Simulating the effect of injection well on groundwater table in unconfined aquifer using numerical model of isogeometric analysis and optimization of injection rate with PSO algorithm

Abstract

Currently, one of the solutions to improve the decline in the water table of aquifers owing to their uncontrolled extraction is artificial recharge schemes, especially in arid and semi-arid regions. The present study applies a nonuniform rational B-spline functions (NURBS) based isogeometric method to investigate the effects of an injection well on the rise of water table, unconfined aquifer hydrodynamic parameters and injection rate. Also, the best injection rate was determined by linking the isogeometric analysis (IGA) simulation model to the particle swarm optimization (PSO) model. Initially, the IGA method was implemented to simulate a water table over an unconfined aquifer with two extraction wells with discharges of 1142.85 and 1428.57 m 3  day −1 for 210 days and was compared with the Modflow model. The model was tuned and run by constructing an injection well with 8214.28 m 3  day −1 injection rate for 1500 days. The IGA simulation results with ME = −0.0096, MAE = 0.0111 and RMSE = 0.0146 indicated the accuracy of the model in water table simulation compared with the Modflow model with ME = −0.0188, MAE = 0.023 and RMSE = 0.0284. The findings revealed that the use of an injection well has a positive impact on raising the water table, with a rise of 1.24 m in the observation well. Investigating the effect of the parameters on the water table changes indicated that the injection rate directly affects changes in the water table. The water table rises by more than 43% with a 50% increase in the injection rate. Also, compared with the effect of the injection rate on the water table, the transmissivity and specific yield effects were not substantial. The results of IGA–PSO showed that the best value of injection rate is 2000 m 3 day −1 , whereby the aquifer water table rises by an average of 37.12 cm.