Conclusion

View the most recent version.

The Canadian Council of Ministers of the Environment Water Quality Index is the main surface water quality indicator used on a national scale. It is therefore important to learn more about this indicator by performing various analyses to understand and explain its behaviour in order to be able to use it better in the future.  To do this, we used four sets of real data thanks to cooperation of certain provinces. We also performed a simulation study to observe the index behaviour in presence of a considerable number of parameters and samples using synthetic data.

It is important to understand that the conclusions we will draw depend on the findings from the four datasets made available to us.

First of all, we studied the behaviour of each parameter by means of a descriptive analysis. The descriptive analysis has shown that: a) the distribution of the majority of the parameters is log-normal the vast majority of parameters followed a log-normal distribution; b) a small number of parameters had a large variability, which could increase the value of the third term of the index; c) for some parameters, it was more difficult to respect the guidelines that for the others (aluminum, lead and pH for the Newfoundland and Labrador dataset, phosphorus and cadmium for the Ontario dataset, cadmium for the British Columbia dataset and phosphorus for the Quebec dataset), which may yield lower index values; d) there are few significant relations between the parameters considered in pairs, and therefore there is no redundancy in the index calculation; and e) the majority of parameters have great importance in the index calculation.

We then studied the index behaviour by means of a sensitivity study. This analysis has shown that: a) the index is sensitive to the number of parameters and samples used to calculate it and therefore the higher the number of parameters/samples, the lower the index value; b) the contribution of the first term is much more significant that the contributions of the other two terms, and that the contribution of the second term is the least important; c) the correlation is strongest between the second and the third term of the index and generally the correlation between terms considered in pairs depends on the number of parameters and samples used in the calculation; d) use of three samples instead of four per year does not have much influence on the index categorization result; e) there is possibly good index variability within certain stations, regardless of the number of samples used in the calculation; and f) the season has an impact on the index calculation.

The results of the simulation analysis are consistent with the results obtained for the four sets of real data.

Moreover, during the study we noted that the guidelines used by water quality experts are not always the same depending on the province and this may have an impact on the index calculation.

In conclusion, the Canadian Council of Ministers of the Environment Water Quality Index is of interest since it includes many issues relating to water quality and at the same time facilitates presentation of results to the public at large. However, the index depends of many components (the number of parameters and samples, the choice of parameters selected, the season, the guidelines), which have an impact on its calculation. Depending on what is used in the index calculation, difference values can be obtained for the same station. Therefore, it is essential that a sampling protocol for these components be established and respected so that the index can be used in the best possible way in order to ensure comparable results.