Recent DSS applications for lake and reservoir management have included pollution control, water supply, and hydropower operation. Henderson-Sellers [1991] reviewed the use of DSS in water quality management for lakes and reservoirs and discussed future directions. A trend toward function and object-oriented programming languages was identified, as well as the future need to address climate change. Gobler and Rossouw [1991] applied a DSS, which included a reservoir eutrophication model, to develop phosphorus control strategies for South African reservoirs. Salewicz et al. [1991] also applied a DSS to a case study in South Africa, evaluating the operation of a multiple reservoir system.
Georgakakos [1991] demonstrated the effectiveness of a DSS for real-time management of the Southeastern U.S. reservoir system. The DSS included optimization and simulation models, as well as data management and interactive graphics systems. During design, emphasis was placed on the potential usage of the DSS within the responsible agencies. Soncini-Sessa et al. [1991] presented a two-level knowledge-based DSS for reservoir management. One purpose of the DSS was to aid model selection and parameter estimation for use on the planning level, in which long-term (strategic) policies are set. Another purpose was to facilitate model use on the management level, in which control decisions are made to meet short- to medium-term goals based on the long-term directives. A prototype DSS was applied to determine optimal policies for Lake Como in Italy.
Simonovic and Grahovac [1991] and Barritt-Flatt and Cormie [1991] described the evolution of a DSS for reservoir and energy management at Manitoba Hydro in Canada. The authors reviewed much of the interaction which occurred during the design and implementation process. Ultimately, systems analysis, computer graphics, and expert systems were integrated to support the user in incorporating flow forecasts, power demand forecasts, energy market forecasts, and system maintenance in making operating decisions. Simonovic and Grahovac [1991] also evaluated the impact of the technology on system operation and identified future needs, which included the use of parallel processing to allow real-time implementation of the DSS. Subsequent work related to this project has included the use of expert systems for model parameter selection ( Simonovic et al. [1992]), interactive computer graphics for expert-system verification ( Bender et al. [1993]), and a decision support system for long-range stream flow forecasting ( Bender and Simonovic [1994]). This case study is perhaps the most well-documented DSS application in the water resources literature.