Water Resources Management 13: 315–334, 1999. © 2000 Kluwer Academic Publishers. Printed in the Netherlands. 315 Feedback Method of Control for Estuary Management I. KAAN TUNCOK 1 and LARRY W. MAYS 2 1 Stanley Consultants, Inc., Phoenix, AZ 85016, U.S.A. 2 Department of Civil Engineering, Arizona State University, Tempe, AZ 85287, U.S.A. (Received: 23 July 1998; in final form: 31 August 1999) Abstract. A feedback method of control has been used to develop a model for optimal determination of freshwater inflow to bays and estuaries. A modification of the feedback method of control was implemented which makes the technique applicable to certain constrained optimal control problems. The modified feedback model for estuary management consists of a hydrodynamic-transport salinity model, HYD-SAL, coupled to a dynamic programming optimization model. The constraints of the model are the monthly freshwater inflows and the salinities. A quadratic criterion representing a weighted sum of squared deviations from target salinity and freshwater level is chosen as the object- ive function. The constrained optimal control algorithm employs a penalty function that uses a similar quadratic criterion as the objective function. This algorithm has performed efficiently for computing the optimal freshwater inflows into the Lavaca-Tres Palacios Estuary in Texas while satisfying the freshwater requirements for other components in the system. Key words: estuary management, feedback method of control, optimal control. 1. Introduction Estuaries are important because they provide areas of nursery habitats for juvenile forms of marine species, for sport and commercial fishing and for other recreational activities. The provision of sufficient freshwater inflow to estuaries is a vital factor in maintaining estuarine productivity. The objective of this paper is to develop a modeling approach to determine time-varying strategies for this freshwater inflow. Mathematical estuarine management models that can determine an optimal bal- ance of freshwater inflows between the upstream water demands and needs for fisheries have been the focus of numerous studies. Martin’s (1987) linear pro- gramming model for determining the monthly freshwater inflow needs for seven major bays and estuaries in Texas is one of them. In Martin’s model the spatial distribution of salinity is ignored, and the complicated hydrodynamics of inflow and salinity are replaced by the simple monthly averaged regression models. Drake (1990), proposed a salinity control model for Breton Sound Estuary in Louisiana to maintain the salinity to the desired target levels by controlling diversion gate settings. Tung et al. (1990), developed a nonlinear chance-constrained model and used