Water shortages and concerns about water supply are prevalent in many parts of the world. Analyzing the factors that affect water reserves over time is crucial. This study developed a machine learning model for groundwater level forecasting based on various factors influencing water storage. The manuscript proposes a new method for automatically handling missing data and calculating temporal variables before including them in the training dataset. Subsequently, an ensemble learning approach was applied to build the water level forecasting model. The results demonstrate that the proposed model can accurately predict the trend of water level changes in storage areas such as aquifers and lakes. In particular, the forecasting performance of the proposed model, measured by R2 and MSE, shows outstanding results with an R2 score of 0.99 and an MSE of 0.004. This suggests significant potential for implementing water level forecasting models in real-world applications.