Journal of Water Resources Planning and Management

In operating a ground‐water management system it is likely the true aquifer response will deviate from the response predicted during the design phase. Efficient operation will require adjustments in operation based on actual system response. The potential of adaptive control for the real‐time management of groundwater hydraulics in the presence of model and parameter uncertainty is demonstrated by an application to target head tracking. Despite uncertainty in the values and spatial distribution of transmissivity, meeting target head levels at selected observation wells by choice of the pumping rates is desirable. Parameter estimation and control are applied sequentially. Head measurements are used to update parameter estimates and modify future pumping choices. It was found that a deterministic approach, with zonation for approximating spatial variability and periodic updating of transmissivity estimates, was capable of meeting the target heads with a reasonable degree of success.

An optimization model for the control of saltwater intrusion in the Jakarta Multiple aquifer system is presented. The optimization model incorporates a quasi‐three‐dimensional, finite‐difference simulation model of the aquifer system. The ground‐water systems are hydraulically coupled via leakage into or out of each aquifer. Finite‐difference methods are used to generate the hydraulic‐response equations of the ground‐water basin relating, for each aquifer, the freshwater and saltwater heads, the location of the interface, and the pumping and recharge schedules. The optimization model minimizes the total squared volume of saltwater in each aquifer of the ground‐water system; the control variables of the model are the location and magnitude of ground‐water pumping and recharge. The model is required to satisfy the current water demand in the basin. The optimization model is solved using MINOS, a projected Lagrangian algorithm, and a modified gradient procedure based on Box's algorithm. The underlying simulation equations are incorporated in each optimization algorithm; the algorithm evaluates the simulation equations to determine the objective and constraint gradients of the planning problem. Planning results are presented. Parametric programming is used to determine the trade‐offs affecting the optimal decisions.

A network flow programming‐based model, KCOM, was developed to simulate water allocation and distribution in the surface water system of Kern County, California. The model will help the California Department of Water Resources plan the proposed Kern Water Bank conjunctive‐use project. KCOM incorporates recharge and pumping features to allow simulation of both surface and subsurface water distribution. Network flow programming‐based simulation was chosen for KCOM because of its generality, flexibility, speed, and superior ability to reflect complex water allocation policies. The water allocation capabilities of network flow programming‐based simulation were enhanced in KCOM by the development of a sequential flow allocation technique, in which sources are allocated in different layers to meet only demands with authorized access to those sources. A validation simulation demonstrates the model's capabilities, and sensitivity analysis underscores the need for improvement of the data used to describe the system. Recommendations are made for the improvement of KCOM.

The general energy management by successive linear programming (GEMSLP) model is developed to solve the optimization problem of the hydropower system operation of a planning horizon longer than a week (Reznicek and Simonovic 1990b). The energy rate function (ERF), denoting the energy obtained from unit release volume, is approximated in the original energy management by successive linear programming (EMSLP) model to be the exclusive function of reservoir storage. This work presents a modification to the original EMSLP model. The GEMSLP model has the capability to take into account the multivariate character of the ERF showing the dependence on discharge and storage. Two ways are investigated to obtain the smooth and differentiable analytical form of the ERF: integral regression analysis and separate analysis of significant relationships. The application of both approaches are illustrated on case studies using the Manitoba Hydro‐system data.

An implicitly stochastic optimization scheme previously described in the literature is extended to consider multiple reservoir systems. The scheme comprises a three‐step cyclic procedure that attempts to improve the initial operating rules for the system. The system requires two sets of contemporaneous streamflow series to be used in the simulation model and synthetically generated series are required for this purpose. The three‐step cycle begins with an optimization of reservoir operations for a given set of streamflows. The optimal operations from the solution are then analyzed in a regression procedure to obtain a set of operating rules. These rules are evaluated in a simulation model using a different set of data. Based on the simulation results, bounds are placed on operations and cycle returns to the optimization model. The cycle continues until one of the stopping rules is satisfied. The use of the scheme to generate operating rules for multiple reservoir systems is illustrated for a two‐river system under a set of 28 different conditions.

The extended streamflow prediction (ESP) procedure of the National Weather Service River Forecast System (NWSRFS) produces long‐range forecasts of streamflow through the use of hydrologic models and historical hydrologic data. An important element of the ESP procedure is converting hydrologic‐model output to estimates of a forecast random variable. In this paper, nonparametric statistical procedures are developed for combining hydrologic models and historical hydrologic data into long‐range streamflow forecasts. Although these procedures are developed for use within the ESP system, they should be broadly applicable to problems of long‐range streamflow forecasting. Two notable features of the procedures developed in this paper are: (1) Climate information is easily incorporated; and (2) hydrologic‐model errors can be accommodated. Results are presented for a test implementation of ESP for a basin in the southeastern United States during the severe drought period of 1988. The relative importance of climate information and soil moisture information for long‐range streamflow forecasting is compared and contrasted.

Despite decades of criticism, the benefit-cost ratio remains one of the most common measures of performance employed in economic evaluation of engineering projects. This paper reviews the problems inherent in this measure of economic performance and suggests alternative measures in its place. It is shown that there are many common circumstances in which the benefit-cost ratio is an unrealiable indicator of a design's economic desirability and comparative economic performance. These problems result largely from the economically arbitrary nature of most classifications of benefits and costs. The ratio measure should be dropped from all serious economic-evaluation manuals and protocols. Some alternative measures are discussed.