A hybrid TOPSIS-agent-based framework for reducing the water demand requested by stakeholders with considering the agents’ characteristics and optimization of cropping pattern

According to the research performed, the agricultural sector is the major water consumer. Therefore, reducing the demand and the use of water in this sector has an effective role in solving the problems related to water scarcity. In this study, a hybrid TOPSIS-agent-based model has been developed to find a solution for this problem in the six management regions of Doroudzan irrigation and drainage network, in Fars province, Iran. Two main groups of agents consisting of the government and stakeholders were considered. All farmers who lived in the same village assumed as an agricultural agent. Some influential parameters such as neighbors’ impacts, training, penalties and incentives were considered in the agent-based model (ABM). In some previous ABM models, the same coefficients applied for all agents. However, without considering the stakeholders’ characteristics and conditions of each agent and region, there might be the threat of making unfeasible theoretical decisions. Therefore, in this study, the TOPSIS method was linked to ABM to determine the uniqueness of these coefficients for each agent. The ABM coefficients were determined by TOPSIS method through the use of the demographic, social, economic and cultural variables and expert's viewpoints. Water demand requested by farmers was calculated based on the existing cropping pattern and the reduction of that was determined by the developed model in different climatic conditions (wet, normal and drought). The results showed that the water requested by the farmers, before applying the ABM, were 7.6, 18.2 and 45%, and after that they were 1.37, 3.5 and 1.09% more than the allocatable water in wet, normal and drought conditions, respectively. In order to reduce the water demand requested by farmers and maintain the effect of management tools (training, penalties and incentives), a cropping pattern optimization model with regard to deficit irrigation was developed.