RESEARCH PAPER Pollution Source Identification in Groundwater Systems: Application of Regret Theory and Bayesian Networks Seyyed Nasser Bashi-Azghadi 1 • Reza Kerachian 2 • Mohammad Reza Bazargan-Lari 3 • Mohammad Reza Nikoo 4 Received: 22 January 2015 / Accepted: 19 January 2016 Ó Shiraz University 2016 Abstract Pollution source identification in groundwater resources is a challenging task due to existing uncertainties in both pollutant source characteristics and mass transport in porous media. To obtain a robust and cost-effective groundwater monitoring configuration, this paper presents a new regret-based optimization model which minimizes the number of monitoring wells and average regret in estimating undetected polluted area. A Monte Carlo anal- ysis is used to consider existing uncertainties in both pol- lution source characteristics and parameters of groundwater quality simulation model. MODFLOW and MT3D, groundwater quantity and quality simulation models, are used to simulate the spatial and temporal variations of a water quality indicator in groundwater. For each non-dominated solution provided by a bi-objective optimization model, the optimal positions of monitoring wells are also determined by minimizing the corresponding regret in undetected polluted area. Furthermore, a Bayesian network (BN) is trained and validated based on results of the Monte Carlo analysis. The trained BN is capable of accurately determining an unknown pollution source using monitoring data in real-time operation of monitoring sys- tem. To demonstrate the efficiency and applicability of the proposed methodology, it is applied to the Tehran Aquifer in the Tehran refinery region, which is highly polluted due to leakage from several tanks of petroleum products. Keywords Groundwater monitoring Á Regret theory Á Bayesian networks (BNs) Á Tehran refinery Á Pollution source identification 1 Introduction Optimal design of groundwater monitoring networks has received increasing attention in the past decades due to the important role of groundwater resources in supplying water demands in arid and semiarid regions, and the considerable cost of remediating polluted groundwater systems (He et al. 2008). Where multiple potential pollution sources threaten aquifers, rapid identification of the main pollution source and its characteristics is vital to take contaminant control and containment actions. Identification of an unknown pollution source is a challenging task due to existing uncertainties in pollutant source characteristics and flow and mass transport processes in porous media (Datta et al. 2009). Coupling a groundwater simulation model with a stochastic optimization technique would be the best option for designing water quality monitoring networks. Mini- mizing the number of monitoring wells, volume of con- taminated groundwater, and construction, operational, and & Reza Kerachian kerachian@ut.ac.ir Seyyed Nasser Bashi-Azghadi s.naserbashi@yahoo.com Mohammad Reza Bazargan-Lari bazargan.lari@yahoo.com Mohammad Reza Nikoo nikoo@shirazu.ac.ir 1 School of Civil Engineering, Iran University of Science and Technology, P.O. Box 16765-163, Narmak, Tehran, Iran 2 School of Civil Engineering and Center of Excellence for Engineering and Management of Civil Infrastructures, College of Engineering, University of Tehran, Tehran, Iran 3 Department of Civil Engineering, East Tehran Branch, Islamic Azad University, Tehran, Iran 4 Department of Civil and Environmental Engineering, School of Engineering, Shiraz University, Shiraz, Iran 123 Iran. J. Sci. Technol.Trans. Civ. Eng. DOI 10.1007/s40996-016-0022-3