’s website for Oceangoing vessel regulation (