<p>In this study, groundwater exploitation risk was calculated by integrating the normalized vulnerability and hazard susceptibility of groundwater resources to identify and categorize groundwater management and protection-priority zones. Based on the presence of alluvial and karstic aquifers in the study area, the DRASTIC and EPIK indices were employed to determine the intrinsic vulnerability. The combined results of these two vulnerability indices, along with a transformed groundwater-quality degradation indicator derived from the Groundwater Quality Index, were used to calculate the overall vulnerability. To determine hazard susceptibility, a Bayesian Network simulation was conducted based on land use, slope, geological formations, and natural landforms. To calibrate and optimize the intrinsic vulnerability indices, a hybrid model known as CNN-GWO was utilized. The calibration results indicated that this method produces favorable outcomes for optimization by maximizing the correlation between the index values and nitrate concentration in the region. The combination of the intrinsic vulnerability and groundwater-quality degradation indicator of the aquifer also demonstrated that the highest overall vulnerability, with a normalized value of 0.92, was observed in the western section of the alluvial aquifer, followed by the elevated regions in Dorfak. The simulation results of the Bayesian Network demonstrated that natural landforms associated with karstic cave and sinkhole activities had the greatest impact on the level of hazard susceptibility, with the highest normalized susceptibility value calculated at 0.55. Ultimately, the results of risk zoning showed that the highest normalized risk value of 0.36 was observed in the hydrological network in the alluvial fan of the plain and the developed karst landforms in the Dorfak highlands. The final risk map was interpreted as a spatial decision-support tool for groundwater management rather than a direct measurement of confirmed exploitation risk.</p>