Data sources: review and selection

View the most recent version.

Site selection
Parameter selection
Guideline selection
Sample numbers, timing and collection period
Data management, calculation and verification

Water quality data used in the calculation of the freshwater quality indicator in the CESI 2006 report were obtained from a number of existing water quality monitoring programs across the country (Text table 2). These programs are managed by federal departments, provincial departments, and under federal-provincial agreements. They were originally established for many different reasons. Currently, there is no national network of water quality monitoring sites designed specifically for the purposes of reporting the state of Canada’s water quality in a fully representative way at different geographic scales across Canada.

Text table 2  Monitoring programs that provided data on ambient water quality from 2002 to 2004

Each program monitors a specific array of parameters designed to suit the program’s objectives and resource constraints. These monitoring programs track ambient concentrations¹ of major ions² (e.g., chloride and sulphate), nutrients (e.g., phosphorus and nitrogen), metals (e.g., mercury), organic compounds (including pesticides and industrial chemicals), and other parameters (e.g., dissolved oxygen, suspended solids and pH). Sampling frequencies also differ among networks, with program needs, resource constraints, and ease of access to sites being important determinants.

Site selection

For the freshwater quality indicator in the CESI 2006 report, data from 370 sites across all provinces and territories were selected from the available water quality monitoring sites that met the desired sampling frequency for the 2002 to 2004 period. Different sampling frequency criteria were applied to sites in the North and those in the South (see Sample numbers, timing and collection period).

The Great Lakes are treated separately in the freshwater quality indicator because of their disproportionate size and unique surface water quality monitoring program. For the Great Lakes dataset, the WQI was calculated using data collected by Environment Canada’s Great Lakes Surveillance Program. Conducted on a two-year rotation, this program sampled Lake Erie, Lake Huron, and Georgian Bay in April/May 2004, and lakes Ontario and Superior in April/May 2005. Approximately 320 sites were sampled on the Great Lakes, with the sites being divided between the lakes as follows: Lake Ontario (100), Lake Erie (70), Lake Huron and Georgian Bay (90) and Lake Superior (60). Of these sites, approximately 20 percent are sampled for organic parameters. Eighteen water quality parameters and three sediment quality parameters were included in the calculation of the WQI score.

The 62 sites in the acid rain monitoring program in the Atlantic region were grouped into 7 clusters. This was done to reduce the influence on the national indicator of these numerous small and neighbouring sites, all subject to the same specific water quality concern. The sites were grouped into clusters based on their proximity. For the lakes in each cluster, the average WQI score, weighted by lake area, was calculated. The average lake area for each cluster was also calculated. The lake in each cluster with an area and WQI score closest to the average was selected to represent all sites in that cluster. Other sites in the cluster were then cut from the dataset. For those clusters with river sites, one river of average WQI score and flow was selected to represent the rivers of that cluster.

Figure 1  Locations of CESI 2006 water quality monitoring sites across Canada

Parameter selection

The parameters used in the WQI calculations can be linked to the main stressors on water quality across Canada, including urban development, agriculture, forestry, mining, smelting, pulp and paper mills and other industrial facilities, deposition of atmospheric pollutants, and dams (Environment Canada, 2001).

Decisions regarding parameters to use for national reporting of the WQI were made by provincial, territorial, and federal water quality experts. The decisions were based on local knowledge of stressors potentially affecting water quality in the region, or at each site, using available monitoring data for 2002 to 2004. Only parameters relevant to the protection of aquatic life were included. This excludes bacterial counts, for example, which are primarily of concern for human health. For all jurisdictions except British Columbia, a common suite of parameters was applied to all sites within the jurisdiction or monitoring program. Site-specific selections of parameters were made in British Columbia, with four parameters (dissolved oxygen, phosphorus, pH, water temperature) included at each site wherever available (refer to Text table 5 for details regarding the parameters used in each jurisdiction).

Guideline selection

Nationally, guidelines are developed according to the methodology outlined by the CCME science-based protocols for guideline derivation and endorsed by the CCME (CCME, 1991). Some provinces and territories have directly adopted the CCME guidelines for their needs, while others have developed their own guidelines using similar protocols to those of the CCME. Typically, water quality guidelines are based on laboratory toxicity studies showing effects on various aquatic life (fish, invertebrates, plants) from different concentrations of a constituent in the water.

For the CESI 2006 report, calculation of the freshwater quality indicator relied largely on the use of existing water quality guidelines for the protection of aquatic life. Most of the guidelines used are based on chronic exposure. In a few instances, guidelines were applied for short-term exposure.³  Guidelines were selected on a site-specific or jurisdictional basis by teams of regional water quality experts from the suite of generic guidelines available from various sources⁴  and from existing site-specific guidelines for the parameters of local interest (Appendix 1). The principle behind guideline selection is to choose those that are most “locally relevant”, meaning appropriate to local aquatic life. Background levels of naturally occurring substances and other characteristics of water, such as hardness and temperature that can affect the toxicity of some of the substances of concern. It is recognised, however, that generic guidelines (i.e., those not derived for a specific site) are often conservative to provide a high level of protection through the use of uncertainty factors, depending on the quality and availability of toxicological information for the substance. Thus, natural concentrations of some substances may exceed these guidelines.

Site-specific guidelines based on background concentration procedure (CCME, 2003) were used in the Northwest Territories and some Nunavut sites (i.e., rivers). In these cases, the upper range of the local natural background level for selected parameters was statistically estimated and found to be greater than the recommended guideline. The CCME Canadian Water Quality Guidelines were found to be locally relevant and used at a few sites on watercourses near outlets of lakes (e.g., Great Bear Lake, Lake Hazen).

The rapid assessment approach,⁵ another site-specific method for areas with high natural background levels (e.g., turbidity), was used to generate a benchmark based on long-term monitoring data (not toxicity studies). This approach was carried out for many parameters for sites in British Columbia and may be done in future in other areas (e.g., Northwest Territories).

Sample numbers, timing and collection period

Annual fluctuations in meteorology and hydrology can have a considerable impact on water quality and, consequently, on the resulting index ratings when applied for individual years. Thus, ratings were based on three years of data in order to dampen temporal variability and reflect a more general state of water quality. The years 2002 to 2004 were the most recent available years across all monitoring programs.

Minimum sample numbers for the three-year reporting period were established for lake, river and northern sites (Text table 3). Sites that did not meet these minima were excluded from the national reporting of the indicator in the CESI 2006 report.

Text table 3  Sample frequency requirements for WQI application in the 2006 CESI report

In temperate lakes, the water column can become thermally stratified, or layered by temperature, during the summer and winter. Mixed conditions are typical during early spring and late fall. Chemical contaminants can also stratify in lakes, with their concentrations being determined in part by water density, which is in turn determined by water temperature. Lakes were sampled at least twice annually, once in the spring and once in the fall. If these spring and fall samples were not available, several samples were taken at various depths during another season. The results of these samples were weighted by the volume of water at the sampled depths and then averaged. Weighting by volume, however, was not always possible. As a final option, samples were taken at the surface of the lake.

In rivers and streams, surface sampling is generally considered to be representative of the water column, which is normally well mixed. However, sampling may need to be repeated more often throughout the year to better capture water quality variability. The CCME technical guidance document (CCME, 2001) recommended a minimum of four samples per year, accounting for seasonal or hydrological variability, based on the original testing of the index.

In northern and remote locations, routine water sampling can be costly and challenging, as it is sometimes dangerous and difficult to access sites, and weather conditions can be extreme. As a result, monitoring sites are often sampled less frequently. In addition, a sensitivity analysis conducted on several northern rivers revealed that having fewer samples (i.e., 9) than the required minimum (12) in a three-year period did not produce WQI scores that were significantly different (Glozier et al., pers. comm.). For these reasons, the minimum sampling frequency for rivers in the North was reduced from 12 (as used in southern Canada) to 9 for the 2002–2004 period and reported separately. These criteria apply to sites that fall north of a line delineated by McNiven and Puderer (2000).

Data management, calculation and verification

Water quality data from each of the monitoring programs are stored in provincial or federal databases, managed by the respective environment departments. Basic site information (e.g., name and location) and water quality data were extracted from available databases, by regional and provincial data providers, and transferred to “WQI calculators” i.e., spreadsheets programmed to calculate WQI ratings. These calculators allow users to select input parameters, guidelines, and sample periods (with options allowing guidelines to be modified by hardness, pH, or temperature, when appropriate).

Suspected outliers in the datasets were identified and validated by verifying field forms and books to check for accuracy of data entry, by ensuring that reported units were correct, by consulting stream flow and meteorological records, and/or by comparing with the levels of other parameters in the dataset (e.g., turbidity, total suspended solids, major ions) that could explain the unusually high or low values of some parameters. Unless identified as likely erroneous, outliers were left in the dataset.

After validation of the dataset, calculations were verified and then peer reviewed. Environment Canada experts then transferred site information, WQI ratings and details on the application (i.e., data source, parameters, guidelines, sample numbers and dates, and contact information) onto templates for incorporation into a central database. Statistics Canada experts reviewed site data to ensure that the number of samples, timing, and locations met the methodology requirements. This information was then used to generate the freshwater quality histograms and map of monitoring site locations by staff at Statistics Canada, the National Water Quality Monitoring Office, and the Strategic Information Integration Directorate of Environment Canada.

The ratings and calculation methods (i.e., parameters included, guidelines used, site information) compiled into the national database were then verified for each site by each data provider to detect any errors introduced during the integration of this information.

Notes

1. Concentration of substances in the aquatic environment, as opposed to effluent discharges.

2. Positively or negatively charged molecules that occur naturally in water as a result of geochemical weathering of rocks, surface runoff, and atmospheric deposition. The eight major ions — calcium, magnesium, sodium, potassium, bicarbonate, carbonate, sulphate, and chloride — account for most of the total dissolved solids in surface waters.

3. In Quebec, the guideline used for turbidity is for short-term (acute) exposure.

4. Sources include Prairie Provinces Water Board (PPWB), 1992; Ontario Ministry of the Environment (OMOE), 1994; CCME, 1999; Alberta Environment, 1999; BCMOE, 2001; Le ministère du Développement durable, de l’Environnement et des Parcs (MDDEP) du Québec, 2006; Williamson, 2002; and United States Environmental Protection Agency (USEPA), 2005.

5. See Environment Canada, 2006. Technical guidance document for Water Quality Index practitioners reporting under the Canadian Environmental Sustainability Indicators (CESI) initiative – 2006 update. Unpublished draft. Ottawa, Ontario.