The Georgia Basin-Puget Sound Airshed Characterization Report 2014: chapter 11


11. Transboundary Transport

Roxanne Vingarzan, Bill Taylor, Sarah Hanna (Environment Canada) and Bob Kotchenruther (Environmental Protection Agency Region 10)

Previous chapters have shown how ambient air quality at a given location is largely determined by air emissions from local sources and those from areas upwind of the site. Generally, emission sources are local and regional in scale, but, on occasion, pollutants are transported from longer distances. For example, there is evidence that smoke from wildfires in Northern California has impacted air quality in southern British Columbia. There is also evidence of pollutants originating in Asia being transported across the Pacific Ocean on westerly winds. In this chapter, air quality within the Georgia Basin/Puget Sound airshed is examined in the context of transboundary flows. Particular attention is given to the transport of pollutants between Georgia Basin and Puget Sound, as well as medium range transport within North America and long range trans-Pacific transport.

11.1 Local Scale Transboundary Transport

Synoptic scale weather patterns and mesoscale influences play a large role in the transport of pollutants over relatively small distances and within confines of the Georgia Basin/Puget Sound airshed. Three methods used to determine potential source regions of pollutants observed at a given location are described below. One relies on surface wind observational data, while the other two are achieved through meteorological and chemical transport modelling.

For air quality monitoring sites equipped with anemometers, observational wind data are analysed to determine the frequencies of winds blowing from various compass directions. Prevailing and seasonal wind patterns become evident, indicating probable pathways for upwind pollutants. The wind data are then used in conjunction with measured concentrations of contaminants at the site to infer the most likely source region of the emissions.

The second method of analysis involves mesoscale meteorological modelling, whereby wind trajectories are used to trace the motion of the air through the three dimensional atmosphere. Source-receptor relationships can be established by tracing these back trajectories of air parcels from a particular receptor back to their origins. Using a technique called cluster analysis, a composite of back trajectories taken over a period of time reveal typical meteorological patterns around the receptor site and can provide valuable insight into common transport pathways and possible source regions affecting the air quality at a given location.

The third method involves photochemical air quality modelling, which is described in detail in Chapter 10, “Regional Air Quality Modelling”, where the impact of introducing emission controls on ambient air quality in the region is simulated through policy scenarios. Here, the air quality simulation involves completely turning off (“zero out”) the anthropogenic emissions on one side of the international border to study the effects on the receiving airshed.

11.1.1 A Study of Transboundary Transport at Christopher Point

Vingarzan et al. (2007) examined transboundary transport of pollutants at Christopher Point on Vancouver Island, one of three study sites selected under the Border Air Quality Strategy. Christopher Point lies on the southern-most tip on the Canadian side of Juan de Fuca Strait, which is divided by the Canada-U.S. border. The Strait is a vital and busy marine transport route from the Pacific Ocean to the major international port facilities at Vancouver, Seattle and surrounding areas in the Georgia Basin-Puget Sound region.

Emissions sources which may contribute to air quality degradation at Christopher Point are many and varied, including: commuter and commercial traffic and industrial activities in the nearby provincial capital city of Victoria; wood burning sources dispersed through the region; a naval shipyard at Esquimalt and a naval firing range to the immediate west of the study site; heavy marine traffic through the Strait of Juan de Fuca; a pulp mill and port at Port Angeles; and three large oil refineries in the northern Puget Sound area of Washington State. In addition, the site is exposed to large, but somewhat more distant, mixed sources from the Lower Fraser Valley of British Columbia and the Seattle-Tacoma area of Washington State.

In the summer, elevated ozone was found to occur during stagnant periods, however, in winter it was associated with westerly winds and post-frontal subsidence, indicating an influx of high altitude background ozone. Elevated PM2.5 concentrations at Christopher Point occurred during stagnant meteorological conditions throughout the year. Despite occasional episodes of poor air quality, Christopher Point did not exceeded air quality standards and objectives during the study period.

Air circulation over the area, and hence wind-borne pollutant transport, is affected by large-scale synoptic weather patterns, land and sea breeze cycles and the channelling effects of the straits. Meteorological data acquired from Environment Canada’s weather station at Race Rocks (approximately 2.5 km south of Christopher Point) indicate that the predominant wind direction is from the west, with winds channelling through the Juan de Fuca Strait. Secondary wind directions are from the north-east and south-west. In the winter, the wind direction is predominantly from the northeast, and in summer it is almost exclusively from the west.

A full year (Sept 2005-06) of measurements of ambient concentrations of nitric oxide (NO), nitrogen dioxide (NO2), sulphur dioxide (SO2), tropospheric (ground-level) ozone (O3), black (elemental, inorganic) carbon (BC) and fine-fraction aerosols (PM2.5) were obtained.

Principal component and wind sector analysis were used to identify likely pollutant sources and transport directions. Four broad patterns of elevated levels of air pollutants were identified.

  • The first was associated with elevated levels of NOx, O3, and SO2 and winds corresponding to source areas in north-western Washington State, the Puget Sound, Juan de Fuca Strait and Port Angeles. A combination of industrial, urban and marine traffic sources likely influenced this pattern.
  • The second was associated with elevated levels of O3 and stronger winds from cleaner areas such as the open Pacific, the Olympic Peninsula, the Juan de Fuca Strait and western Vancouver Island. This pattern was thought to reflect contributions from background sources.
  • The third pattern was associated with elevated concentrations of NO2, BC and PM2.5 and wind directions corresponding to Canadian source areas, such as the city of Victoria, south-eastern Vancouver Island, Vancouver and the Lower Fraser Valley.
  • The fourth pattern occurred under calm conditions which were strongly associated with NO2 and PM2.5 but also with elevated levels of NO, BC and SO2. This pattern is likely the result of local wood burning, the city of Victoria and ship traffic in the vicinity of the study site.

Using a geographic sector approach, aggregated wind sector analysis was employed to quantify the contribution of each measured pollutant from broadly defined source sectors to identify pollutant transport across the Canada-U.S. border. Four source sectors were defined in relation to Christopher Point: Canada, U.S., Pacific/Juan de Fuca and Local (Figure 11.1).

 

Figure 11.1 Source sectors used in aggregated wind sector analysis for the Christopher Point measurements (Vingarzan et al., 2007).

Figure 11.1 Source sectors used in aggregated wind sector analysis for the Christopher Point measurements (Vingarzan et al., 2007). (See long description below)

Description of Figure 11.1

Figure 11.1 is a relief map showing southwestern British Columbia and northwestern Washington State. The location of Christopher Point at the southern tip of Vancouver Island is marked with a cross. Three lines radiate out from Christopher Point and demarcate the three source sectors used in aggregated wind sector analysis for the Christopher Point measurements.  The Pacific/Juan de Fuca sector is separated from the Canada sector by a line running northwest from Christopher Point along the northern shore of the Strait of Juan de Fuca and moving out into the Pacific Ocean at Barkley Sound.  The Canada sector is separated from the US sector by a line running northeast from Christopher Point, passing through the north end of Orcas Island and running past the south tip of Harrison Lake.   The US sector is separated from the Pacific/Juan de Fuca sector by a line running west southwest from Christopher Point, through the northern end of the Olympic Peninsula, and passing just south of Ozette Lake.

 

As shown in Figure 11.2, on an annual basis, the Canadian sectors (Canada + local) dominated BC concentrations (58%), while the U.S. source sector dominated SO2 (58%) and O3(49%) concentrations. Contributions from Canadian and U.S. sectors were relatively similar for PM2.5, NO, and NO2. Local contributions accounted for 10-16% of the total concentration of each pollutant. The Pacific/Juan de Fuca sector contributed the least, accounting for up to 13% of the concentration, with the highest contribution being for ozone. Contributions from each sector varied seasonally, with the U.S. source sector dominating during the spring and Canadian source sectors dominating during the winter (Vingarzan, et al., 2007).

 

Figure 11.2 Wind sector analysis showing percentage contributions by each sector to pollutants measured at Christopher Pont during the entire year. Canadian contributions are shown as local (green) and transported (violet). (Adapted from Vingarzan et al., 2007)

Figure 11.2 Wind sector analysis showing percentage contributions by each sector to pollutants measured at Christopher Pont during the entire year. Canadian contributions are shown as local (green) and transported (violet). (Adapted from Vingarzan et al., 2007). (See long description below)

Description of Figure 11.2

Figure 11.2 is a stacked bar chart with bars for ozone, PM2.5, NO, NO2, SO2, and BC divided into percent contributions from the Pacific/Juan de Fuca, US, local, and Canada sectors. 

For ozone the percent contribution from the Pacific/Juan de Fuca sector was approximately 12%, the percent contribution from the US sector was approximately 50%, the percent contribution from the local sector was approximately 10%, and the percent contribution from the Canada sector was approximately 28%.        

For PM2.5 the percent contribution from the Pacific/Juan de Fuca sector was approximately 10%, the percent contribution from the US sector was approximately 42%, the percent contribution from the local sector was approximately 13%, and the percent contribution from the Canada sector was approximately 35%.

For NO the percent contribution from the Pacific/Juan de Fuca sector was approximately 5%, the percent contribution from the US sector was approximately 50%, the percent contribution from the local sector was approximately 15%, and the percent contribution from the Canada sector was approximately 30%.

For NO2 the percent contribution from the Pacific/Juan de Fuca sector was approximately 5%, the percent contribution from the US sector was approximately 47%, the percent contribution from the local sector was approximately 16%, and the percent contribution from the Canada sector was approximately 32%.

For SO2 the percent contribution from the Pacific/Juan de Fuca sector was approximately 5%, the percent contribution from the US sector was approximately 57%, the percent contribution from the local sector was approximately 17%, and the percent contribution from the Canada sector was approximately 21%.

For BC the percent contribution from the Pacific/Juan de Fuca sector was approximately 4%, the percent contribution from the US sector was approximately 37%, the percent contribution from the local sector was approximately 16%, and the percent contribution from the Canada sector was approximately 43%.

 

The results of this study indicate that air pollution moves across the international border, and it is closely linked to weather patterns and time of the year. Overall, the transboundary impact appears to be relatively equal, with Canada and the U.S. both contributing to pollutants crossing the international boundary.

11.1.2 Back Trajectory Analysis at Saturna Island and the Lower Fraser Valley

Due to the varied topography of the west coast, high resolution (4 km) meteorological model simulations by Environment Canada were used to calculate detailed back trajectories for Saturna Island, Abbotsford, and Chilliwack to determine the nature and extent of transboundary transport of ozone. Cluster analysis was performed to determine preferred transport pathways (Brook et al., 2011). Figure 11.3 shows one-day mesoscale back trajectory clusters for ozone at the Saturna Island CAPMoN site.

 

Figure 11.3 Clustered one day mesoscale back trajectories for O3 at Saturna CAPMoN (December 2003-March 2005) (Brook et al., 2011).

Figure 11.3 Clustered one day mesoscale back trajectories for O3 at Saturna CAPMoN (December 2003-March 2005) (Brook et al., 2011). (See long description below)

Note: The mean path (C1 - red, C2 - green, and C3 - blue) for three distinctive clusters of mesoscale back trajectories are illustrated here. The air mass arrival height was set at 50m. Vertical dimensions are not shown here.

Description of Figure 11.3

Figure 11.3 is a line map of southwestern British Columbia, western Washington, and northwestern Oregon.  It extends from approximately the location of Denny Island in the north to approximately the location of Salem Oregon in the south and as far east as Walla Walla Washington.  The mean path (C1 - red, C2 - green, and C3 - blue) for three distinctive clusters of mesoscale back trajectories for Saturna Island are illustrated. The air mass arrival height was set at 50m. Vertical dimensions are not shown. 

The C1 path arrives from the southeast having originated at the southern end of Willapa Bay.  From Willapa Bay it sweeps in a gentle curve east and north to Olympia, and continues in a direction just east of north up to Whidbey Island.  It then turns northwest and passes the southwestern side of Orcas Island before reaching Saturna.

The C2 path arrives from the northwest having originated at the northern tip of Vancouver Island, just below Hope Island.  From there it runs first south to the Brooks Peninsula before turning southeast along the west coast of Vancouver Island down to Nootka Sound.  It continues in the same southeasterly direction across Vancouver Island, crossing the coast at Duncan before arriving at Saturna.

The C3 path arrives from the north having originated on the east coast of Vancouver Island just south of Nanaimo. From Nanaimo it crosses the Strait of Georgia to Boundary Bay and then turns south to arrive at Saturna. 

 

Maximum summer ozone concentrations (6 hr averages near 60 ppb) were associated with a slow moving, clockwise rotating, stagnant flow (C3), typical of high pressure conditions and summertime regional stagnation. In contrast, elevated fall, winter, and early spring ozone levels (C2) (6 hr average concentrations in the range of 40 ppb) were associated with a longer range, fast moving, north-westerly flow (Vingarzan et al., 2007). Springtime ozone concentrations at Saturna Island were associated with variable flow directions ranging from north-westerly in early spring, giving way to southerly (C1) over the Puget Sound in Washington State and local stagnant flows as summer approaches. A similar flow pattern was found for Abbotsford (not shown) (Brook et al., 2011).

Further east in the Lower Fraser Valley at Chilliwack (Figure 11.4), two flow directions were associated with elevated summer ozone (6 hour average maxima between 60 and 70 ppb), a north-easterly outflow (C3) and a lighter, roughly south-westerly flow (C1) crossing the Whatcom County area of Washington State into British Columbia (Brook et al., 2011).

 

Figure 11.4 Clustered one day mesoscale back trajectories for O3 at Chilliwack Airport (December 2003-March 2005) (Brook et al., 2011).

Figure 11.4 Clustered one day mesoscale back trajectories for O3 at Chilliwack Airport (December 2003-March 2005) (Brook et al., 2011). (See long description below)

Note: The mean path (C1 - red, C2 - green, and C3 - blue) for three distinctive clusters of mesoscale back trajectories are illustrated here. The air mass arrival height was set at 50m. Vertical dimensions are not shown here.

Description of Figure 11.4

Figure 11.4 is a line map of southwestern British Columbia, western Washington, and northwestern Oregon.  It extends from approximately the location of Bella Coola in the north to approximately the location of Portland Oregon in the south. It extends as far east as the Washington-Idaho border, and as far west as Port Hardy.  The mean path for three distinctive clusters of mesoscale back trajectories (C1 - red, C2 - green, and C3 - blue) for Chilliwack airport are illustrated. The air mass arrival height was set at 50m. Vertical dimensions are not shown. 

The C1 path arrives from the southwest having originated in the Strait of Juan de Fuca just south of Victoria BC and then travelling in an almost straight line to Chilliwack airport. 

The C2 path arrives from the south having originated near Longview, Washington.  From Longview it travels in a northeasterly sweep up to the Snoqualmie area and then due north to Chilliwack airport.

The C3 path arrives from the northeast having originated in the interior of BC (somewhere in the vicinity of Salmon Arm). It travels briefly to the southwest and then south before making a long southwesterly sweep into Chilliwack airport.

The Saturna and Chilliwack mesoscale trajectory clusters for ozone were compared to clusters derived from lower resolution Canadian Meteorological Centre (CMC) trajectories (not shown) which cover longer time periods. In both cases CMC and mesoscale trajectories were able to identify the same general dominant flow patterns influencing maximum summer ozone concentrations, however the mesoscale trajectories more accurately described flows around geographical features (Brook et al., 2011).

The lower resolution CMC trajectories were also used to evaluate PM2.5 transport to Chilliwack (Figure 11.5). These showed that for Chilliwack in the spring, summer and fall, higher PM2.5concentrations were linked with two transport patterns associated with lighter, more stagnant flows. One of these pathways was from the east-by-southeast (ESE), while the other was a light flow from the west-by-northwest (WNW:L) direction, passing over the Vancouver metropolitan area. The lowest PM2.5 levels were associated with SSW and WSW flows (Brook et al., 2011).

 

Figure 11.5 (a) CMC trajectory cluster mean vectors of PM2.5 data at Chilliwack derived for the time period 1995 to 2005 and (b) Seasonal box-and-whisker PM2.5 plots by trajectory clusters at Chilliwack (Brook et al., 2011).

Figure 11.5 (a) CMC trajectory cluster mean vectors of PM2.5 data at Chilliwack derived for the time period 1995 to 2005 and (b) Seasonal box-and-whisker PM2.5 plots by trajectory clusters at Chilliwack (Brook et al., 2011).

Notes: The “L” attached to directional flows in Figure (a) refers to light winds. Seasonal breakdowns in Figure (b) are as follows: March - April - May (MAM), June - July - August (JJA), September - October - November (SON) and December - January - February (DJF). Based on hourly data.

Description of Figure 11.5

Figure 11.5 is composed of two panels. Panel A on the left shows an line map of North America extending from the tip of the Baja Peninsula to the Arctic Ocean and as far east as Lake Michigan.  Five CMC trajectory cluster mean vectors for PM2.5 data at Chilliwack are shown.  These are derived for the time period from 1995 to 2005.  The WNW trajectory cluster arrives from the west-by-northwest having originated in the Gulf of Alaska just south of the Alaska Peninsula. The WNW:L trajectory cluster (the “L” refers to light winds) also arrives from the west-by-northwest, although with a slightly increased northerly component.  The WNW:L trajectory cluster originates of the west coast of Haida Gwaii.  The WSW trajectory cluster arrives from just south of west having originated in the offshore waters to the west of Vancouver Island.  The SSW trajectory cluster arrives from the south-by-southwest having originated offshore of northern California. The SSW:L trajectory cluster also arrives from the south-by-southwest having originated just off the coast of Oregon.  The ESE trajectory cluster arrives from east-by-southeast having originated just southeast of Chilliwack at the Canada-US border.

Panel B is a box and whisker plot showing PM2.5 concentration (in µg/m3) broken down into four seasons  (March - April - May, June - July - August, September - October - November, and December - January - February) and with each of the four seasons broken down by the trajectory clusters shown in Panel A.  The plot is based on hourly data.

For March - April - May the data is as follows:

  • The ESE trajectory cluster had a median of approximately 7 µg/m3, a mean of approximately 8 µg/m3, an interquartile range of approximately 5-10 µg/m3, and a full range of approximately 2.5-18 µg/m3
  • The SSW trajectory cluster had a median of approximately 5 µg/m3, a mean of approximately 5 µg/m3, an interquartile range of approximately 3-5 µg/m3, and a full range of approximately 2-7.5 µg/m3.
  • The SSW:L trajectory cluster had a median of approximately 5 µg/m3, a mean of approximately 5.5 µg/m3, an interquartile range of approximately 3-7 µg/m3, and a full range of approximately 2-12 µg/m3.
  • The WNW trajectory cluster had a median of approximately 6 µg/m3, a mean of approximately 6 µg/m3, an interquartile range of approximately 4-7 µg/m3, and a full range of approximately 2.5-11 µg/m3.
  • The WNW:L trajectory cluster had a median of approximately 7 µg/m3, a mean of approximately 7 µg/m3, an interquartile range of approximately 5-8 µg/m3, and a full range of approximately 2.5-14 µg/m3.
  • The WSW trajectory cluster had a median of approximately 5 µg/m3, a mean of approximately 5 µg/m3, an interquartile range of approximately 4-6 µg/m3, and a full range of approximately 2-8 µg/m3.

For June - July - August the data is as follows:

  • The ESE trajectory cluster had a median of approximately 10 µg/m3, a mean of approximately 10 µg/m3, an interquartile range of approximately 7-14 µg/m3, and a full range of approximately 3-19 µg/m3
  • The SSW trajectory cluster had a median of approximately 5 µg/m3, a mean of approximately 5 µg/m3, an interquartile range of approximately 4-5.5 µg/m3, and a full range of approximately 2.5-7.5 µg/m3.
  • The SSW:L trajectory cluster had a median of approximately 5.5 µg/m3, a mean of approximately 6 µg/m3, an interquartile range of approximately 4-7 µg/m3, and a full range of approximately 2.5-14 µg/m3.
  • The WNW trajectory cluster had a median of approximately 6 µg/m3, a mean of approximately 7 µg/m3, an interquartile range of approximately 5-8 µg/m3, and a full range of approximately 2.5-15 µg/m3.
  • The WNW:L trajectory cluster had a median of approximately 8 µg/m3, a mean of approximately 8.5 µg/m3, an interquartile range of approximately 6-11 µg/m3, and a full range of approximately 3 -15 µg/m3.
  • The WSW trajectory cluster had a median of approximately 5 µg/m3, a mean of approximately 5.5 µg/m3, an interquartile range of approximately 3-6 µg/m3, and a full range of approximately 2-12.5 µg/m3.

For September - October - November the data is as follows:

  • The ESE trajectory cluster had a median of approximately 9 µg/m3, a mean of approximately 10 µg/m3, an interquartile range of approximately 5.5-12.5 µg/m3, and a full range of approximately 3-18.5 µg/m3
  • The SSW trajectory cluster had a median of approximately 5 µg/m3, a mean of approximately 5.5 µg/m3, an interquartile range of approximately 4-6 µg/m3, and a full range of approximately 2.5-10 µg/m3.
  • The SSW:L trajectory cluster had a median of approximately 6.5 µg/m3, a mean of approximately 7 µg/m3, an interquartile range of approximately 4.5-8 µg/m3, and a full range of approximately 2.5-14 µg/m3.
  • The WNW trajectory cluster had a median of approximately 6.5 µg/m3, a mean of approximately 6.5 µg/m3, an interquartile range of approximately 5-7.5 µg/m3, and a full range of approximately 3-12 µg/m3.
  • The WNW:L trajectory cluster had a median of approximately 7 µg/m3, a mean of approximately 8 µg/m3, an interquartile range of approximately 5-11 µg/m3, and a full range of approximately 3-18.5 µg/m3.
  • The WSW trajectory cluster had a median of approximately 5.5 µg/m3, a mean of approximately 6 µg/m3, an interquartile range of approximately 4-5.5 µg/m3, and a full range of approximately 2.5-12 µg/m3.

For December - January - February the data is as follows:

  • The ESE trajectory cluster had a median of approximately 5 µg/m3, a mean of approximately 5.5 µg/m3, an interquartile range of approximately 2.5-7.5 µg/m3, and a full range of approximately 1.5-12.5 µg/m3
  • The SSW trajectory cluster had a median of approximately 5 µg/m3, a mean of approximately 5 µg/m3, an interquartile range of approximately 3-7 µg/m3, and a full range of approximately 2.5-10 µg/m3.
  • The SSW:L trajectory cluster had a median of approximately 5 µg/m3, a mean of approximately 5 µg/m3, an interquartile range of approximately 3-7 µg/m3, and a full range of approximately 2-10.5 µg/m3.
  • The WNW trajectory cluster had a median of approximately 5.5 µg/m3, a mean of approximately 5.5 µg/m3, an interquartile range of approximately 3.5-7.5 µg/m3, and a full range of approximately 2.5-10 µg/m3.
  • The WNW:L trajectory cluster had a median of approximately 5 µg/m3, a mean of approximately 5 µg/m3, an interquartile range of approximately 3-7 µg/m3, and a full range of approximately 2-12 µg/m3.
  • The WSW trajectory cluster had a median of approximately 5 µg/m3, a mean of approximately 5 µg/m3, an interquartile range of approximately 3.5-7.5 µg/m3, and a full range of approximately 2-9.5 µg/m3.

 

CMC cluster analyses were also performed for several sites in the U.S. IMPROVE network, located south of the BC border in Washington State. Results (not shown) indicate that national park sites tended to have the highest PM2.5 during the summer period, while more urban-influenced sites in the Puget Sound, had maximum PM2.5levels during the winter. The latter is consistent with wood burning and low boundary layer heights during the cold season. In some cases, the flow direction associated with maximum PM2.5 concentrations coincided with westerly flow across the WISE (North Cascades NP), while in other cases it coincided with transport from Puget Sound (Olympic NP). Note however that these sites were limited to 2-3 years of data and, as such, the level of uncertainty is higher than for the Canadian sites.

11.1.3 Back Trajectory Analysis at Snoqualmie Pass

The Washington State Department of Ecology conducted back trajectory analysis on roughly five years of twice-daily wind fields using the Penn State/NCAR MM5 mesoscale meteorological model at 12 km resolution. This analysis was done in support of the Washington State Visibility Protection State Implementation Plan (SIP). Trajectory analysis was applied to identify probable source contribution areas for two long-term IMPROVE monitoring sites used for the Visibility SIP review. Twelve hour trajectories terminating at Snoqualmie Pass and Mount Rainier for specific worst and best case visibility days during 1997, 1998 and 1999 were examined. The study also examined the seasonal distribution of areas whose emissions are expected to contribute to the air quality at these two sites (Van Haren, 2002).

The Snoqualmie Pass back trajectories provide a good example of transboundary transport. Figure 11.6 clearly shows trajectories entering the Puget Sound airshed from a number of source regions including the Canadian Georgia Basin. Analysis indicates that the transboundary trajectories originating in the Georgia Basin are more frequent in spring and autumn.

 

Figure 11.6 Distribution of autumn 12-hour trajectories to Snoqualmie Pass, Washington (Van Haren, 2002).

Figure 11.6 Distribution of autumn 12-hour trajectories to Snoqualmie Pass, Washington (Van Haren, 2002). (See long description below)

Note: Each trajectory is 12 hours long and terminates at approximately 150 meters above a monitoring site. Segments of trajectories greater than 500m above the surface are ignored.

Description of Figure 11.6

Figure 11.6 is a screenshot of the output from the Penn State/NCAR MM5 mesoscale meteorological model at 12 km resolution for 12-hour trajectories to Snoqualmie Pass. The underlying image is a line map of southern British Columbia, Alberta, and Saskatchewan, as well as the entirety of Washington, Oregon and Idaho. The Northern third of California, Nevada, and Utah are also included, as are western Wyoming and Montana. Overlaid on this map is a trajectory density plot. There is a note that each trajectory is 12 hours long and terminates at approximately 150 meters above a monitoring site. Segments of trajectories greater than 500m above the surface are ignored.  The highest number of trajectories originate from within Washington State, however there is a significant density originating from the southern part of Vancouver Island and the eastern Strait of Juan de Fuca.  There is also a significant density originating from northwestern Oregon.  Trajectories arrive from as far away as the southeastern corner of Alberta, northern Idaho, and northwestern California.

The summer trajectories (not shown) are shorter in comparison to the other seasons, suggesting smaller contributions from transboundary sources.  This is not the case for the autumn trajectories shown in Figure 11.6, in which the paths travelled are longer, and there is a definite increase in the number of air parcels entering the airshed through the Haro Strait portal.

An analysis of the trajectories associated with the worst 20% of visibility cases at Snoqualmie is shown in Figure 11.7. The trajectories on the worst days spend a large fraction of time over the populated areas of Puget Sound (Van Haren, 2002).

Figure 11.7 12-hour trajectories associated with the worst 20% of visibility cases at Snoqualmie Pass, Washington (Van Haren, 2002)

Figure 11.7 12-hour trajectories associated with the worst 20% of visibility cases at Snoqualmie Pass, Washington (Van Haren, 2002). (See long description below)

Notes: ”IQR” (blue box) is defined as the inter-quartile range; outliers are not necessarily statistically significant

Description of Figure 11.7

Figure 11.7 is a screenshot of the output from the Penn State/NCAR MM5 mesoscale meteorological model at 12 km resolution for 12-hour trajectories to Snoqualmie Pass associated with the worst 20% of visibility cases. The underlying image is the same as in figure 11.6, that is a line map of southern British Columbia, Alberta, and Saskatchewan, as well as the entirety of Washington, Oregon and Idaho. The Northern third of California, Nevada, and Utah are also included, as are western Wyoming and Montana. Overlaid on this map is a trajectory density plot. There is a note that each trajectory is 12 hours long and terminates at approximately 150 meters above a monitoring site. Segments of trajectories greater than 500m above the surface are ignored.  Most of the trajectories in figure 11.7 spend significant time in the southern portion of Puget Sound. The trajectories are much shorter than in figure 11.6 with the farthest arriving from southern Vancouver Island, south and central Washington State, and the coastal area just north of Astoria.

 

The 12-hour long trajectories on the worst days are noticeably shorter than any of the seasonal or the best day trajectories. The best days (not shown) are characterized by higher wind speeds coming from the west or southwest. For Snoqualmie Pass, the trajectory analysis suggests that local sources dominate, although there are also a series of trajectories associated with poor visibility extending down from Haro Strait (Van Haren, 2002).

11.1.4 The "Zero Out" Emissions Modelling Studies

One means of estimating transboundary transport of pollutants is through “zero out” simulations, whereby the emissions from one source region are set to zero, and the results are then compared to an unperturbed simulation.  It should be noted that for pollutants that undergo chemical transformations, zero out simulations do not provide a realistic source attribution due to the dynamic nature of atmospheric chemical transformations, but none the less the difference between the two simulations provides an estimate of the effects of the emissions from the source region on the receiving airshed. The CMAQ modelling system was applied at 12 km resolution in a zero out simulation for Georgia Basin/Puget Sound, where Canadian and American anthropogenic emissions were each in turn set to zero to determine the effects of transboundary flows of ozone and PM2.5 on the other country (RWDI, 2003). The base cases used in this study are the same as those described in the previous chapter, Chapter 10: Regional Air Quality Modelling, including a summer case (Aug 9-31, 2001) and a winter case (Dec 1-13, 2002). The summer 2001 base case consists of a variety of summer weather conditions, including several days of hot stagnant air accompanied by episodes of poor air quality. The winter case also covers a period during which stagnant conditions conducive to elevated PM2.5 were observed. It should be noted that ozone results were not presented for the winter case due to the mitigating effects of low temperatures and solar radiation on photochemical production.

“Zero out” U.S. emissions (adapted from RWDI, 2003)

For the summer base case, ozone concentrations increased substantially in the Seattle area and Puget Sound due to weakened O3 titration by NO during stagnant meteorological conditions. Downwind of these urban and industrialized areas, ozone concentrations were significantly lower due to the absence of precursor NOx emissions; however, some O3 and PM2.5 production continued to occur as a result of biogenic sources of NOx, VOCs and ammonia (NH3) over the U.S. portion of the domain. Peak ozone levels in the Lower Fraser Valley were lower by about 10 ppbv than with the U.S. emissions left on, while peak PM2.5 levels in Metro Vancouver were lower by about 14 µg/m3. Due to the onshore westerly flows, O3 and PM2.5 plumes that formed over the Lower Fraser Valley and the Strait of Georgia generally moved eastward parallel to the Canada/U.S. border rather than back and forth across the border. These results suggest that, under stagnant meteorological conditions, U.S. emissions contribute to Canadian ozone and PMprecursors. In general, however, the impact of transboundary pollutant and precursor transport is short-ranged, limited to within tens of kilometers either side of the border. Once the wind flows started from the southwest, there was no significant transport of ozone from Canada to the U.S.A. (RWDI, 2003).

For the winter 2002 simulation, PM2.5 concentrations were very low during the stagnant phase around Seattle, Tacoma and the Puget Sound. When easterly winds changed to northerly winds for a period of time, the PM2.5 plume moved from the Lower Fraser Valley and Strait of Georgia to the northern tip of Olympic Peninsula and to Seattle. Other minor intrusions into the U.S. occurred all along the valleys that line the Canada/U.S. border during these periods. In contrast, PM2.5 levels over Vancouver Island were lower than the base case, indicating that elevated PM2.5 levels from the U.S. typically travelled northward over Vancouver Island. On the other days during this episode, there was little or no cross-boundary impact from Canada to the U.S. due to the predominantly easterly to southerly wind flows (RWDI, 2003).

“Zero out” Canadian emissions (adapted from RWDI 2003)

With the Canadian emissions set to zero, an increase of ozone over the summer base period was observed in the Metro Vancouver area and marine traffic routes due to weakened NO titration effects. Decreases in ozone occurred downwind of these areas. Large scale reductions in PM2.5 were seen around Metro Vancouver and the east coast of Vancouver Island. Some ozone and PM2.5formation occurred throughout the Canadian portion of the domain due to biogenic emissions of VOCs and NOx. Due to onshore westerly flows that dominated the coastal regions in the U.S. and Canada during this episode, ozone and PM2.5 plumes that formed over the urban areas of Seattle and Portland moved generally eastward, parallel to the Canada/U.S. border, resulting in relatively little transboundary transport into the Lower Fraser Valley and the southern portion of the Strait of Georgia. At times, PM2.5 levels in north-central Washington were 15 to 20 µg/m3 lower than the base case results. This suggests that a polluted airmass from the Canadian side of the border, likely associated with marine emissions, moved southward into the U.S. Some relatively weak, short-range transport of ozone and its precursors from the U.S. travelled to the Lower Fraser Valley and southern portions of the Strait of Georgia from August 11-15.  Afterward, when major wind flows came from the southwest, the transboundary impacts on ozone levels were quite low because of fairly high wind speeds and subsequent ventilation (RWDI, 2003).

As with the summer episode, there was a significant reduction in PM2.5 around Metro Vancouver and eastern Vancouver Island during the stagnant phase. When easterly flows dominated the entire domain, PM2.5 plumes that formed over the Seattle and Portland regions moved offshore to the west, over the Pacific Ocean resulting in no significant transboundary impacts from the U.S. on Canada. During December 6-10, the wind flow patterns veered to south-easterly, causing the PM2.5 plume from the Seattle region to move north-westward across the straits of Georgia and Juan de Fuca and over the southern coast of Vancouver Island. Compared to the winter base case, the PM2.5 concentrations from the “zero out” Canadian simulations in these areas are quite high with peaks of around 24 µg/m3, meaning that about 50 to 60% of base case levels can be attributed to transport from the U.S. (RWDI, 2003).

Conclusions of the“Zero Out” CMAQ Simulations

Both of the zero out cases showed limited trans-boundary impacts on ozone and PM2.5 concentrations during the summer 2001 episode. One of the reasons is that the majority of the low-level wind flows during this episode are east-west in nature, such that the air pollutants are transported more or less parallel to the Canada /U.S. border instead of crossing back and forth. In addition, sea-land breezes and the mountain-valley flows can induce local transboundary impacts but generally prohibit the longer distance transport of pollution during this episode.

The winter case showed a higher frequency of longer range transport. For example, when the wind flow pattern was from the southeast, the airmass travelled greater distances northward from the Seattle region to Vancouver Island, the Strait of Georgia and southern Vancouver Island (RWDI, 2003).

The results of these Zero Out  simulations identified different transboundary pollutant transport regimes in the region and have shown that  that these are governed by both geography and specific meteorological conditions. While longer-range transport does occur, this likely happens less often than the more local-scale transboundary effects that occur along the B.C. / Washington border, and particularly in the vicinity of the Lower Fraser Valley and southern Vancouver Island (RWDI, 2003).

11.1.5 Additional "Zero Out" Modelling Simulations

Two additional zero out modelling studies using the lower resolution CHRONOS and AURAMS models simulated transboundary transport of PM2.5 in the summer months. The CHRONOS model (21 km grid spacing) was applied nationally to examine the impact of U.S. PM2.5 emissions on Canada, and vice versa, for the summers of 2003 and 2004. Figure 11.8 shows that the maximum Canadian influence on PM2.5 in the Puget Sound area is in the range of 0.2 to 1.0 μg/m3. Similarity, the influence of U.S. emissions on Canadian PM2.5 ambient levels were between 0.2 and 1.0 μg/m3, with a peak between 1.0 to 2.5 μg/m3 at the border between Georgia Basin and Puget Sound airshed in BC (Bouchet et al., 2011).

 

Figure 11.8 CHRONOS map of (a) the maximum influence of Canadian emissions on ambient PM2.5 levels in the U.S. and (b) the maximum influence of U.S. emissions on ambient PM2.5 levels in Canada for summer 2004 (expressed as relative sensitivity of PM2.5 levels in μg/m3) (Adapted from Bouchet et al., 2011)

Figure 11.8 CHRONOS map of (a) the maximum influence of Canadian emissions on ambient PM2.5 levels in the U.S. and (b) the maximum influence of U.S. emissions on ambient PM2.5 levels in Canada for summer 2004 (expressed as relative sensitivity of PM2.5 levels in μg/m3) (Adapted from Bouchet et al., 2011)

Note: These maps allow the identification of U.S. areas that are sensitive to changes in Canadian emissions (and vice versa) and provide a first estimate of their level of sensitivity

Description of Figure 11.8

Figure 11.8 has is composed of two panels.  Both are line maps encompassing British Columbia and Washington State as well as the immediately adjacent areas of the Yukon, Northwest Territory, Alberta, Montana, Idaho, and Oregon.  Both panels show the result of CHRONOS zero out modelling results.  In both cases the maps are colored according relative sensitivity of PM2.5 levels in μg/m3 with the scale ranging from -0.1 to >25.0 (in increments of -0.1, 0.0, 0.1, 0.2, 0.5, 1.0, 2.5, 5.0, 10.0, and 25.0). 

The left panel shows the maximum influence of Canadian emissions on ambient PM2.5levels in the U.S. On the northeastern side of Puget Sound the influence of Canadian emissions is on the order of 1-2.5 μg/m3. Along the southern and western shores of Puget Sound and along the southern shore of the Strait of Juan de Fuca the influence of Canadian emissions is on the order of 0.5-1 μg/m3. In the furthest south and west areas of the airshed the influence of Canadian emissions is on the order of 0.2-0.5 μg/m3.

The right panel shows the maximum influence of US emissions on ambient PM2.5levels in Canada. In the western side of the lower Fraser Valley immediately adjacent to the US border the influence of US emissions is on the order of 2.5-5 μg/m3. For the most of the remainder of the lower Fraser Valley and the metro Vancouver area the influence of US emissions is on the order of 1-2.5 μg/m3.  Long the northern and eastern edges of the lower Fraser Valley the influence of US emissions is on the order of 0.2-1 μg/m3.  Along the southern and eastern edges of Vancouver Island the influence of US emissions is on the order of 0.2-0.5 μg/m3.

The figure has a note that these maps allow the identification of U.S. areas that are sensitive to changes in Canadian emissions (and vice versa) and provide a first estimate of their level of sensitivity.

 

The AURAMS (42 km) model (not shown) also showed that both the predicted impact of Canadian emissions on ambient U.S. PM2.5 levels and the predicted impact of U.S. emissions on ambient Canadian PM2.5 levels for the year 2015 were not very different from the 2004 base case, although this particular study did not consider any of the proposed regulations that are being assessed as part of the Regulatory Framework for Air Emissions in Canada (Bouchet et al., 2011).

The results of these trans-boundary studies suggest that transport of both ozone and PM2.5 occurs between the two airsheds throughout the year, depending on prevailing weather patterns. Significant transport of air pollutants appeared to occur primarily across the Canadian and U.S. portions of the Lower Fraser Valley, in a 100 km wide band along the border, with transport occasionally reaching westward to Vancouver Island.

11.2 Medium Range Transboundary Transport

The studies described in the previous section have shown how meteorological conditions can affect the transport of pollutants within and between the airsheds of the Georgia Basin and Puget Sound. However, there is evidence that pollutants can travel from locations far away from this region and have an effect on the local air quality. The next study documents medium range transport of smoke from wild-fires in California.

11.2.1 Californian forest fire plumes over Southwestern British Columbia

A particularly severe fire season in Northern California during the summer of 2008 resulted in two documented cases of smoke being transported to Oregon, Washington and southwestern British Columbia. McKendry et al. (2011) used ground-based and satellite lidar, sunphotometry and high altitude chemistry observations at Whistler B.C. and Mt. Bachelor in Oregon to characterize the optical, physical and chemical properties of plumes of smoke from the Northern California wild fires. Southerly winds carried dense smoke northward in early July and then again in early August. Satellite based measurements showed extensive smoke haze spreading into southwestern British Columbia from June 30 to July 3 and August 4-7. Meteorological conditions at the time consisted of a stable layer in the lower to mid-troposphere due to a temperature inversion which acted as a vertical constraint on the smoke plumes, while being advected northward.

During both events, ground level concentrations of PM2.5 at Vancouver International Airport reached about 15 μg/m3, or roughly three times background concentrations. The highest PM2.5concentrations were recorded at Hope on July 2, 2008 (~40 μg/m3). However, on August 6-7, PM2.5 concentrations across the WISE did not exceed 15 μg/m3 (McKendry et al., 2011).

This study highlights the fact that poor air quality in Georgia Basin/Puget Sound can be affected by events taking place in distant regions well beyond the control of local authorities. The study also points to the significance of medium range transport of smoke into the region, in light of projected increases in forest fire frequency and severity in western North America under future climate change scenarios.

11.3 Long Range Transboundary Transport

Local and regional sources have by far the largest impact on pollutant concentrations in any area. However, contaminants from one region of the globe can be swept up to higher altitudes and transported across oceans and continents by strong winds aloft. Air masses transported over British Columbia typically originate over the Pacific Ocean and are usually relatively clean, exhibiting the characteristics of northern hemispheric background concentrations. Large-scale events of transported pollutants are episodic in nature and tend to be relatively infrequent.

Transport of dust and pollutants from East Asia and Siberia has been well documented in recent years. Pollutant transport from Asia to western North America takes on the order of 5-6 days (Jaffe et al., 1999; McKendry et al., 2001) and occurs mainly in the free troposphere with subsidence bringing the air down near North America. Although there is substantial variation from year to year, on average the continuous contribution of Asian particulate matter to 24hr PM2.5 levels in Western North America is on the order of 0.2 to 1.0 µg/m3, with the greatest influence occurring at elevations above 500 m (Van Curen and Cahill, 2002). An EPA study of trans-Pacific transport using CMAQmodelling gave similar results, suggesting that Asia contributes 0.6 -1.6 μg/m3 to monthly average PM2.5 levels in the western United States (Keating et al., 2005). 

Although Asian particulate matter impacts North America throughout the year, there is typically elevated transport in the springtime when meteorological conditions are well suited to lofting plumes above the boundary layer in Asia and subsequent transport across the Pacific (Fischer et al., 2009). Measurements at Whistler Mountain in the Georgia Basin have shown that Asian contributions to mean springtime sulphate at this elevated site are on the order of 0.31 μg/m3 with levels reaching as high as 1.5 μg/m3 during significant transport episodes (van Donkelaar et al., 2008). Additional measurements at Whistler Mountain by Leaitch et al. (2009) associated sulphate, nitrate and organic material with coarse dust particles of Asian origin.

Very large Asian dust events are less frequent, occurring only every three to five years (Keating et al., 2005). In 1998 large dust storms in the Gobi Desert contributed about 50% of the total PM2.5 load measured in the LFV, which reached a maximum value of 44 μg/m3 (McKendry, 2006).  In 2001, another large dust event raised the mean PM10 concentration at 110 IMPROVE sites throughout the continental U.S. by an average of 9 μg/m3 (Jaffe et al., 2003).

Although rare, dust plumes over North America have even been observed to originate from as far away as the Sahara Desert. Using model output, satellite imagery, surface monitoring observations and sun photometer data, McKendry et al. (2007) confirmed that dust from a March 2005 Sahara dust storm crossed Asia and the Pacific and then gradually subsided along the west coast of North America approximately 14 days later. Although the event had only a minor effect on ambient PMconcentrations in BC, this was the first time that a long range (~19,000 km) intercontinental pathway for pollutants from the Sahara had been documented.

In addition to particulate matter and its precursors, other air pollutants are also transported across the Pacific under the specific meteorological conditions. Studies of background ozone in the Pacific Northwest indicate that there is a measurable contribution from Asian sources. The maximum contribution to surface ozone occurs in the springtime and is on the order 3-10 ppb (Vingarzan, 2004; Weiss-Penzias, 2004).  A study of transpacific ozone pollution using GEOS-Chem (a global 3-D chemical transport model) gave similar results, suggesting that Asia contributes approximately 5 -7 ppbv at the surface in the springtime over the western North America (Zhang et al., 2008). The 2000-2006 rise in Asian anthropogenic emissions increased this influence by 1-2 ppbv. Although this general contribution is relatively small, individual episodes can significantly enhance background ozone levels and can contribute to exceedences of air quality standards. Large Siberian wild fires during the summer of 2003 had a significant impact on ozone levels in BC and Washington State.  At Enumclaw, 8-hour average ozone concentrations reached 96 ppb on June 6, 2003, of which 15 ppb was attributed to the Siberian fire plume. In the LFV, ambient ozone concentrations reached 80 ppb during the same event and were likely enhanced to a similar extent by the Siberian fire plumes (Jaffe et al., 2004).

Sensitivity simulations based on emission change scenarios for ozone precursors indicate that a 10% increase in Asia would increase the surface ozone in the U.S. region by 0.1 to 0.2 ppbv at the surface and by 0.2-0.4 ppbv at 650 and 200 hPa (Wild and Akimoto, 2001). Larger contributions were found in the spring and fall. Results from GEM-AQ/EC model showed that although O3 enhancement by Asian emissions could range as high 10 ppbv episodically, a 20% reduction in O3 precursors would only result in a 0.3-0.5 ppbv O3reduction in the western part of Canada.

In terms of policy-relevant maximum daily 8-h average ozone (MDA8 O3), sensitivity simulations (Reidmiller et al., 2009) in which anthropogenic O3 precursor emissions were decreased by 20% in four source regions: East Asia, South Asia, Europe and North America show that the greatest response of MDA8 O3 to the summed foreign emission reductions occurs during spring in the West at 0.9 ppbv. East Asia was found to be the largest contributor to MDA8 O3 at all ranges of the O3 distribution for most regions (typically ~0.45 ppbv), followed closely by Europe. While the impact of foreign emissions on surface O3 in the U.S. is not negligible and is of increasing concern, reductions in domestic O3 precursor emissions remain a far more effective means of decreasing MDA8 O3 by a factor of 2 to nearly 10 relative to foreign emissions reductions.

While a number of studies have shown no clear pattern in trends of PM2.5transport from Asia (NRC, 2009; McKendry, 2006), it seems reasonable to consider that rapid economic growth and emission increases in Asia will lead to increased concentrations of pollutants in air masses transported across the Pacific (Keating et al., 2005). Likewise a continuing rise in anthropogenic emissions from Asia is expected to increase the background level of ozone (Vingarzan, 2004; Zhang et al., 2008). There are some indications that background ozone levels over the mid-latitudes of the Northern Hemisphere have continued to rise over the past three decades of approximately 0.5-2% per year (Vingarzan, 2004). A better understanding of these processes will hopefully lead to a more complete picture of long range transport in the future. At present however, it remains difficult to characterize precisely how pollution transported aloft ultimately affects air quality or ecosystems at ground-level and to partition observed pollution into domestic and foreign components (NRC, 2009).

11.4 Chapter Summary

Several studies based on wind observations and back trajectory modelling indicate that pollutants cross the Georgia Basin/Puget Sound international boundary throughout the year. The Christopher Point study linked elevated levels of contaminants with their probable source regions through surface wind analysis, revealing a strong seasonal dependence on weather patterns. Back trajectory analysis at three locations in the Georgia Basin showed that high ozone concentrations are associated with stagnant meteorological conditions in the summer. In other seasons, high ozone may occur as the result of downward vertical mixing and transport from the upper troposphere resulting from the passage of synoptic scale weather fronts. Elevated PM2.5 concentrations can occur at any time of year during stagnant conditions, and fine particulate matter can be transported under light wind conditions.

Modelling results showed that there is regular but very localized transboundary transport, particularly across the international border in the Lower Fraser Valley during stagnant conditions conducive to smog episodes. These flows are not normally associated with exceedences of air quality standards and objectives on either side of the border. Other flow patterns in the fall, winter, and spring, and during unstable periods in the summer, can carry pollutants and precursors across the international boundary on a more episodic basis. “Zero out” emissions modelling scenarios, in which Canadian and U.S. emissions were respectively turned off, were conducted to estimate the transboundary effects of air pollution. The zero out simulations indicate that, for the specific meteorological and synoptic patterns during the selected base case, local/urban-scale air quality impacts from transboundary transport occur along the border (within about 50 km) with some frequency during the summer. However, the incidence of longer-range (over 100 km) regional transport within the airshed was low. The winter simulations indicate a higher frequency of longer range regional transport. For example, under a southeast wind pattern, plumes were shown to travel from the Seattle area to Vancouver Island. In general, Canadian and U.S. regions within the airshed exerted a similar degree of influence on each other in terms of pollutant transport.

Air quality in the Georgia Basin/Puget Sound region can also be influenced by medium range transport of pollutants originating well outside of the airshed. During the devastating wildfires in Northern California in the summer of 2008, smoke plumes trapped under a temperature inversion in the lower to mid-troposphere were observed to be transported under the influence of southerly winds into southwestern B.C., resulting in elevated PM2.5 levels in the Lower Fraser Valley.

The long range transport of pollutants across the Pacific contributes to background levels of ozone and PM2.5 in Georgia Basin/Puget Sound. Periodically, Asian dust events and large wild fires can have short term detrimental effects on ambient air quality.

11.5 References

Bouchet, V., Buset, K., Bloxam, R., Carou, S., Cousineau, S., di Cenzo, C., Davignon, D., Farrell, C., Jiang, W., Kellerhals, M., Makar, P., Morneau, G., Pelletier, S., Smyth, S., and Vingarzan. R., 2011. “Chapter 6. Model Scenarios and Applications”. In: Canadian Smog Science Assessment. Volume 1. Atmospheric Science and Environmental Effects. Environment Canada and Health Canada.(Executive summary) (Full report available upon request) Environment Canada, Science and Technology Branch, 4905 Dufferin St, Downsview, Ontario, M3H 5T4

Brook, J.R., Farrell, C., Friesen, K., Langley, L., Kellerhals, M., Morneau, G., Reid, N., Vingarzan, R., Waugh, D., and Wiens, B., 2011. "Chapter 7. Air quality at the regional and local scale: The what, where, why and how of concentration variations". In:Canadian Smog Science Assessment. Volume 1. Atmospheric Science and Environmental Effects. Environment Canada and Health Canada.(Executive summary) (Full report available upon request) Environment Canada, Science and Technology Branch, 4905 Dufferin St, Downsview, Ontario, M3H 5T4

Van Curen R.A. and Cahill, T.A., 2002. Asian aerosols in North America: Frequency and concentration of fine dust. Journal of Geophysical Research 107 (D24), 4804, doi:10.1029/2002JD002204.

van Donkelaar, A.,  Martin, R.V., Leaitch, W.R., Macdonald, A.M., Walker, T.W., Streets, D.G., Zhang, Q., Dunlea, E.J., Jimenez, J.L., Dibb, J.E., Huey, L.G., Weber, R., and Andreae, M.O., 2008. Analysis of aircraft and satellite measurements from the Intercontinental Chemical Transport Experiment (INTEX-B) to quantify long-range transport of East Asian sulfur to Canada. Atmospheric Chemistry and Physics 8: 2999 -3014.

Fischer, E.V., Hsu, N.C., Jaffe, D.A., Jeong, M-J., and Gong, S.L., 2009. A decade of dust: Asian dust and springtime aerosol load in the U.S. Pacific Northwest.  Geophysical Research Letters 36, L03821. doi:10.1029/2008GL036467.

Van Haren, F., 2002. Review of the Washington State Visibility Protection State Implementation Plan - Final Report. Washington State Department of Ecology, November 2002. pp 82-97.

Jaffe, D.A., Anderson, T., Covert, D., Kotchenruther, R., Trost, B., Danielson, J., Simpson, W., Berntsen, T., Karlsdottir, S., Blake, D., Harris, J., Carmichael, G. and Uno, I., 1999. Transport of Asian Air Pollution to North America. Geophysical Research Letters 26, 711-714.

Jaffe, D., Bertschi, I., Jaeglé, L., Novelli, P., Reid, J. S., Tanimoto, H., Vingarzan, R., and Westphal, D. L., 2004. Long-range transport of Siberian biomass burning emissions and impact on surface ozone in western North America.  Geophysical Research Letters31, L16106, doi: 10.1029/2004GL020093.

Jaffe D., Snow J., and Cooper O., 2003. The April 2001 Asian dust events: Transport and substantial impact on surface particulate matter concentrations across the United States. Eos, Transactions, American Geophysical Union 84 (46): 501 - 516. Nov. 18th, 2003.

Keating, T., J. West, and D. Jaffe, 2005. Air Quality Impacts of Intercontinental Transport, EM (Environmental Management), October, 28-30.

Leaitch, W.R., Macdonald, A.M., Anlauf, K.G., Liu, P.S.K., Toom-Sauntry, D., Li, S.M., Liggio, J., Hayden, K., Wasey, M.A., Russell, L.M., Takahama, S., Liu, S., van Donkelaar, A., Duck T., and Martin, R.V., Zhang, Q., Sun, Y., McKendry, I., Shantz, N.C., Cubison, M., 2009. Evidence for Asian Dust Effects from Aerosol Plume Measurements during INTEX-B 2006 near Whistler, BC, Atmospheric Chemistry and Physics 9: 1-24.

McKendry, I. G., 2006.  Background concentrations of PM2.5 and Ozone in British Columbia, Canada. University of British Columbia report prepared for the British Columbia Ministry of the Environment, Vancouver, B.C.

McKendry, I.G., Hacker, J.P., Stull, R., Sakiyama, S., Mignacca D., and Reid, K., 2001. Long-range transport of Asian dust to the Lower Fraser Valley, British Columbia, Canada. Journal of Geophysical Research 106 (D16): 18,361-18,370.

McKendry, I.G., Macdonald, A.M., Leaitch, W.R., van Donkelaar, A., Zhang, Q., Duck, T., and Martin, R.V., 2008. Trans-Pacific dust events observed at Whistler, British Columbia during INTEX-B, Atmospheric Chemistry and Physics 8: 10275-10300.

McKendry, I.G., Strawbridge, K., O’Neill,  N., McDonald, A.M., Liu, P., Jaegle, L., Jaffe, D., Fairlie, D., Anlauf K., and Westphal, D., 2007. A Case of Trans-Pacific Transport of Saharan Dust to Western North America, Journal of Geophysical Research 112, D01103, doi:10.1029/2006JD007129.

McKendry, I., Strawbridge, K., Karumudi, M. L., O'Neill, N., Macdonald, A.M., Leaitch, R., Jaffe, D., Sharma, S., Sheridan, P., and Ogren, J., 2011. Californian wildfire plumes over Southwestern British Columbia: lidar, sunphotometry, and mountaintop chemistry observations. Atmospheric Chemistry and Physics 11: 465-477.

National Research Council (NRC), Committee on the Significance of International Transport of Air Pollutants, 2009. Global Sources of Local Pollution: An Assessment of Long-Range Transport of Key Air Pollutants to and from the United States.

Reidmiller, D., Fiore, A. Jaffe, D., Bergmann, D., Cuvelier, C., Dentener, F., Duncan, B., Folberth, G., Gauss, M. Gong, S., Hess, P., Jonson, J., Keating, T., Lupu A., Marmer, E., Park, R., Schultz, M., Shindell, D., Szopa, S., Vivanco, M. Wild, O., and Zuber, A., 2009. The influence of foreign vs. North American emissions on surface ozone in the US., Atmospheric Chemistry and Physics 9: 5,027-5,042.

RWDI, 2003. Pacific Northwest International Air Quality Modelling Project, Phase 2 report, Environment Canada, Pacific & Yukon Region, Vancouver, BC, August 28, 2003.

Vingarzan, R., 2004. A review of surface ozone background levels and trends. Atmospheric Environment 38: 3,431 - 3,442.

Vingarzan, R., Fleming, S. Meyn, S., Schwarzhoff, P., Belzer, W., 2007. Air Quality and Transboundary Transport at Christopher Point, British Columbia. Meteorological Services of Canada, Environment Canada.

Weiss-Penzias, P., Jaffe, D.A., Jaegl, L., and Liang, Q., 2004. Influence of long-range transported pollution on the annual and diurnal cycles of Carbon monoxide and ozone at Cheeka Peak Observatory, Journal of Geophysical Research 109, D23S14, doi:10.1029/2004JD004505.

Zhang, L., Jacob, D., Boersma, K., Jaffe, D., Olson, J., Bowman, K., Worden, J., Thompson, A., Avery, M., Cohen, R., Dibb, J., Flock, F., Fuelberg, H., Huey, L., McMillan, W., Sigh, H., and Weinheimer, A., 2008. Transpacific transport of ozone pollution and the effect of recent Asian emission increases on air quality in North America: an integrated analysis using satellite, aircraft, ozonesonde, and surface observations. Atmospheric Chemistry and Physics 8: 6117-6136.

Report a problem or mistake on this page
Please select all that apply:

Thank you for your help!

You will not receive a reply. For enquiries, contact us.

Date modified: