Analysis of TransAlta Reservoir Contributions to North Saskatchewan River Levels at Edmonton

1 The North Saskatchewan River Basin

The North Saskatchewan River drains 28,110 km²^{[1]Water Survey of Canada, station 05DF001 metadata. wateroffice.ec.gc.ca} of the Canadian Rockies and foothills before reaching Edmonton. Its headwaters lie at the Saskatchewan Glacier in Banff National Park,^{[2]The Canadian Encyclopedia, 'North Saskatchewan River.' thecanadianencyclopedia.ca} and it gathers water from dozens of tributaries as it flows east across the foothills. Two large reservoirs -- Abraham Lake (Bighorn Dam) and Brazeau Reservoir -- sit astride the upper basin and intercept roughly one-third of the total drainage area. The remaining two-thirds is fed by uncontrolled tributaries: the Clearwater, Ram, Nordegg, Cline, Siffleur, and Mistaya rivers, along with numerous smaller ungauged streams.

Drainage Area Breakdown

Drainage areas: Water Survey of Canada station metadata^{[1]Water Survey of Canada station metadata. wateroffice.ec.gc.ca}

Although the two dam catchments represent only one-third of the total drainage area, they exert an outsized influence on Edmonton's river levels because they store and release water on demand. The natural tributaries, by contrast, follow the seasonal snowmelt cycle and are essentially dry by September.

2 TransAlta's Dam Infrastructure

TransAlta Corporation operates two major hydroelectric facilities on the upper North Saskatchewan system. Together, these dams have a combined generating capacity of 470 MW and can store 1.9 billion cubic metres of water -- enough to fill approximately 760,000 Olympic swimming pools.

MW-to-Discharge Calibration

Because TransAlta does not publish real-time discharge data from its dams, discharge must be inferred from AESO generation records.^{[10]Alberta Electric System Operator (AESO). aeso.ca} These calibrations were developed using 564 days of concurrent data (May 2016 through November 2017) where both AESO generation records and Environment Canada discharge measurements were simultaneously available at each dam's outflow gauge.^{[1]Water Survey of Canada station metadata. wateroffice.ec.gc.ca} Both relationships explain more than 93% of the variance (R² = 0.938 for Bighorn, R² = 0.969 for Brazeau), confirming that generation output is a near-deterministic proxy for water release volume.^{[9]Authors' analysis of WSC and AESO data. See Methodology appendix.} Full calibration details are available in the Methodology appendix.

3 The Core Finding: Dam Flow Contribution

This is the central result of the analysis. Using 10 years of concurrent gauge records (October 2016 through June 2026) and generation data, we calculate the fraction of Edmonton's total river discharge that originates from TransAlta's two dams versus natural (uncontrolled) tributaries.

4 Tributary Analysis

To understand why dam contributions dominate so heavily by late summer, it is useful to examine the natural tributaries individually. Each follows a similar pattern: peak flows during spring snowmelt (May-June), rapid decline through July, and near-baseflow conditions from August onward.

Major Natural Tributaries

Clearwater River (2,257 km²). The largest uncontrolled tributary. It drains the front ranges east of the main divide and provides meaningful flow through June and July, but drops to low levels by August. Even at its peak, it contributes only about 8% of the total basin area.

Whirlpool Point headwaters (1,921 km²). This gauge captures the mainstem North Saskatchewan upstream of the Bighorn Dam influence. Fed by glacial melt from the Columbia Icefield and surrounding peaks, it provides a base-level contribution that is relatively steady but modest in absolute terms.

Ram River (1,851 km²). A foothills tributary with a flashy, snowmelt-driven hydrograph. By September its median flow contribution to Edmonton is negligible.

Nordegg River (868 km²). A smaller foothills stream. Peak contributions occur in June; by fall it runs at baseflow.

Cline River (826 km²). Drains the area between Abraham Lake and the Siffleur. Its flows are captured entirely upstream of the Bighorn Dam gauge, making it part of the "dam-controlled" system in practice, though not behind the dam itself.

Siffleur River (512 km²). A small headwaters tributary. Its contribution is proportional to its drainage area and follows the standard snowmelt curve.

Mistaya River (248 km²). The smallest individually gauged tributary. Glacier-fed with relatively stable flow, but its small catchment limits its absolute contribution.

5 Current Seasonal Levels vs Targets

Using 10 years of 5-minute level data (October 2016 through June 2026) at the Edmonton gauge (Environment Canada station 05DF001), the following monthly medians establish the baseline performance of the river under current dam operations.

The chart reveals that the river currently meets the 3.75 m target only in June and July. By August, levels have already dropped below target, and by October the median is nearly 0.65 m below the recommended threshold. This pattern tracks directly with the decline in natural tributary flows described in Section 4.

Rating Curve: Discharge to Level

The relationship between discharge and river level at Edmonton is non-linear but well-characterized. The following table, derived from 10 years of concurrent observations (October 2016 through June 2026), provides the translation between flow rate and gauge height.

The key row is 275-300 m³/s, which corresponds to a mean level of 3.75 m. This confirms the discharge target: approximately 280-290 m³/s of total flow at Edmonton is needed to sustain the recommended river level.

6 What It Takes: Three Target Scenarios

We evaluate three possible level targets to illustrate the range of feasibility. Each scenario calculates how much additional dam discharge would be required beyond current median operations, the total volume of water involved, and what fraction of the combined reservoir storage it represents.

Days at or Above Target (current operations)

At the recommended 3.75 m target, the river is currently at or above that level only 40% of season days -- and a mere 4% of October days. A reliable river for navigation and recreation requires a significantly higher percentage.

Additional Dam Discharge Required

7 Recommended Generation Schedule

To achieve the 3.75 m target, the following monthly generation levels are recommended for each dam. "Now" values represent current monthly median generation from AESO records; "target" values are calculated from the MW-to-discharge relationships calibrated in Section 2.

The most significant increases are needed at Brazeau Dam, particularly in September (+106 MW) and October (+148 MW). No change is needed in June or July, when natural flows already sustain the target. Bighorn's increases are modest throughout. The Brazeau increases reflect the fact that TransAlta currently scales back generation in fall — precisely when downstream need is greatest and natural tributaries have dried up.

8 Summary and Call to Action

Appendix: Data Sources and Methodology

Hydrometric data accessed via wateroffice.ec.gc.ca. AESO data accessed via aeso.ca.

R² Regression Analysis

Five regression models were tested to quantify how well different combinations of upstream inputs predict Edmonton's discharge:

Model E, incorporating all available upstream inputs, explains 98.3% of the variance in Edmonton's discharge with an RMSE of 20.6 m³/s. Model D, using the Highway 759 gauge alone (which captures both dam and natural flows downstream of the foothills), achieves 97.2% -- confirming that this intermediate gauge is a near-deterministic predictor of Edmonton conditions. Model A shows that dam outflows alone, with no other information, explain over half of Edmonton's discharge variance.

For a full discussion of methodology, calibration techniques, and limitations, see the Methodology & Data Sources page.