Abstract
Understanding the relationships between large-scale climate circulation and precipitation and river discharge at the local level is of central importance for deciphering hydroclimatological processes. Annual maximum river discharges in Europe are more sensitive to the variability of atmospheric circulation than mean discharges. However, there is a knowledge gap regarding the specific influence of atmospheric patterns on maximum discharge dynamics in the Middle and Lower sections of the Danube basin, especially in Serbia and Romania. In this study, historical climate data and the influence of atmospheric patterns (Greenland-Balkan Oscillation index, Mediterranean Oscillation index and North Atlantic Oscillation index) were analysed to understand their impact on hydrological extremes. The results showed significant correlations between discharge and atmospheric oscillations, in particular a statistically significant decreasing trend of maximum monthly discharge at Orsova station (trend: −0.0548, p-value: 0.0305) in the period 1961–2018. Correlation analyses showed significant positive correlations between runoff and atmospheric indices in different time periods and stations, underlining the complexity of these relationships. In addition, GEV analysis underlined the stability of maximum discharges at Bezdan and Bogojevo stations over time, with notable instability observed at Orsova and Ceatal Izmail stations. These results emphasize the importance of considering both temporal and spatial variations in river discharge dynamics when assessing the impact of climate change on hydrological extremes.