Development of Simulation-Optimization Models for Agricultural Contaminant Loading Management Considering Effects of Groundwater Pumping

Abstract

In agricultural regions, a significant amount of groundwater has been used but agricultural activities, such as greenhouse farming, have often threatened its quality. Therefore, it is necessary to suitably manage the agricultural contaminant loading for sustainable groundwater use in those regions. However, pumping condition should also be considered in the management because groundwater pumping can change the fate of contaminant in the subsurface such as its leaching to the water table and migration in the aquifer. In this study, based on the field investigation and monitoring, an agricultural contaminant loading management model was developed in order to determine the optimal permissible contaminant loading mass for a given pumping condition, using simulation-optimization method. Periodical on-ground contaminant loading on non-point sources such as fertilizer application in greenhouse was simulated by integrating the 1-D analytical solution for solute transport in the unsaturated zone and the 3-D numerical model for groundwater flow and solute transport in the saturated zone. Backward transport simulation was applied to the model in order to evaluate the relative importance of contaminant sources quantitatively. Genetic algorithm was linked to this integrated simulation model as optimization technique. This model could be useful in the agricultural contaminant loading management in the agricultural regions where many potential non-point sources were located at. Using this model, the optimal contaminant loading designs obtained under various pumping conditions were compared in order to examine the effects of pumping conditions in determining the optimal contaminant loading. The results demonstrated that the optimal contaminant loading designs were determined differently according to the given pumping conditions. Another management model was developed to manage permissible on-ground contaminant loading mass and pumping rates simultaneously. This model cannot consider only dynamics between fate of contaminant and pumping but also various conditions such as different usage of, or demand on, each pumping well and contaminant source in a single of optimization process. The optimal design determined from this model allowed more amounts of both of contaminant loading and groundwater pumping than any other optimal design suggested previously. In addition, in the agricultural regions where groundwater has been used intensively in a specific period of time, such as rice-growing season, it must be particularly important to consider such pumping condition in the agricultural contaminant management. Therefore, the model to simultaneously manage agricultural contaminant loading and groundwater use under time-variant pumping condition was also developed. For this, the method to approximate the contaminant leaching to the fluctuating water table caused by a regular schedule of groundwater pumping was suggested and transient groundwater flow simulation was applied. In the optimal design obtained under the time-variant pumping condition, the contaminant loading was restricted considerably because a relatively large amount of contaminant leaching to the shallow depth of water table during the period without groundwater pumping, a strong inflow of contaminant to the wells driven by the large amount of pumping during the period with groundwater pumping, and a sudden increase of contaminant leaching immediately after stopping the operation of pumping. Particularly, the optimal design obtained under the no-pumping condition in this study was to imitate some previous studies about agricultural contaminant management which had not considered any pumping condition, in order to demonstrate the importance of considering pumping condition in the agricultural contaminant loading. The result showed that the agricultural contaminant loading management without considering pumping condition could fail in the regions where groundwater use has been common.