Author(s)

L. Fagherazzi, D. Fay & J. Salas

Abstract

The hydrological data used in the design, planning and operational studies of water resource schemes are often limited to historical records, which are usually short, incomplete, sparsely distributed in space, and poorly synchronized. Moreover, a particular sequence of flow observations rarely reoccurs in identical form in a future period. If important characteristics of the historical series such as annual sequences of low and high flows are poorly modeled, design and/or planning shortcomings may then result, because their effects were underestimated. To better grasp drought and flood variability, the Lake Ontario – St. Lawrence River Study used both historical and synthetic hydrological scenarios. A set of stochastic hydrological Net Basin Supplies series for the Great Lakes, and local inflows of the St Lawrence River’s major tributaries upstream from Trois-Rivieres sequences was developed, based on the statistical properties of the historical series. The set constitutes a large number of potential hydrological scenarios that could occur due to the observed natural variation in climate. The simulated series were then used, along with the observed 1900-2000 data, to design and evaluate the adequacy of newly proposed Lake Ontario multiobjective management strategies. In addition, four hydrologic scenarios incorporating climate change fields on seven climate variables were developed for the Great Lakes and Ottawa River. Scenarios from General Circulation Model (GCM) output changes, covering the range of climate change variability for the future 30-year period 2040-2069, were used to test the robustness of the management strategies under possible future climates. This paper summarizes the special hydrological characteristics of the Great Lakes – St. Lawrence River System, the major modeling hypothesis and retained strategies used to prepare a stochastic generation model, and the preparation of climate change scenarios for a major tributary, the Ottawa River basin. It also illustrates their use in evaluating Lake Ontario multiobjective management strategies. Keywords: stochastic generation model, shifting mean and ARMA process, climate change scenarios, Lake Ontario regulation plans.

Keywords

stochastic generation model, shifting mean and ARMA process, climate change scenarios, Lake Ontario regulation plans.