With a steadily growing population and increasingly limited natural freshwater resources, water-scarce regions must implement smart interventions to sustainably manage their water resources. Using Groundwater Vistas software, a modelling framework for Kuwait City urban aquifer was developed to address this issue. This framework was employed to create a numerical model of the shallow aquifer beneath the urbanised aquifer of Kuwait City. The model was calibrated using groundwater levels from 15 calibration points across the modelled area. Two key factors influenced the calibration: the constant head boundary condition applied upstream in the model domain and the horizontal hydraulic conductivity. The model performed satisfactorily achieving a Nash-Sutcliffe efficiency coefficient of 0.983, a root-mean-square error of 1.134 m, and a Kling-Gupta efficiency of 0.971. These results were then used to simulate two complex hydrogeological processes affecting the Kuwait City aquifer: unplanned dewatering schemes and incidental recharge from anthropogenic watering practices. It was found that altering the pumping rate between 250 and 2250 m³/day, resembling the dewatering process, could lower the groundwater level by approximately 1.21–1.79 m in coastal areas. Conversely, an unplanned recharge of 1500–6000 m³/day could substantially raise groundwater levels by approximately 2.5–3 m in inland areas. However, these findings should be cautiously approached, as certain constraints may significantly influence the model's reliability. These constraints include the limited availability of data records and the possible existence of unknown sources/sinks to the aquifer. The outcomes of this study should aid water managers in establishing reliable groundwater control decisions in the study area and other areas with similar hydrogeologic characteristics.