The Don River has been called the “most urban river in Canada” because of its position at the centre of the largest metropolitan region in the country (Toronto, Ontario, Canada), and its nearly completely urban watershed. The river, and the Lake Ontario harbour into which it flows, has a heavy metal legacy from anthropogenic activities around the watershed. While sanitary sewage is routed to treatment plants outside of the watershed, only about 20 per cent of the total stormwater flows are routed through management facilities such as ponds. Deposited sediments have been significantly enriched with copper, lead, and zinc, and heavy rainfall events have resulted in an increase of up to four times the base level loading. Transport from the river has accounted for over 90 per cent of the total load to the harbour which has been identified by the International Joint Commission as a polluted Area of Concern. In order to support policy and decision-making, the objectives of the study were to assess the sampled spatial and temporal variability of metals in the Don River and subsequently to develop a modelling strategy to describe flood metal transport dynamics.
A number of data sets were used to characterize the Don River watershed metal concentrations. For assessing concentration trends in water over time, the Ontario Surface Water Quality Monitoring Network station at a Water Survey of Canada gauge on the Lower Don River and other available data sets were used. Additionally, one time grab samples were collected in April 2014 at locations throughout the Don River and its tributaries to improve the spatial coverage of the available data. For assessing concentrations in bed sediment, the best record came from sediments dredged from the Keating Channel, which connects the river to the harbour. The Toronto Region Conservation Authority (TRCA) has measured metal concentrations in these sediments since 1987. These results were compared with other data sources such as grab samples taken in the winter of 2008 by the Ontario Geological Survey and sediment samples taken in April 2014 at the same locations as the water samples for this study.
Watershed hydrology was simulated using a calibrated Storm Water Management Model (SWMM) provided by the TRCA. Eight rain gauges maintained by the TRCA were used to supply the rainfall data to the model. For the study, the hydrologic model was further validated using a five-month period, and the pollutant buildup/washoff modules were activated to estimate total suspended solids (TSS) and associated metal loads to the study reach. The pollutant modules required a land use layer to be added, and classes that included residential, commercial, industrial, government and institutional, parks and recreational areas and open areas. All land use areas were assigned TSS buildup and washoff parameters. Metal concentrations in surface water runoff were simulated using the pollutant co-fraction approach. This approach assumes that metals are associated with TSS and requires the definition of co-fractions of the TSS load that represent the metal loads. Cofractions were calculated for each metal using measured data.
A model setup tool was developed to link the SWMM model with the Environmental Fluid Dynamics Code (EFDC) to allow a seamless transition from catchment hydrology to study the in-stream processes in the Lower Don River. This roughly nine km long section of the channel has been heavily altered due to the growth of the city around it. EFDC was selected for hydraulic routing based on its suitability for modelling the dynamics of metal transport in rivers. For the current study, a tool called the SWMM to EFDC model setup tool (STEMS) was written in MATLAB™ to initialize the EFDC model using the output files and geospatial layers from PCSWMM™. Sediment transport simulations for the Lower Don River were performed with an active sediment bed, using a total of 105 active cells. Various combinations of sediment parameters were tested in EFDC to obtain a reasonable agreement of modelled TSS concentrations with the observed data. Parameterization of the EFDC model was also required to simulate the processes of metal sorption and transport, including the use of partitioning coefficients to model the sorption processes.
The SWMM model showed a good agreement with the observed hydrologic data for the validation period, and the EFDC model matched well with the SWMM model results and the measured data.
While the sampling data for metal concentration in the Lower Don River is relatively sparse, the EFDC model captured the range of TSS concentrations better than the SWMM model. Largely as a result of the improved sensitivity of TSS, metals concentrations were also more sensitive to flow in EFDC (Figure 1). The results also showed that certain areas of the creek are characterized by significant sediment deposition and may currently be hotspots with high metal concentrations. Future sampling should be more spatially targeted to assess the distribution of metal concentration as a function of sediment deposition and erosion within the river.
The study showed that metals pollution in the Don River is improving for some metals respects while remaining flat or getting worse for others (Figure 2). Lead seems to be a success story. Sediments were highly polluted in the late 1980s and early 1990s, but more recently dredged sediment in the Keating Channel shows that the channel has progressively been cleaned up to the point where lead concentrations in most samples are now at or below the Lowest Effect Level (LEL) for invertebrates. Lead concentrations are typically below Canadian Water Quality Guidelines (CWQG) and are relatively insensitive to flooding. This indicates that the sources of lead in the catchment are not strongly correlated with overland flow.
Zinc similarly shows a reduced concentration in sediment, although the reduction is less significant than for lead. Sample medians for zinc remain above LEL thresholds, and the concentrations in most water samples are below the CWQG. Of the three metals, zinc has the highest correlation with flow, suggesting that it is being added to the system from overland flow or through resuspension within channels.
Copper clearly remains a problem. Concentrations in sediment samples at the mouth of the river sometimes exhibit severe impairment levels, meaning that this pollutant affects most of the invertebrate community; and median concentrations in the water typically exceed the CWQG. The distribution of copper in all of the channels indicates that the problem is widespread, and the comparison with samples indicates that the problem has become worse over the last decade in the upper watershed.
The Don River watershed faces significant environmental challenges due to urbanization and the associated flooding, erosion, and degradation of water quality. Lead pollution in the Don River is less of a problem than it was in past decades. Policy changes and sediment dredging in the mouth of the river have reduced lead concentrations in the water and sediment to the point where they no longer exceed guidelines except occasionally during floods. Zinc and copper pollution are increasingly a problem in the Don River, with copper in particular exceeding lower guidelines throughout the watershed and severe effect guidelines during floods and at hotspots within the watershed.
Sediment transport in flashy urban rivers is intermittent, meaning that most of the transport will occur in a relatively short period of time. Metals associated with sediment are likely to be transported in short bursts, both as they are flushed off of land surfaces and re-suspended in the stream. Targeted monitoring programs are needed to accurately capture and model such dynamics in an urban river system.
The STEMS tool successfully integrated the EFDC hydrodynamic model with the SWMM hydrologic model. STEMS provides an efficient setup process for the EFDC model based on the results of the SWMM model and can be applied on any river network where a SWMM model is available. This approach is recommended to take advantage of the more accurate representation of instream hydraulic and chemical processes in EFDC.
Mansoor, S. Z., Louie, S., Lima, A. T., Van Cappellen, P., MacVicar, B. (2018). The spatial and temporal distribution of metals in an urban stream: A case study of the Don River in Toronto, Canada. Journal of Great Lakes Research, 44, 1314-1326.
Contact: Bruce MacVicar, Department of Civil and Environmental Engineering
For more information about WaterResearch, contact Julie Grant.