Groundwater Flow and Contaminant Transport Modeling in the Selgi Watershed, Woleka Sub Basin, Ethiopia

Abstract

Owing to urbanization, modern agriculture, population growth, and global warming, groundwater supplies are under significant stress. The quality and potential of groundwater have declined over time. In this context, groundwater models are crucial for tracking these changes. This study employs the MODFLOW and MT3DMS models to simulate nitrate concentration and groundwater levels. Data were collected from twelve hand-dug wells every quarter of the year for nitrate concentrations, while eighteen wells provided monthly measurements of groundwater levels. The default metrics found in the Visual MODFLOW 2015 model were utilized to evaluate model performance. After the flow models were calibrated and validated, the results demonstrated a very good fit, and the transport modeling also performed well. A sensitivity analysis using the PEST model indicated that hydraulic conductivity is a sensitive parameter, varying between 0.01 and 0.15 m/day, while other parameters remained constant. The study employed sorption and reaction-type simulations to calibrate the Transport Model owing to the mobility of nitrate species, with advective transport being the predominant mechanism. The ideal parameters for nitrate calibration included a distribution coefficient (Kd) of 0.000001 1/mg/L, a reaction rate of 0.001 1/day, an effective velocity of 0.14 m/day, and a longitudinal dispersion of 100 meters. Hydraulic heads were observed to be low downstream and high upstream in the watershed. Hydraulic heads were observed to be lowest downstream and highest upstream in the watershed. However, only minor changes were noted in the simulated nitrate concentration within the watershed. This finding may assist decision-makers in understanding groundwater level variability and nitrate contamination, which can guide future fertilizer application rates and the development of new shallow wells.

Keywords: Calibration, MODFLOW, MT3DMS, Kd, PEST, and Validation