Kamali et al. (2025) Evaluating the environmental and economic effects of groundwater quotas and markets using multi-agent system simulation and centralized optimal model

Identification

• Journal: Agricultural Water Management
• Year: 2025
• Date: 2025-10-06
• Authors: Asghar Kamali, Mohammad Hossein Niksokhan, Mojtaba Ardestani
• DOI: 10.1016/j.agwat.2025.109853

Research Groups

Faculty of Environment, University of Tehran, Tehran, Iran

Short Summary

This study evaluates the environmental and economic impacts of groundwater quotas and markets in the Ardabil Plain, Iran, using a multi-agent system simulation (MASS) model and a centralized optimal model (COM). It finds that decentralized, self-interested agents under uniform quotas perform poorly compared to a centralized optimal approach, while permit trading can enhance economic benefits and, in some scenarios, reduce environmental violations.

Objective

• To evaluate the environmental and economic performance of different groundwater use quotas and groundwater markets in the Ardabil Plains aquifer, comparing a decentralized multi-agent system simulation (MASS) approach with a centralized optimal model (COM) approach.

Study Configuration

• Spatial Scale: Ardabil Plain, Iran, spanning 990 square kilometers. The aquifer has a maximum thickness of 220 meters. The hydrological model uses 504 grid cells, each 1.75 kilometers by 1.75 kilometers. The study involves 27 farmer agents, each representing clusters within a sub-basin.
• Temporal Scale: Hydrological models (SWAT-MODFLOW) run on monthly stress periods. The groundwater model was calibrated over five years (2017–2022) and verified over one year (2022–2023). Policy evaluation is based on average annual benefits over one-year management periods, with simulations conducted for ten different sets of agent productivity parameters.

Methodology and Data

• Models used:
  • Multi-Agent System Simulation (MASS): Integrates an agent-based platform (NetLogo) with a calibrated groundwater model (MODFLOW) and a watershed model (SWAT) using Python-based codes. Agents are self-interested and maximize individual benefits.
  • Centralized Optimal Model (COM): Integrates MATLAB with the calibrated MODFLOW and SWAT models, using a Particle Swarm Optimization (PSO) algorithm to maximize total agent profits while meeting environmental constraints.
  • MODFLOW: Groundwater flow model.
  • SWAT: Soil and Water Assessment Tool (watershed model).
• Data sources:
  • Groundwater level trends from 32 piezometric stations for MODFLOW calibration.
  • Aquifer recharge data derived from the SWAT model.
  • Local financial data for crop costs (e.g., 150 million, 80 million, and 1 billion Rials per acre for Wheat, Barley, and Potato, respectively) and selling prices (e.g., 125,000, 120,000, and 150,000 Rials per kilogram for Wheat, Barley, and Potato, respectively).
  • Ten sets of randomly generated agent productivity parameters (crop yield and irrigation requirements) to account for uncertainty.

Main Results

• Simulations with shortsighted, self-interested, and heterogeneous agents (MASS) under a uniform groundwater quota policy perform poorly compared to the Centralized Optimal Model (COM).
• Trading of groundwater permits among farmers generally leads to enhanced economic benefits (gains from trading) and, occasionally, a decrease in environmental violations (streamflow depletion).
• The percentage of gains from trading and the increase in groundwater pumping due to market creation are highest at minimal allocation levels (e.g., 100 millimeters) and decrease as allocation levels increase.
• The MASS with free-access policy yields the highest average annual benefit (AAB) of approximately 151 trillion Rials but results in the highest streamflow violation (VSF) of about 90%.
• The COM, representing an optimal solution with complete information and centralized regulation, achieves an AAB of approximately 120.3 trillion Rials with 0% VSF.
• To reduce VSF to 20% or less, a maximum water usage quota slightly exceeding 100 millimeters is required across all scenarios.
• To achieve an AAB comparable to the COM's 120.3 trillion Rials, a quota exceeding 200 millimeters is needed, which results in VSF rates slightly over 39% for MASS with quotas and slightly under 37% for MASS with a groundwater market.
• Under higher quotas, permit trading can lead to increased streamflow violations due to overall system-wide pumping increases, as farmers with surplus allocations sell water, counteracting the environmental benefits of improved water productivity.

Contributions

• This study's primary contribution is the integration of a groundwater market into both the Centralized Optimal Model (COM) and the Multi-Agent System Simulation (MASS), building upon previous work that focused on comparing other policy instruments.
• It provides a comprehensive evaluation of the environmental and economic performance of groundwater quotas and markets under both centralized and decentralized decision-making frameworks, considering heterogeneous agents and physical aquifer characteristics.
• The research offers practical insights for policymakers by identifying optimal quota levels that balance economic benefits for agents with environmental sustainability within a multi-agent simulation context.

Funding

The authors received no funding for this study.

Citation

@article{Kamali2025Evaluating,
  author = {Kamali, Asghar and Niksokhan, Mohammad Hossein and Ardestani, Mojtaba},
  title = {Evaluating the environmental and economic effects of groundwater quotas and markets using multi-agent system simulation and centralized optimal model},
  journal = {Agricultural Water Management},
  year = {2025},
  doi = {10.1016/j.agwat.2025.109853},
  url = {https://doi.org/10.1016/j.agwat.2025.109853}
}