7.	HYDROLOGIC AND HYDRAULIC EVALUATION

7.1	Integrated Great Lakes - St. Lawrence River Model

The evaluation of Lake Ontario regulation plans, the practicality of proposed criteria, and the hydrologic impacts on the interests, require computer simulation of water levels and flows of the Great Lakes-St. Lawrence River system downstream as far as Trois-Rivères, Quebec. In its 1993 final report to the IJC, the Levels Reference Study Board recommended among other things, that work continue on upgrading models used for simulation, forecasting and regulation to formulate a comprehensive water supply and routing model that includes the whole basin through Trois-Rivières, Quebec and includes the influence of the Ottawa River. While separate computer models exist for different parts of the system, and substantial progress has been made in the development of a coordinated routing and regulation model for the upper lakes, additional work is required to develop and integrate Lake Ontario regulation plan(s) and St. Lawrence River components into the model to simulate water levels and flows of the entire system. As the focus of the work is on Lake Ontario outflow regulation and since Lake Ontario outflows are regulated on a weekly basis, it is proposed that the simulation time-step be quarter-monthly, which approximates a week and is the period of much of the available hydrologic data. With this time step it is possible to ignore short-term effects such as those caused by winds and transients set up by flow changes. The completion of the coordinated, system-wide regulation and routing model will enable simulations, at a quarter-monthly time scale and an appropriate degree of accuracy for this study, under different regulation scenarios and hydrologic conditions.

To examine the short-term effects (i.e., within a quarter-month) of regulation on the St. Lawrence River upstream of Cornwall-Massena, a 2-D hydrodynamic model will be developed of the river from near Kingston to the Moses-Saunders dam. This model will be used to investigate detailed and short-term effects of flow changes on levels and velocities that would be needed to answer commercial navigation, recreational boating and environmental questions as well as operational hydraulic, hydropower generation and ice formation questions. {Topographic and bathymetric data needed for such a model will be collected as a separate part of the work as described in section 4.2}

7.2	Modeling of the St. Lawrence, Ottawa River and Other Tributaries

Regulation of the outflows of Lake Ontario affect water levels and flows of the St. Lawrence River well downstream of the project. The existing IJC regulation criteria require that water level and winter ice conditions at the Port of Montreal and operations during the annual flood discharge from the Ottawa River be taken into account in regulating Lake Ontario outflows. Operating experience has shown that ice conditions on Lac St. Pierre and spring runoff from downstream tributaries can also affect Lake Ontario regulation. Questions have been posed about the potential impacts that flow variations resulting from Lake Ontario regulation may have on the natural environment downstream. To assess the effects of Lake Ontario regulation and potential changes to the regulation criteria, sufficiently accurate water level and flow modeling of the St. Lawrence River downstream to Trois-Rivières is essential.

A review of the available methods to simulate the hydraulic relationships of the Montreal archipelago and the river downstream to Lac St. Pierre and the development of possible improvements within the coordinated routing model will be conducted. Factors to be examined include the modeling of ice and aquatic vegetation growth on the hydraulics of the channels. Environment Canada, Quebec Region has completed much of the development of a detailed 2-D hydrodynamic model for the St. Lawrence River downstream of Cornwall including the Montreal archipelago. The fine spatial resolution of about 60 m of the existing configuration of the model would entail significant resources for long-term simulations, however, this resolution may be able to be increased for more efficient hydraulic routing. It may be possible to vary the resolution of this model whereby selected critical supply periods could be simulated at a high resolution for flood or habitat impact modeling. To use this model for floodplain delineation, habitat and possibly other criteria review studies, a digital elevation model (DEM) (vertical resolution of ±25cm ) needs to be developed of the floodplain of the St. Lawrence River between Beauharnois and Trois-Rivières. The DEM would be needed for more accurate modeling of the flow-level relationship in the river as needed for habitat studies and in the delineation of flooded areas under the spectrum of expected hydraulic conditions. {This DEM will form part of the work described in section 4.2}. The applicability of this hydrodynamic model will be investigated and, if appropriate, adapted for use in this study.

The time series of hydrologic data needed by the downstream routing model needs to be generated. To accomplish this, sets of recorded and/or simulated outflow data need to be obtained or developed to simulate Ottawa River and downstream tributary outflows consistent with those generated for Lake Ontario water supplies. For the climate change case, a suite of hydrologic models for the tributaries to Lake St. Francis, the Ottawa River, the Richelieu River and several other major tributaries downstream of Montreal need to be obtained or developed to translate the precipitation and temperature data from the climate models into outflows to the St. Lawrence from these basins. The models that are used by one or more of the agencies involved with the Ottawa River regulation group can be used to facilitate the process.

7.3	Great Lakes Supply Scenarios

7.3.1

Generation of Hydrologic Sequences for the Existing Climate

Lake Ontario Regulation Plan 1958-D was developed and tested using historical water supplies to Lake Ontario for the period 1860-1954 adjusted to the then current diversion and hydraulic conditions. Since regulation began in 1960, more extreme supplies have been recorded. They include the low supplies in the mid-1960s, and higher supplies in the 1970s, mid-1980s and parts of the 1990s. As a result, level and flow conditions outside the design range that is reflected in the existing IJC criteria were experienced. With the existing criteria, these situations lead to regulation under criterion (k) with outflow management through discretion of the Board of Control and the Commission. Since the climate factors that produce supply sequences are random in nature, it is unlikely that the historical sequence will ever be repeated. Periods of higher and lower supplies will occur in the future due to the natural variation in climate, even without the effects of anthropogenic increases of greenhouse gases in the atmosphere. To design a regulation plan that would be more useful under a wider range of supplies, a different design approach is needed. To account for this natural variability in supplies, it is proposed that extensive set of synthetic hydrologic sequences be developed based on the statistical properties of existing historical supply and related data sets. A similar approach was used with success in a recent study by Hydro Quebec (Rassam et al, 1992, GLERL 1992) to analyze the spillway facilities at the outlet of Lake St. Francis, however, that work did not include the Ottawa River or downstream tributaries. The Hydro Quebec study synthesized a sample equivalent to 50,000 years. It is proposed that this synthetic data set, that represents the distribution of the potential hydrology, be used for the design and evaluation of the proposed new criteria and Lake Ontario regulation plans.

The first step in this work would be to update (through 1999) the coordinated historical supplies for each of the Great Lakes, flows for the Ottawa River and other downstream tributaries, flows through the major diversions, and flow retardation factors for ice and aquatic vegetation in the connecting channels and the St. Lawrence River. Some of these data (e.g. tributaries between Cornwall and Trois-Rivières) may need to be simulated by hydrologic modeling based on existing precipitation and temperature data. The next step would be to conduct statistical analyses of the structure (e.g., mean, standard deviation, autocorrelation, cross-correlations ) of the data series and review the stochastic models developed for the Hydro Quebec study to determine if they are still appropriate (assuming that work would be made available). New stochastic models that can be used to synthesize the outflows from the Ottawa River and other downstream tributaries, and other needed sample series, will need to be developed. The recent work for Hydro Quebec was conducted by their staff, with the advice and assistance of a group of experts from INRS-Eau and Canadian and U.S. universities.

7.3.2

Climate Change

Water supplies and related hydrologic variables representing the most current climate change scenario(s) resulting from atmospheric change research will be generated and used to test the regulation plan and proposed criteria. A similar effort, now nearing completion for the Great Lakes as part of an IJC reference study, can be used if appropriate. Lacking from the existing work is modeling of the hydrologic impacts of climate change on the St. Lawrence tributaries downstream of Cornwall (e.g., Ottawa River basin). The proposed study would use the hydrologic models obtained or developed as part of section 7.1 to simulate the hydrologic effects of climate change on these basins.

A qualitative assessment of changes due to demographic and other possible factors will be made to illustrate how such changes may affect water supplies and related hydrologic factors and their potential impact on regulation. The development of water supply series that simulate the effects of climate change will be carried out by the agencies (e.g., Environment Canada, GLERL) with experience in this field.

7.4	Review Existing Regulation Plan, Investigate New Techniques

Plan 1958-D, the regulation plan presently in effect, was developed using the recorded sequence of water supplies to Lake Ontario for the period 1860-1954. This same sequence of supplies and a similar regulation plan were used in the mid-1950s to design the channel excavations and structures in the upper St. Lawrence River that would provide the needed levels and flows for navigation, ice management and satisfy the criteria in the IJC orders. It was anticipated at the time that more extreme high or low supplies would lead to level and flow conditions outside the criteria. This has been evident since the mid-1960s, and it has not been possible with the existing plan and channel capacities to satisfy all of the existing regulation criteria with the more recent and different sequence of supplies. The recent review of the regulation plan by the St. Lawrence Board (ISLRBC, 1997) showed that Lake Ontario levels outside the range stipulated by criteria (h) and (j) would occur regardless of the regulation plan, given the extreme supplies experienced in the past 40 years and the other constraints in the IJC orders.

During the Levels Reference Study, a number of changes to Plan 1958-D were investigated. Following completion of the Levels Reference study, further investigations on improved regulation plans to replace to Plan 1958-D were carried out by the International St. Lawrence River Board of Control and in 1997 a new regulation plan called Plan 1998 (previously referred to as Plan 35P) was recommended by the Board. The changes incorporated into Plan 1998 attempted to improve the levels and flows for the major users or interests in system, without causing adverse effects on other interests, all within the constraints of the existing IJC regulation criteria. Since the needs of the recreational boating interests and the environment are not explicitly recognized in the IJC criteria, the regulation plan changes attempted to meet these recreational boating and environmental needs insofar as they did not conflict with the existing regulation criteria. While Plan 1998 was not adopted by the IJC due to insufficient information on the environmental impacts and the Commission's judgement that the plan did not provide sufficient improvement over the existing situation. Nevertheless, Plan 1998 did contain new methods that incorporate the knowledge gained from operating experience and several other technical improvements which should be assessed as to their utility.

To compare water levels and flow conditions with regulation to those that would have occurred without regulation, a model of the pre-project or unregulated Lake Ontario outlet hydraulic relationship will be used along with existing downstream hydraulic conditions. As both ice and relative crustal movement affect the relationship between lake levels and unregulated outflows, these two factors will be considered, in a quantitative manner where possible, in the unregulated condition model. This model will demonstrate the extent that Lake Ontario regulation has affected water levels and flows in the Lake Ontario - St. Lawrence River system.

Plan 1958-D is based on the traditional rule-curve method. New techniques are available in operating multi-purpose water control works. For example, the St. Lawrence Board developed and tested a regulation plan using an optimization technique to take into consideration the needs or preferences of a number of users in the Lake Ontario - St. Lawrence River system. In that model, the Lake Ontario outflow is determined each week based on optimization of the degree of satisfaction for each of the interests with the expected hydrologic conditions.

It is proposed that investigation be made into new potentially advantageous outflow regulation techniques in addition to those considered by the St. Lawrence River Board of Control in their recent work. This work would include a review of regulation methods that use forecasts of supply in their operation and the existing supply forecast methods that are available. Ideally, this will result in a more proactive approach towards regulation.

A review of available hydrologic forecasting techniques for the Great Lakes and the Ottawa River and other downstream tributaries will be made to support the investigation and development of new regulation plans, as well as to assist in the regulation of outflows by the Board of Control under discretionary and extreme conditions development. The existing work and expertise in this area from a number of government agencies (e.g. NOAA, USACE, Env.Can., DFO Hydro Quebec etc.) will be investigated. Adaptation of one or more forecast methods to use in making outflow regulation decisions may be made and tested. The ability to assess the accuracy and usefulness or benefit of forecasts will depend upon the availability of historic data required by the forecast method.

7.5	Iterative Evaluation of Regulation Plans and Criteria

It is proposed that new regulation plan(s) be developed and evaluated to determine to what degree they meet the new or updated criteria developed in the study. If the new regulation criteria are to be satisfied by the regulation plan for the chosen hydrologic design conditions, the criteria and regulation plan may have to be developed in concert. If the new plan does not have to fully satisfy the criteria, the criteria can be set prior to the plan development. If the plan cannot meet all of the new criteria, some method of determining the relative importance of criteria must be developed to use to test plan changes and determine which plan best meets these new criteria. The Analytic Hierarchy Process, for example, could be used to weight the criteria. This would allow the quantitative comparison of different plans. Sensitivity analyses could then be applied to the weights to determine the robustness of the comparisons.

With a larger design supply set using the extensive synthetic hydrologic series, the nature of the criteria may more appropriately be in terms of minimizing or maximizing the frequencies of specific conditions. For example, a hydropower criterion might be to minimize the frequency or total amount of spillage. This approach would be different than setting absolute threshold values that are not to be exceeded.

Since the needs and preferences of the various interests are different and at times in opposition, development of a comprehensive set of criteria and a matching regulation plan satisfying all the interests will not be a simple task. There is a need to demonstrate what levels and flows are physically possible with the current physical regulatory works and channels, through simulation of regulation for the wide range of possible hydrologic conditions. An understanding of the reality and practicability of certain level or flow conditions could help promote better dialogue amongst the interest groups and the acceptance of the needs of others and the eventual needed compromise among the groups. This would be an iterative process likely involving workshops, public meetings, and regulation plan development and testing.

7.6	Study Organization, Costs and Schedule

It is recommended that the Great Lakes - St. Lawrence Regulation Office of Environment Canada and the Buffalo District of the U.S. Army Corps of Engineers be requested to lead this study. These two offices have extensive operating experience related to Lake Ontario regulation.

Other agencies having expertise for this work and recommended to be involved are:

• the Great Lakes Environmental Research Laboratory of NOAA to assist in the area of climate change and Great Lakes hydrologic supply simulation,
• hydrology staff from Quebec Region of Environment Canada, Department of Fisheries and Oceans Canada and Quebec Ministère de l' Environment to assist with development of models of Ottawa River and downstream tributary hydrology and St. Lawrence hydraulics,
• staff from Hydro Quebec with expertise to develop or guide the stochastic models of supplies and flows. Consultants with appropriate expertise could also be used, but the experience that Hydro Quebec has in this field would be most valuable.

Table 7a. Time and Cost Estimate - Hydrologic Model and Evaluations (U.S. $K)

Table 7b. Time and Cost Estimate - Hydrologic Model and Evaluations (Cdn $K)