Linking large-scale low-frequency atmospheric variability to compound flood events in Emilia-Romagna

Extreme weather and climate events - such as floods - are among the most impactful hazards for both human society and natural systems. Their accurate representation and prediction remain challenging, especially when such events are driven by interacting climatic factors that operate on different spatial and temporal scales. In this context, large-scale atmospheric circulation patterns and associated teleconnections play a crucial role in modulating regional hydro-meteorological extremes, offering a potential source of predictability that is still not fully exploited or understood. This study examines extreme flood events in Emilia-Romagna, a flood-prone region of northern Italy, with a particular focus on the link between local flood dynamics and broader atmospheric variability.

Floods are here categorised into two types: preconditioned floods, which occur when extreme precipitation is preceded by abnormally high levels of soil moisture, and non-preconditioned floods, in which no such antecedent conditions are present. To detect these extremes, a peak-over-threshold method is applied to ERA5-Land reanalysis data spanning the period from 1979 to 2024, capturing significant anomalies in both precipitation and soil moisture. The role of atmospheric low-frequency variability is explored through the analysis of geopotential height anomalies at 500 hPa, derived from ERA5: results indicate that preconditioned flood events are frequently associated with persistent and more intense negative geopotential height anomalies near the affected region, particularly during the winter months, highlighting the role of seasonal atmospheric circulation patterns in enhancing flood risk. Additionally, hydrological impacts are assessed using EFAS river discharge data (1992–2024), showing that preconditioned events are associated with widespread and elevated discharge levels across Emilia-Romagna. Finally, an event coincidence analysis reveals that soil moisture preconditioning is more frequent during winter months, consistent with enhanced synoptic-scale persistence during this season.

Overall, the study offers new insight into how persistent large-scale circulation anomalies modulate regional flood risk through both dynamic and hydrological pathways. These findings underscore the importance of incorporating atmospheric teleconnections into flood early-warning systems, and suggest avenues for improving predictive capabilities not only in Emilia-Romagna but also in other areas with comparable climatic and hydrological settings.