Crop pattern optimization in a multi-reservoir system by combining many-objective and social choice methods

A.B. Dariane,M. Ghasemi,F. Karami,A. Azaranfar,S. Hatami
DOI: https://doi.org/10.1016/j.agwat.2021.107162
IF: 6.7
2021-11-01
Agricultural Water Management
Abstract:Crop pattern determination is an essential step toward proper water resources management in irrigated agricultural development, especially in arid and in semi-arid regions. In this paper, we propose a method for determining short-term optimal crop pattern by developing a suitable model for a four-reservoir system in the Karun basin including Karun 1, 3, and 4, and Gotvand reservoirs in Iran. The model satisfies multiple objectives including hydropower, municipal, industrial, agricultural, and environmental purposes. To achieve this goal, we examined two heuristic many-objective algorithms including SPEA2 and NSGA-III. Many-objective algorithms provide a Pareto front of optimal solutions. However in practice we need to know an exact solution of crop pattern. To overcome this problem, a novel application of social choice (SC) is accomplished. The results indicate the superiority of NSGA-III over SPEA2. Comparison of the results for the minimization objective functions shows that the average values of municipal, agricultural, environmental and hydropower obtained with the NSGA-III algorithm are 0.22, 6.73, 0.49 and 1.72, respectively, whereas in SPEA2 they are 0.72, 9.47, 2.05 and 4.72. This finding also emphasizes the importance of improvements necessary in solution algorithms for complicated water resources problems.
agronomy,water resources
What problem does this paper attempt to address?