The role of regional groundwater systems on the Great Lakes is investigated along Lake Ontario’s north shore. Given that the ORMGP jurisdiction covers much of Lake Ontario’s north shore catchment area, we have been approached to apply our database and numerical models to proportion chloride (salt) loadings from a variety of sources with specific attention to direct groundwater discharge along the lake shore.

Direct groundwater discharge to Lake Ontario (along the shoreline) remains considerably uncertain, with some historical estimates provided by Haefeli (1970; 1972), Singer (1974), Ostry (1979), Ostry and Singer (1981), McCulloch (1973), and Hodge (1978).

Our analysis suggests long-term direct groundwater discharge loadings range from 2-14 kilotonnes/year of chloride to the north shore of Lake Ontario. Indirect loadings via groundwater discharge to streams along the north shore are estimated to contribute an additional 10-50 kilotonnes/year of chloride.

When correcting for drainage area, groundwater-sourced loadings amount to 100-600 kilotonnes/year. For context, chloride loadings from all sources to Lake Ontario (groundwater, surface water, contaminant point sources, etc., yet excluding inflow from Lake Erie) are estimated to range from 1700-2500 kilotonnes/year (Chapra et al., 2009).

read 2022 hard-copy report here

The North Shore

From the Trent river east, the northern catchments drain into the Bay of Quinte, which then mix with Lake Ontario near its easterly outlet. So for the most part, the northern ~225km shoreline that extends from Hamilton to Belleville is here considered the “north shore”.

This leaves only a 6,520 km² drainage area of the 64,030 km² Lake Ontario catchment (10% by areal cover, purple-ish area below), while covering (at least) 22% of Lake Ontario’s entire shoreline.

The North Shore is relatively urbanized being centered on Toronto. With almost half of north shore protected as greenbelt (48%), the remaining land used is split between the existing developed land and the “whitebelt” areas (lands both un-developed and un-protected), 28% and 24%, respectively. The total length of mapped roads within the north shore catchment area (not shown) is a little over 25,000 km.

Land use designations within the North Shore catchment area.


Road salt application

For context, rates of municipal road salt application are 11.6 tonnes per kilometer serviced roads. Accounting for the road length along the North Ontario shore, this amounts to 290 kta road salt, and that’s not counting for private application (residences, parking lots, etc.).

From: Products Division, Environment and Climate Change Canada (2021)


Data collection

As of this snapshot, 131,771 chloride samples were present in the ORMGP database, of which 66,828 (from 2,675 locations) had screen information and were taken from areas assumed to be up-gradient of Lake Ontario.

Locations of Chloride samples grouped by their source. Shaded in grey is the contributing area and stretch of Lake Ontario shoreline being investigated.


Categorization

Locations fell into 4 categories:

  1. deep wells (screened depth>20m)
  2. shallow wells (screened depth<20m)
  3. surface water (mostly PQMN)
  4. unknown (or not yet classified)


And were assigned to the regional formations:

  1. Late Stage Glaciolacustrine-Glaciofluvial
  2. Halton Till
  3. Mackinaw/Oak Ridges
  4. Channel—Silt
  5. Channel—Sand
  6. Upper Newmarket
  7. Inter Newmarket Sediment
  8. Lower Newmarket
  9. Newmarket Till/Northern Till
  10. Thorncliffe
  11. Sunnybrook
  12. Scarborough
  13. Bedrock—Undifferentiated

At every location, the maximum measured Chloride value was collected.

This was (primarily) to counter the complication of handling seasonal concentration regimes:

Van Meter et.al., (2019)


Chloride concentrations

The distribution of all maximum measured Chloride [mg/L] looks log-normally distributed, with a mean of 23.7 mg/L (n = 2,675). Each bin is then colour-coded to the measurement source.

Distribution by source

The distributions are compared using a two-sample Kolmogorov-Smirnov (KS) test. The KS tests whether two distributions can be said come from the same population, this is deemed true when the p-value>0.05.

It appears that only the deep groundwater and unknown sources can be said come from the same distribution. Unknown–surface also show some possibility. It is the shallow water sources that appear to differ the most, which is (interestingly) showing lower concentrations. Normally shallow wells tend to have higher concentrations, however these shallow wells are mostly located to the north, within the Oak Ridges Moraine area up-gradient from major urbanization.

Distribution by formation

In a similar fashion, distributions are presented on the basis of which geological formation the samples were taken from.

In general, the KS test seems to imply that the samples are indistinguishable by formation, as the KS test rarely reject the null hypothesis.

Distribution by sampling date.

Seasonal declines in water quality sample have been observed in winter months (Mazumder et.al., 2021). This decline is apparent with our data, mainly for surface water samples, however there’s no apparent need to reduced the dataset on the basis of sampling date .

Chloride trend

Finally, mapped water quality locations having greater than 100 measurements, spread over 10 years were tested for temporal trends. It quite clear that groundwater chloride concentrations are on the rise.

Numerical modelling

Great Lakes Intake Program

Since the mid 70s, drinking water intakes along the great lakes began collecting continuous water quality measurements at a point when Great Lake chloride levels were at their maximum. Since then there was a steady decline until the mid 90s a slight incline until about 2005, when chloride levels look to steady. The data made available (and shown below) are posted through 2019.

From this, the current steady concentration of Lake Erie \((c_\text{e})\) and Lake Ontario \((c_\text{o})\) are set to 18 and 22.5 mg/L, respectively.

Hydrograph separation

Next, indirect groundwater discharge is estimated using hydrograph separation. Stream flow measurements in the North Shore contributing area are “separated” into their constituent components: groundwater discharge and overland runoff using automated proceedures.

3D groundwater flow models

Numerical groundwater flow models give insight into the proportion of groundwater originating from either direct sources (along the Lake Ontario shoreline) or indirect sources via groundwater discharge to streams.