Flood early warning systems are crucial for disaster risk reduction strategies, enabling communities to take timely action against threats. However, the effectiveness of these systems depends on accurate and timely hydrological data, particularly river discharge and water level measurements. Unfortunately, many regions face significant challenges in obtaining hydrological data, especially discharge data, due to outdated rating curves, high equipment costs, and logistical constraints. In contrast, water-level measurements offer reduced uncertainty and are often more accessible, providing an alternative to hydrological modeling in data-scarce regions. To address these limitations, we developed and validated the Discharge-to-Water Level Transformation (DWLT) method, which uses the monthly duration curves to transform discharge simulations from the GEOGLOWS ECMWF Global Hydrological Model into water level predictions using data from over 19,000 ground- and satellite-based river gauge stations. The results indicate that the water levels generated by DWLT are closely aligned with the observed water levels, especially when using satellite measurements, which offer a valuable alternative when ground-based data are scarce. Despite quality issues such as spikes and zero-level inconsistencies in ground-based data and temporal limitations such as short monitoring periods and infrequent measurements in satellite-based data, the methodology shows promising potential for large-scale and local hydrological applications. This work supports future flood forecasting and water resource management efforts, highlighting water level as an effective variable in hydrological modeling.