Detection and characterization of precipitation extremes and geohydrological hazards over a transboundary Alpine area based on different methods and climate datasets

Abstract

Extreme hydrometeorological events are an increasing concern in Central Europe, particularly in the European Alps, where they pose significant threats to ecological and socio-economic systems. Risk assessment and management can benefit from a better understanding of how the choice of climate datasets and extreme definition affect the identification of extreme meteorological events capable of triggering hazardous events. Considering precipitation as a key triggering factor for geohydrological hazards, such as landslides and floods, this study compares different meteorological products and statistical definitions for the identification and characterization of extreme precipitation events over 2003–2020 in a transboundary Alpine region between Austria and Italy. In particular, the temporal and spatial match of detected precipitation events and available hazard records is analysed. Three statistical metrics of 1 d precipitation are adopted for identifying the most intense precipitation events, each of them accounting for a different aspect of extreme conditions, i.e., the spatial extent of intensities, the local intensity peak and the magnitude. The metrics are calculated from four gridded meteorological products with spatial resolution ranging from 1 to 9 km. These products include observation-based data (SPARTACUS-TST) and radar-aided products (INCA) as well as reanalyses (CERRA-Land and ERA5-Land). The sets of most intense precipitation events identified by each combination of datasets and statistical approaches are evaluated against hazard occurrences in both space and time. Extreme events defined by the local intensity, i.e., the spatial 99th percentile over the study area, show the highest correspondence with observed hazards. In terms of datasets, the two 1 km observation-based products, especially INCA, offer a more detailed and reliable representation of precipitation intensities and exhibit a stronger association with hazards. Contrarily, the 9 km ERA5-Land reanalysis shows the lowest performance in precipitation extreme identification across all metrics and in their match with recorded hazards, both spatially and temporally. The spatial scales resolved by the meteorological data and/or accounted for by the metrics applied to define extreme events play a crucial role in capturing processes able to trigger geohydrological hazards in the study area. These findings can support the formulation of methodological recommendations for assessing transboundary risks associated with precipitation extremes and related geohydrological processes in the Alpine regions.