Optimisation model for water allocation in deficit irrigation systems II. Application to the Bembezar irrigation system

In this paper, the proposed optimisation model is applied to optimise water management in the Bembezar system, a small hydrological basin belonging to the Guadalquivir River basin in southern Spain that supplies water to the Bembezar River Irrigation District. In order to apply the model, the irrigation methods and performance in the irrigation district have been analysed through a set of field irrigation ev aluations. Cropping patterns, crop productivity and other relevant agronomi c and economic data have been collected. The influence of irrigation uniformity and the type of distribution of irri gation water on the crop yields, as well as the relationship between crop y ields and irrigation scheduling have been analysed using the proposed model . A deterministic analysis has been carried out in the irrigation district in order to compare optimum water and cropping patterns management with actua l ones. In order to account for the randomness of both climatic and water availabil ity variables, a stochastic data generation has been carried out which cons iders the correlation between these hydrological series. The system is then analysed in a stochastic environment. Several simulations of the optimisat ion process have been carried out using generated data on climatic and wate r availability variables. The result of this analysis demonstrates that when only the satisfaction of the internal demands is considered, high quantities of water are allocated to the irrigation districts resulting in low economic benefits per unit of water used and lower irrigation efficiency. This situation has been compar ed with the solution provided by the hypothesis of a proposed water market in which it is possible to transfer part of the water of the system to othe r alternative uses at a fixed price. In this second hypothesis, water consu mption in the irrigation districts was reduced. (C) 2001 Elsevier Science B .V. All rights reserved.