Journal of Water Resources Planning and Management

Planning is an integral part of water resources development and management. Whether or not particular plans or programs are eventually implemented, the planning process itself forces us to think about what we are or should be doing to address a particular set of problems or needs. The planning process should lead to a better understanding of what will happen if we do or do not act and, if we decide to do something, which of many possible actions is likely to be the best. Such planning requires information. Models are an increasingly important source of information, but such information is never complete, is rarely if ever certain, and hence is never a substitute for the judgment of experienced planners and managers. This paper attempts to reflect on the broader context in which models are used in the practice of planning, the inherent limitations of models in this broader context, and hence the challenges modelers have when addressing the information needs of planners and managers. Some criteria for judging the degree to which models have met these needs are reviewed. The paper concludes with some thoughts on the current state of planning models and supporting technology.

The objective of this paper is to review the state of the art of systems analysis and optimization techniques developed in the field of water resources for the planning and management of a ground‐water system. The areas reviewed include the following: ground‐water management models, inverse solution techniques for parameter identification, and optimal experimental design methods. Emphasis is placed upon ground‐water supply management models, as opposed to models used for ground‐water quality management. The techniques that have been used in the optimization of ground‐water management include: linear programming, mixed‐integer and quadratic programming, differential dynamic programming, nonlinear programming, and simulation. The inverse problem of parameter identification pertains the optimal determination of model parameters using historical input and output observations. Because of data limitation in both quantity and quality, the inverse problem is inherently ill posed. This paper summarizes recent advances made in the inverse procedures and methods developed to alleviate the problems of instability ard nonuniqueness of the identified parameters. The optimal experimental design problem addresses the issue of data requirements and optimal sampling strategies for the purpose of parameter identification. A criterion must be established for the optimal design of a pumping test. The fundamental concept of optimal experimental design and various criteria used for optimization are reviewed.

Review of the use of system‐analysis techniques, and in particular optimization, to design water‐distribution networks reveals that in spite of the considerable development of models in the literature they have not been accepted into practice. This lack of acceptance is present even though a competitive evaluation of the component design models has shown them to be capable of designing realistic networks. The lack of acceptance is attributed primarily to the absence of suitable packaging to make the algorithms useful in a design office environment. This evidence suggests that, from a practice point of view, there is relatively little need for further development of these component design models, other than the packaging. Reliability analysis in water‐distribution network design has not yet entered practice either. In contrast to the component size problem, reliability analysis has not been accepted primarily because of a lack of reliability measure that is both comprehensive in its interpretation of reliability and computationally feasible. More research is needed in this area before reliability can be explicitly incorporated into design procedures. Development of decision support systems (DSS) incorporating optimization and classical simulation models with an interactive graphics capability is seen as being a major priority for future research in the field, particularly if optimization approaches are to become more common in design practice. These DSSs can assist the designer in including and evaluating reliability and generating alternative solutions.

In the broadest sense of systems analysis, the Hydrologic Engineering Center (HEC) has been actively applying systems analysis techniques since its inception in 1964. HEC's basic systems analysis method has been the simulation of watershed and river basin processes. Several other systems analysis activities are an outgrowth of that basic simulation capability: automated parameter estimation algorithms, synthetic streamflow generation, network analysis, branch‐and‐bound search techniques, numerical methods, risk and uncertainty analysis, and expert systems. Systems analysis at the HEC has been an approach to problem solving as much as the actual technical methods. This philosophy of problem solving can be likened to a local hardware store's approach to do‐it‐yourself home repair. The hardware store personnel make a genuine effort to understand the homeowner's problem; then, they offer tools, supplies, and advice to solve the problem. Examples of these systems analysis techniques and their application to hydrologic engineering problem solving are presented in this paper.

During the past three decades, the application of the systems approach to reservoir management and operations has been established as one of the most important advances made in the field of water resources engineering. A primary role of systems analysis is to provide an improved basis for decision making. It has been concluded that a gap still exists between research studies and the application of a systems approach in practice. The objective of this paper is to provide a short review of the mathematical models used in reservoir management and operations, to present conclusions reached by previous state‐of‐the‐art reviews, and to provide two ideas for closing the gap between theory and practice. First, a simple simulation‐optimization model for reservoir sizing has been presented as an example of systems approach respond to practical needs of water resources engineers. The second example illustrates the benefits of knowledge‐based technology with regard to single‐multipurpose reservoir analysis.

Although mathematical programming techniques can be applied to solve certain (type A) problems, they cannot be used to solve complex (type B) problems directly. Instead they can be used to generate alternative solutions that can then be evaluated by the designer and modified according to judgment and intuition. A wastewater‐treatment plant design context is used to illustrate that, although a minimum‐cost model may produce an unsatisfactory design, other model formulations can be used to generate good alternatives. In particular, two model formulations are presented based on recent results in robust optimal design and in modeling‐to‐generate‐alternatives (MGA) techniques. It is argued that the alternative designs generated by these techniques may serve as starting points in an interactive design process, where the engineer uses trial and error to modify solutions, and create new ones, based on modeled and unmodeled issues.

Today, mathematical modeling has become an accepted part of the process of establishing and evaluating alternative scenarios for water‐quality management. The proliferation of microcomputers, the general availability of tested hydrodynamic and water‐quality models, and accessibility to user‐oriented graphics software have placed such decision‐support tools at the disposal of the resource manager. They have also raised concerns about uncertainties in model structure and predictive capabilities, as well as the quality of data and information required for reliable decisions. Even though modeling has enhanced the decision process, there still are questions concerning the proper role of such tools and how they can be most effectively used in water‐quality decision making. This paper traces the historical development of water‐quality modeling of natural surface‐water systems, from the 1960s to the present, citing some of the major advances and pitfalls in bringing computer‐aided decision support to its present state. It seeks to highlight capabilities now available to analysts and decision makers in the form of integrated packages of models, data bases, and interpretive graphics software. It also offers some caveats, lest the user‐decision maker become enamored with these newly discovered tools without realizing their limitations.

This paper considers water resources planning, specifically the case in which there is limited capability to expand water supply capacities by means of structural solutions. Under such circumstances, all measures that facilitate more efficient water use becomes very important. One of the most critical tasks of water resources planners is the evaluation of the potential for rationalization of water use. The problem is, however, that such practitioners have limited access to information that describes rationalization options available to individual water users. Regarding future water requirements, there is no method to precisely determine their values, and thus water resources planners must proceed on the basis of indirect information only. A dialogue between water administration and water users is always required, but in this case it is of a special importance. To this end, a water resources allocation model has been formulated, based on concepts of fuzzy sets and satisficing theories. A numerical example illustrates the application of the proposed methodology.

The lower Colorado River system, which is managed by the Bureau of Reclamation (BuRec), comprises three major dams (Hoover, Davis, and Parker). There is a significant hydroelectric capacity in the southwestern system, but to make optimum use of the resource, it must be integrated with available thermal and nuclear sources. The major competing demands on the system are water supply, flood control, and energy production. BuRec and the Western Area Power Administration have undertaken an effort to explore alternative management strategies for the three dams. A model that incorporates a hierarchy of objectives as well as physical, institutional, and political constraints imposed on the system has been developed. It may be used to provide guidance to support the managers of the system in real time as they make hourly operating decisions over the next week. The model may also be used to explore alternative operating strategies for the lower Colorado. The process of formulating, developing, and implementing the model has affected the evolution of improved operations for the lower Colorado.

Throughout 20 years, especially the recent decade, system engineering has been extensively researched on and applied to the water conservancy and hydroelectric power in China, involving the development and utilization of water resources, irrigation and drainage, flood control and water‐logging prevention, water‐supply and water‐quality protection in their survey, planning, design, construction, operation, and management. These applied fields deal with most of the theories and methods of optimization, such as mathematical programming, simulation techniques, system dynamics, utility decision‐making theory, large systems analysis, hierarchical analysis, model systems, and so on. System engineering brings into play an important role in the undertakings of water conservancy and hydroelectric power in our country. Some achievements and developments are reviewed in this paper. The optimization methods of large‐scale and multipurpose systems, expert systems, and decision‐making support systems will be vigorously developed in China in the days to come.