The water crisis is one of the most important conflicts of the day, and Iran is also in the field of water resources from countries with tension. Following the drop in groundwater levels in aquifers, most of the country's plains have experienced water crisis and water scarcity and their groundwater aquifers have been limited. Among the water stresses, Hashtgerd Plain, which has been selected as the study area. In order to exit the plain from the aforementioned situation, water management should be undertaken with basic management measures, which requires the provision of water resources predictive tools through these measures, one of which is the Bayesian network; In this study, to create a water-borne water prediction model, the predicted variable, groundwater surface parameter in the following month and predictive variables, temperature, precipitation, evaporation, aquifer feeding, The discharge level of the aquifer and the groundwater level were considered in the current month. To reduce information uncertainty, two predictive approaches were used with categorized data (discrete data) and prediction with explicit data (continuous data), and by examining the precision of the pre- The aquifer nose by the two approaches referred to provides explicit precision approach to observational hydrographs. As a result, this approach has been selected as a superior approach for predicting groundwater level due to the application of management scenarios in which the values ​​of the coefficient of explanation 0.94 and the RMSE is 0.207. The approach used in this study can be used in other water resources stresses.