Groundwater Management in Ardabil Plain Using Agent-Based Modeling

Modeling socio-hydrological interactions are one of the essential requirements for water resources management in water-stressed areas. The Ardabil aquifer (Northwestern Iran) is one of the restricted aquifers under intense development activities. The water table is dramatically declining and leading the area to an environmental disaster. In this study, a simulation-optimization model has been developed for solving the Ardabil groundwater problem, which simulates groundwater level changes and determines the optimal water exploitation values. These models have been linked by a new method in the MATLAB which provides access to various MODFLOW packages and takes up less memory. The simulation-optimization model has been then linked to an agent-based model, which simulates agents’ behavior and their interactions. For this purpose, firstly the key agents and their desirability have been determined. Then, the particle swarm optimization algorithm is used to estimate the agents’ desired groundwater exploitation values. In the next step, the best solution using the compromised programming method is selected according to the experts' point of view. Finally, the agent-based model provided the final exploitation values, taking into account social pressure and management rules (incentive and penalties). The results show that groundwater demand is reduced up to 22% in comparison to the initial value. This demand reduction resulted in 90 cm of the increase in the groundwater level for the entire plain.