Accurate forecasting of river discharge is critical for the sustainable management of water resources, influencing applications such as irrigation planning, flood and drought mitigation, and infrastructure development. This study investigates the application of the Seasonal Autoregressive Integrated Moving Average (SARIMA) model to forecast minimum monthly discharges of the Danube River, addressing challenges posed by nonlinear and time-dependent hydrological processes. The study utilizes an extensive dataset comprising daily discharge records from ten stations across seven countries, spanning over a century. Monthly minimum discharges were computed and analyzed to identify long-term trends and seasonal patterns. The SARIMA model was selected for its proven ability to capture seasonal variations and optimize forecasting accuracy in da-ta-limited environments. Model performance was evaluated using statistical measures such as mean absolute error and root mean square error with results indicating robust predictive capabilities across the studied stations. The findings reveal significant vari-ability in discharge trends, with notable decreasing trends in minimum flows at several upstream and midstream stations, highlighting potential impacts of climate change and anthropogenic influences. In contrast, downstream stations exhibited relatively stable discharge patterns. These insights underscore the need for adaptive water manage-ment strategies to mitigate the risks associated with decreasing low flows. The study demonstrates the utility of SARIMA models in hydrological forecasting and provides a foundation for future research exploring hybrid modeling approaches incorporating climate variables. The results offer valuable inputs for policymakers and stakeholders in managing water resources under evolving climatic conditions.

Forecast accuracy, minimum flow prediction, model evaluation, river basin management, seasonal variability