This study aims to perform long-term forecasting of monthly water levels from El Yeso Dam, one of the most important water reservoirs in Santiago, Chile. Leveraging a 14-year time-series climate data and satellite imagery, our investigation focuses on understanding the dynamics of water levels and their correlation with environmental factors. Landsat constellation images were extracted using the Google Earth Engine platform to examine the appearance of the dam and calculate the area of the water surface. Several experiments were performed using stochastic time series models such as SARIMAX and Recurrent Neural Networks, including SimpleRNN, LSTM, and GRU. The algorithms were validated through the examination of errors, analysis of performance criteria, and inspection of residuals. According to the results, the reservoir's volume does not exhibit a sustained downward trend. On the contrary, a slight increase in water level is forecasted during the winter months and a significant rise is anticipated for the summer months. This suggests that the consistent temperature increase in the area due to climate change may contribute to the melting of the snow cover and surrounding glaciers, resulting in an increase in water levels. Therefore, this research contributes to the development of effective planning and management strategies for the region's water resources.