Towards a storyline of compound flood events over Italy: the role of precipitation-soil moisture pre-conditioning

The urgency for deeper understanding compound hydro-meteorological extreme events is growing, as these events are increasingly recognized for their potential to exacerbate impacts compared to single hazards. Characterized by the concomitance of multiple natural hazards or drivers, compound events are further intensified by climate change, which influences their severity and increases their frequency of occurrence. These hydro-meteorological extremes pose a significant risk to terrestrial ecosystems and have devastating consequences for socio-territorial systems. In Italy, recent extreme events have highlighted this threat, as demonstrated by the unprecedented sequence of heavy precipitation events in 2023-2024 that led to widespread flooding in the region of Emilia Romagna in central-northern Italy. These low-probability events, which resulted in several fatalities and damages amounting to tens of billions of euros, were amplified by antecedent precipitation that saturated soils, significantly enhancing the runoff response and, consequently, flood severity and extension. Indeed, the pre-condition given by soil imbibition preceding heavy rainfall occurrences is crucial in determining the potential severity of the event, but its comprehensive understanding is still limited.

This study investigates the relationship between precipitation and soil moisture conditions in Italy, with the goal to quantitatively characterize their role in the occurrence of historical and plausible compound hydro-meteorological extremes. We employ state-of-the-art reanalysis datasets (ERA5 and ERA5-Land) to analyze a series of representative extreme precipitation events, focusing on their antecedent soil moisture conditions and estimating the typical temporal scales of the associated co-variation. The link between these hydro-meteorological quantities and riverine flood occurrences is assessed considering streamflow discharge data from EFAS hydrological reanalysis dataset. The prevailing large-scale conditions driving these events, in terms of the 500-hPa geopotential height and the integrated water vapor transport column, are explored to identify the key dynamical features responsible for the occurrence of compound flooding. Additionally, a large ensemble of seasonal numerical weather forecasts is employed to sample the phase space of precipitation-soil moisture conditions applying the so-called UNSEEN (Unprecedented Simulated Extremes using ENsembles) approach. Within this framework, the probability of compound precipitation-soil moisture extremes is assessed through a statistical event coincidence analysis to understand the dominant spatio-temporal patterns of their interaction; moreover, physical storylines of rare, yet plausible, extreme flood events will be built through ensemble pooling.