If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
Human activities have changed the chemical composition of the global atmosphere over the past few decades. Particularly, industrial emissions, intensification of agricultural practices, urban development, and transportation have directly enhanced the levels of pollutant gases and aerosols, which are likely to have serious climatic implications both directly through perturbations of the radiation budget and indirectly through chemical feedbacks . For example, the polluted air may lead to a number of contemporary issues related to human health, ecosystems, national heritage, regional climate change, etc. It has been observed that a significant fraction of air pollutants emanates from fossil fuel combustion and specific industrial activities, which mainly involve large point sources like thermal power plants, steel plants, cement plants, sulfuric acid manufacturing, smelting of copper, zinc and lead ores , etc. Besides, emissions from industrial point sources are controllable while other sources (e.g., soil erosion, forest fires, and road travel) are subject to some unpredictable natural or economic factors. That is why most efforts on air pollution control have concentrated on industrial point sources, especially those sources existing in the metropolitan and urban areas [3-10].
The ambient air quality is usually measured to help understand the state of air pollution in the surrounding area of a point source, which is commonly indicated by the concentration of gaseous pollutants and size or number of particulate matter . A number of studies on ambient air quality have been reported in the past decades. For example, Atkinson and Lewis  developed a number of linear programming models to minimize the costs of emission control to achieve ambient air quality standards, based on the assumption that improvements in ambient air quality are proportional to reductions in regional emissions. Chow  discussed the measurement methods for determining and attaining compliance with ambient air quality standards for suspended particles. Meanwhile, some improvements or recommendations for the ambient air quality standards had been put forward regarding ozone and nitrogen dioxide in view of its implications of health effects [14-19]. Previous research efforts were mostly focused on evaluating the air pollution level of major cities or communities and its implication to human health [20-26]. However, there were few findings to evaluate the potential impacts of air pollutant emission from industrial point sources, especially in a spatial-analysis way. Spatial analysis can reveal the geographical patterns of point sources to help screen out some sources with harmful influences on the surrounding area, and thus support decision-making with respect to air pollution emission mitigation for each individual emitter [27-29]. Besides, the contribution of a given industry source to the local air quality can be estimated through spatial analysis based on its geographical distance to residential district. In this sense, spatial analysis is no doubt an effective approach for impacts analysis of air pollutant emissions.
Therefore, this article is aiming to present an effective approach based on spatial analysis to assess the potential impacts of air pollutant emissions from point sources in the province of Saskatchewan, Canada. Trend analysis will be first carried out to demonstrate the temporal changes in the total number of and the spatial distribution of point sources. Then, the IDW method will be used to generate interpolation surfaces for main air pollutants emitted by industrial sources to inform their spatial emission patterns. Following that, ten representative industrial facilities will be screened out to estimate their impacts on the surrounding residents, aiming to demonstrate the effectiveness of the proposed approach. The main objective of this study is to introduce an effective approach with the aid of spatial analysis to facilitate assessing the potential impacts of point sources on local residents, by quantifying how many people might be affected. This approach can also be used to estimate the potential impacts of newly-planned industrial projects based on their predesigned specifications. Thus, it can provide useful information for supporting the decision-making in terms of the site selection and the migration of local residents if necessary.
In this article, all industrial point sources in the province of Saskatchewan are to be studied. As reported by the Government of Saskatchewan , the total area of Saskatchewan is 651,900 km2, with 570,269 km2 land area and 81,631 km2 fresh water area. Of the ten provinces of Canada, Saskatchewan ranks fifth in total land area, making up 6.5% of the Canadian total. Its geographical location is illustrated in Figure 1. Saskatchewan's census population for 2011 was 1,033,381, according to Saskatchewan Bureau of Statistics . There was an increase of 65,244 persons (6.7%) from the 2006 census population of 968,157. Saskatchewan is an energy and mineral powerhouse, blessed with a wealth of resources that would be the envy of nations. And it is on the strength of those resources that Saskatchewan has emerged as an economic leader in Canada . Saskatchewan accounts for 28% of the country's primary energy production, the highest of all provinces in Canada. Saskatchewan is Canada's second largest oil producer, the fifth largest oil producer among all American states and Canadian provinces, and the third largest producer of natural gas and coal in Canada. The primary industrial activities cover electricity, mining, oil and gas, chemicals, manufacturing, cement, lime, and other nonmetallic minerals.
The Canada-wide Standards (CWS), developed by the Canadian Council of Ministers of the Environment in June 2000, provide the numeric targets for PM2.5 (particulate matter smaller than 2.5 μm) and ozone and associated statistical forms:
PM2.5: 30 µg m−3, the 3-yr average of the annual 98th percentile of the daily 24-h average concentrations.
Ozone: 65 ppb, the 3-yr average of the annual 4th highest of the daily maximum 8-h average concentrations.
Particulate matter is a problem throughout all seasons and in all regions of Saskatchewan, while ozone is a summer regional problem. Most particulate matter in Saskatchewan results from dust of soil erosion and road travel transported by high wind velocity. Agricultural activities in the spring and fall also increase particulate matter, as do some industrial activities. Forest fires, through the summer months, impact large parts of the province's air quality. However, agricultural and other natural sources are difficult to control, most of efforts turn toward industrial sources. As a formal commitment of air quality to its residents, Saskatchewan has endorsed the CWS for particulate matter and ozone . That means Saskatchewan agreed on the implementation of continuous improvement, pollution prevention, and keeping clean areas clean programs in areas with ambient concentrations below the CWS levels, in accordance with the guidance provided in CWS.
As released in the comprehensive report of Saskatchewan Ministry of Environment , the ambient levels of PM2.5 are 14 μg m−3 in the city of Regina and 9 μg m−3 in the city of Saskatoon, respectively. The ambient levels of ozone for Regina and Saskatoon are 52 and 51 ppb, respectively. It is apparent that these two cities have not exceeded the standard required by CWS. However, this report did not show the air quality of other cities due to deficiency of monitoring data. It is challenging to assess the air quality for the entire province of Saskatchewan. Moreover, the concentrations of PM2.5 and ozone usually vary with the distance between monitoring site and pollution sources (especially industrial facilities). In other words, people who live in the neighborhood of industrial sources may suffer polluted air with higher concentrations of particulate matter and ozone than the standards of CWS. Nevertheless, the industry provided approximately $1.7 billion in revenue in 2010–2011 and was one of the largest contributors to the provincial economy of Saskatchewan, and industrial activity will continue to expand with the increase of investing in resource exploration and development . This may pose a threat to regional air quality and inhabited environment. Hence, it is necessary to estimate the potential impacts of industrial point sources to support decision-making regarding the tradeoff between economic development and air pollution prevention, especially in the context of population expansion.
The air pollutant emission data for point sources was collected from the National Pollutant Release Inventory (NPRI), Environment Canada. The time horizon of dataset spanned over 15 yr from 1994 to 2008, which is to be used for trend analysis. The NPRI contained all criteria air contaminants (CAC) and other related pollutants, in collaboration with all industrial sectors including electricity, chemicals, mining, cement and lime, oil, and gas pipelines and storage, oil sands and heavy oils, upstream oil and gas, petroleum and coal products refining and manufacturing, pulp and paper, wastewater treatment, wood products, and other manufacturing. The geographical information of all industrial facilities was also acquired from the NPRI along with their administrative details such as facility name, owner company, etc. The emission data of 2008 was screened out to illustrate the analyzing process of potential impacts for point sources in terms of PM2.5. The Environment Canada's Wind Energy Atlas offers valuable data about wind energy potential based on numerical simulations that were run on all of Canada. The data of mean wind speed at the height of 30 m in 2008 was obtained from this atlas to facilitate the evaluation of possible transport distance for air pollutants. Census data of 2006 was acquired from Statistics Canada, which is to be aggregated with spatial distribution of point sources to estimate the total number of people influenced by pollutant emissions from industrial activities.
First of all, this study will carry out trend analyses for all point sources based on the NPRI dataset from 1994 to 2008, aiming to inform their temporal and spatial variation. To provide a better understanding of the emission pattern of precursors for particulate matters and ozone, the inverse distance weighted (IDW) method will be employed to generate the interpolated surfaces for main air pollutants such as total particulate matters (TPM), sulfur dioxide (SO2), nitrogen dioxide (NO2), carbon monoxide (CO), volatile organic compounds (VOCs) and methane (CH4).
The potential impacts of air pollution involve not only affecting human health by being inhaled directly, but also pose some indirect influences. For example, the polluted gases emitted from industrial sources can dissolve into rain to form acid rain, which can erode human buildings, damage natural vegetation and crops, threat aquatic life and so on. Therefore, this study will present an effective way to evaluate the effects of air pollutant emission from point sources on the surrounding residents and the major cities.
The IDW method is widely used to estimate an unknown value as the weighted average of its surrounding points, in which the weight is the inverse of distance raised to a power . The IDW can be expressed as follows :
where zu is the unknown value to be estimated at point u, zi is the attribute value at control point i, diu is the distance between points i and u, s is the number of control points used in estimation, and k is the power. The higher the power, the stronger (faster) the effect of distance decay is. In other words, raising distance to a higher power will imply stronger localized effects. In this study, power k is fixed as 2 and the number of control points is set as 12 while generating interpolation surfaces for the main air pollutants.
In general, gaseous pollutants can be transported long distances by air currents. Therefore, the air quality at a given site may result from a mixture of local, regional, and/or distant sources. The contribution of a point source to the monitoring site is inversely proportional to the distance between them. That is why IDW method is introduced in this study to generate the interpolation surfaces for main pollutants. Because of the lack of enough data on air pollutant concentration reported by the NPRI, we assume that the more pollutants a point source emits the higher concentration its ambient environment will have. Therefore, annual emission data of the main air pollutants can be used in this study to approximately reflect their concentration distribution within the boundary of Saskatchewan. Note that the interpolation surfaces cannot be directly used for estimating the actual concentration of the corresponding pollutant at a specific point since they are based upon annual emission data other than observed concentration records.
Potential Impact on Local Residents
Air pollutants emitted from a point source may spread over a large area, and the concentration of the pollutants will decrease gradually as contaminants move through the air. The dispersion distance is strongly depending on the emission rate of the pollution source, average wind speed and geometric stack height. Current research efforts on air pollution mainly focused on the ambient air quality, which is usually measured by the concentration of fine particulate matter (PM2.5) and ground-level ozone. The concentration of the pollutants can be calculated according to Gaussian dispersion model. A widely-used model for the dispersion of gaseous air pollutants is the Gaussian model developed by Pasquill , in which gases dispersed in the atmosphere are assumed to exhibit ideal gas behavior. However, it seems unreasonable to estimate the potential impact of air pollutant emission from a point source by the concentration of pollutants in monitoring cities, especially while the point source is far away from the cities. Therefore, the number of people who may be affected by a point source is proposed in this study to assess its potential impact.
To determine how many people may be involved, we introduce the potential area of influence (PAI), which is defined as a region around a given point source within which the concentration of specific air pollutant is higher than a threshold value (usually delimited by the air quality standard). This region, similar to the dispersion distance, is determined by geometric stack height, wind speed and emission rate. In addition, wind direction is an important factor which forces pollutants to move predominantly downwind. Considering the capriciousness of wind direction, we assume the wind may come from any directions within a natural year period. Thus, the PAI can be refined as a circular area around the point source. Thus, the key thing is to determine its radius. As a result, the total number of people likely to be affected by the point source can be calculated by adding up those people who are covered by the region of PAI. Assume the PAI spans four census districts A, B, C, and D, the population density of the ith district is expressed as Pi (i = 1, 2, 3, 4); the partial area of the ith district covered by the PAI is Si, which can be calculated with the aid of spatial analysis function of ArcGIS ; the number of people who are likely to be affected in the ith district can be estimated by Si×Pi. As a result, the total number of people being affected will be
where rPAI is the radius of the PAI (illustrated as Figure 2).
Unfortunately, the radius of PAI can not be calculated directly since it is subject to several inexact factors (e.g., wind speed, emission rate of industrial facility). The average or estimated values, therefore, should be used in the computing process. The following describes how to estimate the possible maximum PAI radius of PM2.5 for all point sources of Saskatchewan in 2008. Firstly, according to the Gaussian dispersion model , the greatest value of the ground level concentration is always situated along the plume centerline, which can be expressed as:
where is the concentration at distance x from the source along the plume centerline (unit: g m−3), Q is the emission rate of the pollution source (unit: g s−1), u is the average wind speed (unit: m s−1), is the standard deviation of the plume in the y direction (unit: m), is the standard deviation of the plume in the z direction (unit: m), and H is the geometric stack height (unit: m). The standard deviations and are measures of the plume spread in the crosswind and vertical directions, respectively. They depend on atmospheric stability and on distance from the source. Atmospheric stability is classified in categories A through F, called stability classes. Class A is the least stable while Class F is the most stable. In this study, we assume Class D stability to indicate neutral conditions. The standard deviations and are both functions of downwind distance x. Thus they can be computed once atmospheric stability is determined . The maximum emission rate of PM2.5 can be estimated with the highest annual emission in 2008, which was 267.483 tonnes emanated by a facility of Potash Corporation located in Lanigan. Thus, we have the maximum emission rate for PM2.5 as 8.46 g s−1. According to the numerical simulation results in terms of mean wind speed and energy reported by Environment Canada's Wind Energy Atlas , the annual wind speed at the height of 30 m ranges from 4 to 6 m s−1 within Saskatchewan's boundary. Therefore wind speed u can be estimated by taking the mean value of 5 m s−1. Because of lack of stack height information at the time of this study, we assume that the average stack height is 40 m based on the discussion of stack threshold options regarding regional air quality modeling, which was released by Environment Canada . Consequently, we can calculate the concentrations of PM2.5 at different distance to the point source based on Eq. (3); and then estimate the possible maximum radius for PM2.5 by comparing the concentrations with the CWS (i.e., 30 µg m−3). The testing result is shown in Figure 3. At the beginning, the ground level concentration of PM2.5 sharply arises as the distance to the point source increases; the maximum concentration is observed at ∼800 m; after that, the concentration gradually decreases with the increasing of distance. It is approximately at 3565 m where the concentration of PM2.5 starts to drop below the standard required by the CWS. Therefore, the estimate of possible radius for PM2.5 is set as 3565 m. This value will be used to analyze the potential impacts of point sources as mentioned before.
The objective of trend analysis is to exhibit the temporal and spatial changes of industrial activities in Saskatchewan from 1994 to 2008. Figure 4 shows the changing pattern of geographical distribution for all point sources; meanwhile, Figure 5 draws the fluctuation in the number of point sources during the given time horizon. As shown in Figure 4, most of point sources are located in the southern part of Saskatchewan. This spatial pattern coincides pretty much with the population distribution (i.e., most of people live in the south). At the beginning of 1994, there were only 43 point sources in Saskatchewan. After that, the total number did not change until 2002 when the number of industrial facilities had been doubled. Especially in 2003, the number of point sources had jumped to more than 450. This growing trend continued until 2006 when it reached the peak of 587. Although there was a slight decrease in the later two years (i.e., 2007 and 2008), the total number of point sources was still higher than 450. The rapid growth of point sources in 2002 and 2003 is likely to be related to increasing resources extraction, primarily oil and gas.
Concentration Distribution of Main Pollutants
The IDW method is used in this study to generate interpolation surfaces for main air pollutants based on their annual emission from point sources, aiming to reflect their concentration distribution in the context of Saskatchewan. Considering the data integrity and availability, this study chose the inventory emission data of 2008 to create the interpolation surfaces. The results are shown in Figures 6a–6f.
The concentration patterns of TPM, SO2 and NO2 presented a similarity in the southern end of Saskatchewan, especially the mid-south where the highest concentration for these three pollutants was captured. This was closely related to the fact that a power station with the greatest emission amount of the pollutants was located in this area. In addition, higher concentration of TPM could be found in the middle area nearby Humboldt and the eastern area nearby Yorkton and Melville. The pattern of CO tended to move forward to the west and the north, higher concentration of CO was mainly distributed in the area nearby Lloydminster, Meadow Lake, Regina, Prince Albert, Swift Current, and the rural area in the southwestern corner. High emission of VOCs concentrated on North Saskatchewan where very few people lived; however, the regions nearby Saskatoon, Melfort, and Moose Jaw also indicated relatively high levels of VOCs. The hot spots of CH4 concentration mainly sprawled out in some rural areas. In general, most of the major cities did not suffer higher concentration of these pollutants due to the majority of point sources distributed in the rural areas.
Potential Impacts of PM2.5
To estimate the impacts of air pollutant emission from a point source on its surrounding residents, PM2.5 is chosen in this study as a representative air pollutant to demonstrate the proposed method in terms of potential impacts assessment. The possible radius of PAI for PM2.5 in Saskatchewan has been estimated based on the emission inventory of 2008, using the aforementioned method. For each point source, the PAI will be a circular area of fixed size around itself if the same radius is applied (i.e., 3565 m). This may not be consistent with the fact that the transport distance of air pollutant for each point source is subject to various conditions, such as emission rate, stack height, average wind speed, atmospheric stability, etc. These conditions are usually changing with its geographical location, thus the transport distance varies with each individual source. As a result, the potential dispersion areas are usually different in reality. Hence, the interactive relationship between the variety of conditions and the transport distance of air pollutant from point source should be investigated. Because of the data availability, this study only considers the effect of emission rate, which is assumed to be proportional to the total annual emission, on the transport distance of PM2.5. On the basis of the estimating process for the radius of PAI, we sample a number of point sources and calculate their emission rates. For each value of the emission rates, the possible transport distance (i.e., radius of PAI) can be identified by comparing with the CWS for PM2.5. Figure 7 indicates that a significant linear relationship exists between the radius of PAI for a point source and its annual emission of PM2.5. In addition, it is found that the ground level of PM2.5 at any point will be never higher than 30 µg m−3 if the annual emission is ≤70 tonnes.
According to the above analysis, we screen out 10 facilities with their annual emission of PM2.5 in 2008 higher than 70 tonnes. Most of these facilities come from mining sector. The radius of PAI for each facility can be calculated through the regression equation shown in Figure 7. Then, the potential impact of a given point source on local residents, expressed as the total number of people being affected, is estimated depending on the population density of each census region covered by its PAI. The results are listed in Table 1. The spatial distribution of these ten facilities together with their potential impacts of PM2.5 in 2008 is presented in Figure 8. There are a considerable number of people affected by the facility named Co-Op Refinery/Upgrader Complex compared with others. Although the annual PM2.5 emissions of several mining facilities are much higher than the amount emanated from the Co-Op facility, which comes from petroleum and coal sector, their potential impacts on the surrounding people is extremely low. The reason is apparent if we look into the geographical location of these facilities. As shown in Figure 8, the Co-Op facility is located in the city of Regina while most of other facilities are far away from the major cities. That is why the Co-Op facility has a tremendous potential impact in terms of the total number of people being affected. There are also two facilities relatively near the cities (i.e., the Cory mining facility resides in the neighborhood of Saskatoon and the power station of Boundary Dam is in the nearby area of Estevan, respectively), but the actual distances from the cities for both facilities are greater than their radius of PAI. Therefore, less of people will be affected by these two facilities. In brief, the closer an industrial facility to the major cities (usually with considerable high density of population), the higher potential impacts it will have. Hence, it is strongly recommended to build the new facilities with high emissions in rural areas where few people live.
Table 1. Potential impacts of selected facilities.
Annual emission of PM2.5 (tonnes)
Radius of PAI (m)
Total number of people being affected
Potash Corp—Lanigan Division
Mosaic Potash Esterhazy—K2 Site
Mosaic Potash Esterhazy—K1 Site
Poplar river power station
Boundary dam power station
Co-op refinery/upgrader complex
Petroleum and coal
Weyburn oil battery
Oil sands and heavy oils
Mosaic potash colonsay
This research presented an effective approach with the aid of spatial analysis to evaluate the potential impacts of air pollutant emissions from point sources in Saskatchewan. Trend analysis was firstly carried out to demonstrate the temporal changes in the total number of and the spatial distribution of point sources from 1994 to 2008. The results indicated that there was an explosive growth of the total number of facilities from 2001 to 2003, and most of newly-built facilities were distributed in the south of Saskatchewan where the majority of the total population of the province lives. The IDW method was then used to generate interpolation surfaces for main air pollutants emitted by industrial sources in 2008. Similarly, the results showed that most of the hot spots of emissions were situated in the southern part. Following that, we proposed a spatial analysis approach which quantifies the impact of a facility on local residents by estimating the total number of people covered by its PAI. Then, ten industrial facilities were chosen with their annual emission of PM2.5 in 2008 more than 70 tonnes to estimate their impacts on the surrounding residents, aiming to demonstrate the effectiveness of the proposed method. The results showed that those facilities near to major cities had higher impacts than others located in rural areas. In addition, this approach can be applied to estimate the potential impacts of newly-planned industry projects according to their pre-designed specifications. Thus, the results of impact analysis can provide useful information for supporting the decision-making of the site selection as well as the migration of local residents if necessary.
This research was supported by the 111 Project (No. B14008), by the Program for Innovative Research Team in University (No. IRT1127), the Key Project of Ministry of Education (No. 311013), and the Natural Science and Engineering Research Council of Canada.