Taylor & Francis Group
Browse
UAEM_1500060_Supplementary File.docx (3.04 MB)

Monitoring water quality on the central Toronto waterfront: Perspectives on addressing spatiotemporal variability

Download (3.04 MB)
Version 2 2018-10-08, 19:21
Version 1 2018-07-17, 15:31
journal contribution
posted on 2018-10-08, 19:21 authored by E. T. Howell, S. C. Dahmer, K. M. Chomicki, N. Benoit, D. Boyd, G. S. Bowen

Toronto Harbour, adjacent to a large urban centre on Lake Ontario, receives inputs from storm sewers, combined sewer overflows, and urban runoff that lead to contrasting water quality over the waterfront. Toronto’s Inner and Outer Harbours, mesotrophic and meso-oligotrophic, respectively, were investigated in 2008 to assess how water quality conditions were affected by loading gradients, weather and lake circulation. Spatially-intensive measurements of UV fluorescence, turbidity, conductivity, and chlorophyll a, together with lab-based analysis of chemistry at discrete sites, were used to depict patterns and contrasts in water quality in the harbour. Spatially-integrated field sensor data were also employed to examine the efficacy of using discrete water quality sampling to represent average conditions. Nitrogen, total phosphorus, dissolved organic carbon, major ions and E. coli gradients were a recurrent feature among surveys with concentrations decreasing away from the Don River mouth. The limited point-sample data reasonably depicted average conditions among areas of the harbour on the days of survey as did the results interpolated for a long-term monitoring station in the Inner Harbour. The strong variability seen within the Inner Harbour indicates that the most affected water quality conditions are likely under represented by area-wide conditions. Temporal variability in water quality, correlated with the discharge from the Don River, was strong yet under represented by the field-based sampling. Empirical prediction of total phosphorus concentrations in the Inner Harbour, and correlated with Don River discharge, were used to demonstrate both the critical need to address temporal variability in monitoring design and the possibility of using empirical predictive approaches drawing upon field sensor data to fill this gap.

History