Statistics Canada
Symbol of the Government of Canada


Warning View the most recent version.

Archived Content

Information identified as archived is provided for reference, research or recordkeeping purposes. It is not subject to the Government of Canada Web Standards and has not been altered or updated since it was archived. Please "contact us" to request a format other than those available.

Appendix 1: Description of the air quality indicator
Appendix 2: Description of the greenhouse gas indicator
Appendix 3: Description of the freshwater quality indicator

Appendix 1: Description of the air quality indicator

The air quality indicators track measures of Canadians’ long-term exposure to ozone and to fine particulate matter (PM2.5), two key components of smog that have been linked to health impacts ranging from minor respiratory problems to hospitalizations and even premature death.

Air monitoring

Canada has a coordinated air monitoring network with stations across the country. The ozone and PM2.5 data used in this report were collected through the National Air Pollution Surveillance (NAPS) Network, a joint federal, provincial, territorial and municipal program focused on urban air quality, and through the Canadian Air and Precipitation Monitoring Network (CAPMoN), a network operated by Environment Canada that measures rural and remote background levels of air pollutants.

The data collected through NAPS and CAPMoN are subject to strict quality assurance and quality control standards to maintain national consistency. In addition to audits by provincial, territorial and municipal jurisdictions, NAPS sampling stations are also subject to federal audits. This ensures that the data stored in the NAPS database are of the best possible quality.

Stations were grouped into regions by Environment Canada. These regional groupings have changed from previous reports to improve geographical representivity. Stations in eastern Ontario are now grouped with stations in southern Ontario, rather than with the ones in Quebec. Thus, indicator levels for these two regions in this report are not comparable with those in previous reports.

Map A.1 Locations of monitoring stations contributing to the air quality indicators – national and regional. Opens a new browser window.

Map A.1 Locations of monitoring stations contributing to the air quality indicators – national and regional

Ground-level ozone

From 1990 to 2005, 260 monitoring stations across the country reported hourly concentrations of ozone. Data sets from 76 of these stations were sufficiently complete for this period to be used for the national trend analysis (Figure 1). The measurement error for hourly ozone concentrations at individual sampling stations is estimated to be ±10% (Dann and Conway 2005). 

Fine particulate matter

From 2000 to 2005, 162 monitoring stations reported hourly observations for fine particulate matter (PM2.5) concentrations across the country. Data sets from 65 of these stations were sufficiently complete for this period to be used for the national analysis (Figure 3). The measurement error for hourly PM2.5 concentrations at individual sampling stations is estimated to be ±20% (Dann and Conway 2005). 

Monitoring of PM2.5 started in 1984 in only a few Canadian cities, using a good but labour- and resource-intensive filter sampling method. Gravimetric analysis was conducted by passing air through a size-selective filtering medium which was then collected and sent to a certified laboratory for manual weighing. Other methods that continuously monitor and provide in-situ real-time hourly PM2.5 data became available in the mid-1990s and have gradually been deployed to many more sites across Canada. Hence the exposure indicator trend analysis begins in 2000. A comparative analysis between manual weighing and the new automated methods shows good agreement during the warm season.

Computing the exposure indicators

The ozone exposure indicator was calculated using the following steps. For each given station, the ozone was averaged over a running 8-hr period. For a calendar day, this procedure gives 24 8-hr average readings. From these 24 8-hr readings, the daily maximum was then retained. These daily maxima were then averaged over the entire warm season (April 1 to September 30). Finally, these station warm-season averages were averaged nationally or regionally with the value for each station being population-weighted to provide the yearly national and regional exposure indicators covering the period from 1990 to 2005. 

The PM2.5 indicator was calculated on a yearly basis as follows. For each given station, hourly concentrations of PM2.5 were first averaged over a 24-hr period (midnight to midnight), which represents the commonly used unit for assessing exposure to PM2.5. These daily averages were then averaged over the entire warm season (April 1 to September 30). Finally, these station warm-season averages were averaged nationally or regionally with the value for each station being population-weighted to provide the yearly national and regional exposure indicators covering the period from 2000 (the first year that the monitoring data was sufficiently extensive) to 2005. 

Population-weighted concentrations

Population-weighting the average concentration at each station puts more emphasis (or "weight") on the levels in the more populated areas, thereby providing a better indication of the ozone and PM2.5 levels to which a greater proportion of the population may have been exposed.1

The warm-season average concentration (Cn) at a given monitoring station was then multiplied by the population (Pn) living within a 40-km radius of the station (Cn*Pn). All the considered Cn*Pn products were then added together and divided by the total considered population, giving the CESI exposure indicator:

Formula 1


C1 = the warm-season average concentration of the daily maximum 8-hr ozone for the ozone exposure indicator, or the warm-season average concentration of the daily 24-hr average concentration for the PM2.5 exposure indicator at station 1, and

P1 = the population living within a 40-km radius of station 1.

Trend computation

The values of the exposure indicators can vary annually. Despite these annual variations, the value may experience an overall increasing tendency, a decreasing tendency or no tendency at all. This overall tendency is estimated by the slope of a straight line fitted through the actual values of the indicators. The slope of this line and its direction of change is what is meant by the "trend." 

Non-parametric statistical tests were conducted to examine the direction and the magnitude of the annual rate of change in the air quality indicators. The standard Mann-Kendall trend test was used to determine the direction of the yearly changes, and the Sen trend slope estimator was used to assess the magnitude of the observed rates, and also to test whether the slope obtained was statistically different from zero at the 90% confidence level. The Sen method is a non-parametric linear slope estimator commonly used in environmental statistics with time series data.

Interpretation of the trend and statistical significance

For the exposure indicators, trends are only reported if the slope is statistically different from zero. If the slope is not statistically different from zero, it means that a slope of zero is one possibility; as such, there may be no upward or downward trend in the values and any annual variations in the values of the indicator are therefore likely due to random errors alone. No test for the stability of the exposure indicators was conducted.

Interpretation of trends in ozone and PM2.5 exposure indicators should give careful consideration to the slope of the trend lines. The magnitude of statistically significant trend slopes may not always be environmentally important when compared with detection limits, background levels and air quality standards.

In the case of the air quality indicators, studies indicate that adverse health effects can occur even with low concentrations of these pollutants in the air (WHO 2005). As a result, an increase in trend slopes of these indicators, regardless of their magnitudes, may signal the potential for increased health risk. 

Further details on the air quality indicators are provided on the Government of Canada website and the Statistics Canada website.

Appendix 2: Description of the greenhouse gas indicator

The greenhouse gas (GHG) emissions indicator, related data and trends information come directly from Canada's National Inventory Report, 1990–2005 – Greenhouse Gas Sources and Sinks in Canada (Environment Canada 2007a), an annual report submitted by Environment Canada as required under the United Nations Framework Convention on Climate Change (UNFCCC). Greenhouse gas emissions are estimated according to the procedures and guidelines prescribed by the Intergovernmental Panel on Climate Change (IPCC) and are reviewed annually by a United Nations expert review team. The indicator estimates Canada's total annual anthropogenic (human-induced) emissions, released into the atmosphere, of the six GHGs covered under the Kyoto Protocol (see Chapter 3).

The total emissions estimate is calculated by adding the individual estimates for each of the six gases. The individual estimates are all converted to an equivalent amount of carbon dioxide by multiplying the estimated emissions for each gas by a weighting factor called "global warming potential" (GWP) that is specific to that gas. This potential represents the amount of warming over 100 years that results from adding one unit of the gas to the atmosphere, compared with the effect of adding one unit of carbon dioxide. The GWPs for the six greenhouse gases under the Kyoto Protocol are as follows:

  • Carbon dioxide: 1
  • Methane: 21
  • Nitrous oxide: 310
  • Halofluorocarbons: 140 / 11 700
  • Perfluorocarbons: 6500 / 9200
  • Sulphur hexafluoride: 23 900

The emissions for each GHG are estimated by summing the individual estimates for different activities. In general, measurements of the amount of activity (e.g., kilometres driven or amount of a given product manufactured) are multiplied by the emissions per unit for that activity. Estimates of emissions per unit of activity, also known as emission factors, are based on measurements of representative rates of emission for a given activity level under a given set of operating conditions (U.S. EPA 1996). Some emission factors can be calculated for individual industrial facilities; most, however, are more general and are derived from national or international averages.

The indicator does not include emissions from naturally occurring sources (e.g., organic matter decay, plant and animal respiration and volcanic and thermal venting) or the absorption of emissions by natural sinks such as forests and oceans. Emissions and removals from some types of land, such as forests and wetlands, and changes in land use are excluded from the indicator as well.

Environment Canada's Greenhouse Gas Division developed and compiled emission and removal estimates using data from several sources, including Statistics Canada (statistics on energy, transport, livestock, crop production and land), Natural Resources Canada (statistics on mineral production and forestry) and Agriculture and Agri-Food Canada (some agricultural parameters), and other sections of Environment Canada (data on landfill gas capture, hydrofluorocarbon and perfluorocarbon use, ozone and aerosol precursors). Environment Canada engineers and scientists estimate emissions using methods developed by IPCC as well as methods and models developed in-house specifically for estimating Canadian emissions.

Emissions estimates for the various sectors are also reviewed by experts from the organizations that provided the source data, such as Statistics Canada, Natural Resources Canada and Agriculture and Agri-Food Canada. Finally, the information submitted by Canada each year to the UNFCCC Secretariat is subject to external review by a team of experts, and a report of their findings is published by the UNFCCC. The inventory underwent an in-depth review in Canada in 2003, and a "desk" review in 2004 and 2005.

Sources of uncertainty in the estimated emissions include the definitions of the activities that are incorporated in the estimates, methods for calculating emissions, data on the underlying economic activity and the scientific understanding. Uncertainty information is used to set priorities to improve the accuracy of future inventories and to guide decisions about improvement of the estimation methods. The uncertainty about estimates for individual gases, individual sectors or specific provinces will be higher than for the overall national estimate.

Quality assurance, quality control and verification procedures are part of the preparation of the inventory. They take the form of internal checks and external reviews and audits, following international standards. Activities based on these reviews are intended to further improve the transparency, completeness, accuracy, consistency and comparability of the national inventory. The detailed documentation, international reporting guidelines, domestic and international scrutiny and reliance on Statistics Canada energy survey results all contribute to the quality of the GHG estimates.

The complete National Inventory Report, 1990–2005 – Greenhouse Gas Sources and Sinks in Canada is available upon request.

Further details on the GHG emission indicator are provided on the Government of Canada website and the Statistics Canada website.

Box A.1. Statistics Canada’s Greenhouse Gas Emissions Account

Statistics Canada’s Greenhouse Gas Emissions Account forms the basis for Figure 8. Produced following the concepts of the System of National Accounts, it uses many of the same basic data as the GHG inventory compiled by Environment Canada; however, the information is recast into the commodity and industry framework of the System of National Accounts so that the emissions data can be used for economic modelling. In particular, this linkage permits use of Statistics Canada’s national input–output accounts to analyse the interplay between production and consumption of goods and services and the GHG gas emissions that result from those activities. Emissions from the production of goods and services are attributed via the input–output model to the final purchaser.

Statistics Canada’s Greenhouse Gas Emissions Account provides emissions estimates for 119 industries and two categories of household expenditure. In addition to the detailed emissions data produced by sector, several environment–economy "intensity" indicators are derived from Statistics Canada’s Greenhouse Gas Emissions Account, including the GHG intensity of gross industrial output, the GHG intensity of household consumption and the GHG intensity of net exports.

Emissions factors from Environment Canada are applied to Statistics Canada’s Energy Use Account data (which are also based on the System of National Accounts industry and commodity frameworks). The Energy Use Account data come mainly from Statistics Canada’s Industrial Consumption of Energy Survey, transportation surveys, the Report on Energy Supply–Demand in Canada and Natural Resources Canada’s Census of Mines. Additional estimates of emissions that are not linked to fossil fuel consumption are taken directly from the Environment Canada GHG inventory and are applied to the appropriate industries in the System of National Accounts.

The final demand categories outlined in Figure 8 can be defined as follows:

  • Exports: receipts from other provinces and territories or from abroad for sales of merchandise or services. The barter, grant and giving of goods and services as gifts would also constitute exports.
  • Gross fixed capital formation (subdivided into "Construction" and "Machinery and equipment"): the value of a producer’s acquisitions, less disposals, of fixed assets during the accounting period plus certain additions to the value of non-produced assets (such as subsoil assets or major improvements in the quantity, quality or productivity of land) realized by the productive activity of institutional units.
  • Government net current expenditure: economic activities of the federal government (including defence), the provincial and territorial governments, local or municipal governments, universities, colleges, vocational and trade schools, publicly funded hospitals and residential care facilities, and publicly funded schools and school boards.
  • Inventories: consist of stocks of outputs that are still held by the units that produced them prior to their being further processed, sold or delivered to other units or used in other ways, and stocks of products acquired from other units that are intended to be used for intermediate consumption or for resale without further processing.
  • Personal expenditure: represents the purchases of commodities, commodity taxes, wages and salaries and supplementary labour income of persons employed by the personal sector. Includes individuals, families and private non-profit organizations.

Appendix 3: Description of the freshwater quality indicator

The national freshwater quality indicator is based on the Water Quality Index (WQI), which is endorsed by the Canadian Council of Ministers of the Environment (CCME 2001). The WQI is described further on the CCME's website.

In this report, the WQI was calculated for 359 locations in southern Canada and 36 locations in northern Canada for a total of 395 sites. These sites were further grouped by Canada’s major drainage areas. In the 2006 Canadian Environmental Sustainability Indicators (CESI), the WQI was reported for 370 locations nationwide, with 340 in southern Canada, 30 in northern Canada, as well as for 7 basins in the Great Lakes.

The set of monitoring sites was assembled from existing federal, provincial, territorial and joint water quality monitoring programs (Map A.2). These monitoring sites were established for many different reasons, including regulatory requirements, compliance with interprovincial or international agreements and the need to manage local water quality issues. For example, some small lakes in the Maritimes are being monitored because they are located in acid-sensitive areas.

The monitoring sites included in the calculation met the minimum requirements for the timing of the sample collection (2003 to 2005) and the number of samples taken (four per year for rivers and two per year for lakes during spring and fall turnover, over the three-year period). Most of the sites were located in southern Canada and were potentially affected by human settlements, farms, industrial facilities and dams, as well as acid precipitation. Consequently, the monitoring sites are not statistically representative of Canada as a whole. Most were originally chosen for monitoring because they are in areas where there is concern about the effects of human activities on water quality. Saskatchewan, northern Ontario and northern Quebec are large areas that currently have little or no representation in the water quality indicator.

The minimum sample requirement was reduced for sites in northern locations to reflect the reality of water quality sampling in northern Canada and to allow more sites to be included in the indicator for this reference period. Sensitivity analysis showed that the reduction of sample requirements in this case did not impact the WQI results significantly.

Running waters included in this analysis range from small streams such as Prince Edward Island's Bear River, which has an average flow of 0.3 m3/sec and drains an area of about 15 km2, to powerful rivers such as the Mackenzie, which discharges 9910 m3/sec and drains an area of about 1.8 million km2 (MRBB 2004). The lakes also vary considerably in size—from Glasgow Lake (0.24 km2) in Nova Scotia's Cape Breton Highlands to Sipiwesk Lake in Manitoba (454 km2) (Natural Resources Canada n.d.).

Map A.2 Water quality monitoring stations, Canada, 2003 to 2005. Opens a new browser window.

Map A.2 Water quality monitoring stations, Canada, 2003 to 2005

The range of water quality parameters incorporated into the WQI calculations includes

  • nutrients (e.g., phosphorus and nitrogen);
  • metals (e.g., arsenic and zinc);
  • physical characteristics (e.g., pH, dissolved oxygen and turbidity);
  • major ions (e.g., chloride and sulphate); and
  • some organic compounds (e.g., pesticides).

Different subsets of these parameters were selected and applied either uniformly throughout different jurisdictions and regions or, in the case of British Columbia, at individual sites. Generally, Environment Canada and its provincial and territorial counterparts chose which parameters to use in the calculation, based on which parameters had been measured, the human activities of concern and the availability of suitable water quality guidelines. The choices were made by drawing on local knowledge and advice provided by provincial, territorial and federal water quality experts. The parameters used in the WQI calculations reflect some of the main stressors on water quality across Canada noted above. Water quality guidelines were selected from national, provincial and site-specific sources.

Additional work will be required on several aspects of the freshwater quality indicator, such as the representation and distribution of sites across the country, the consistency with which parameters are used in the calculations, the implementation of locally relevant water quality guidelines and the development of water quality trends. How different parameters are combined to produce the index values will also be reviewed and refined.

Further details on the water quality indicator are provided on the Government of Canada website and the Statistics Canada website in the Data Sources and Methods report.


  1. This approach is similar to but more general than the pilot method used for the National Round Table on the Environment and the Economy (NRTEE 2003) discussion paper on the Environment and Sustainable Development Indicators, prepared at Statistics Canada.