Abstract
The project TRANSLATE (climaTe Risk informAtion from eNSembLe weAther and climaTe prEdictions) addresses the need to improve the understanding and predictability of hydrological extremes over the European region. TRANSLATE aims to develop event-based storylines that describe the unfolding of flood events that are unprecedented in the relatively short observational record using seasonal reforecasts of river streamflow from the European Flood Awareness System (EFAS). The aim of TRANSLATE is to develop a tool methodology to be used to develop risk storylines, including causal networks of physical and socio-economic indicators of extreme weather events. The ultimate goal is to produce actionable information that can assist in the implementation of adaptation actions by local stakeholders. The project builds on an integrated methodology that combines ensemble prediction systems with an open-source impact model to estimate the expected damage of unprecedented extremes. The core result is a catalogue of unprecedented flood events with high potential for direct economic damage to the target region, illustrating the patterns of atmospheric circulation associated with high-impact events. The tool methodology can be applied to any catchment where input data can be systematically validated due to its customizable structure.
Results achieved
: Introduction and General Objectives of the PRIN Project In the context of anthropogenic climate change, the assessment of risk associated with extreme hydrometeorological events can no longer be based solely on the assumption that observed historical time series are stationary. So-called “low-likelihood, high-impact” events—or flood events without historical precedent—represent the most complex challenge for infrastructure resilience and civil protection planning. The PRIN project aims to overcome the inherent limitations of observed historical records by leveraging the computational and statistical power of ensemble weather and climate forecasts (ensemble). The fundamental objective is to extract predictive and scenario-based information—not so much for short-term operational forecasting as for quantifying the internal variability of the climate system and identifying plausible but as-yet-unobserved extreme scenarios. 2. The Methodological Framework: The UNSEEN Approach The innovative core of the research is based on the definition and standardization of the UNSEEN (UNprecedented Simulated Extremes using ENsembles) approach. Real historical observations provide us with a single time path, or a single climate trajectory. In contrast, seasonal and climate prediction models generate retrospective simulations (reforecasts) composed of dozens of parallel trajectories (ensemble members), each representing a physically consistent evolution of the atmosphere starting from imperceptibly different initial conditions. In the developed framework, simulations from the SEAS5 system of the European Centre for Medium-Range Weather Forecasts (ECMWF) were used, featuring a 25-member ensemble over a time horizon ranging from 2000 to 2023. Using a technique of statistical concatenation of quarterly blocks extracted from different initialization dates (corresponding to the beginning of April, May, June, and July), the researchers generated a set of 100 surrogate time series. 3. Modeling Chain: From Atmospheric Synoptics to Territorial Impact The added value of the project lies in having developed an integrated modeling chain capable of translating a barometric anomaly at high altitude into an estimate of economic damage across the territory. The process consists of three sequential phases: • Atmospheric Forcing: The SEAS5 ensemble dataset provides large-scale precipitation and temperature parameters. • Hydrological Modeling: The meteorological forcings are processed by the LISFLOOD distributed hydrological model, integrated into the fifth version of the European Flood Awareness System (EFAS5). This step converts the cumulative rainfall over the watershed into daily river discharge. • Impact Modeling: The peak discharges of UNSEEN events are finally fed into CLIMADA (CLImate ADAptation), an open-source platform for probabilistic risk modeling. CLIMADA cross-references the event intensity (the expected water level) with exposure (maps of economic assets, population, and infrastructure) and vulnerability curves, calculating the direct economic impact expressed in monetary terms. 4. Case Study Analysis: Panaro and Reno The application of the basin-scale framework focused on two watercourses in Emilia-Romagna: the Panaro River (analyzed up to the Bomporto monitoring station, with a sub-basin area of 1,775 square kilometers) and the Reno River (Casalecchio monitoring station, with a sub-basin area of 4,828 square kilometers). Despite their geographical proximity and a distance of only 33 kilometers between the two measuring stations, the application of the UNSEEN approach revealed profoundly asymmetric hydrological and meteorological-climatic responses. An examination of extreme quantiles and interannual trends revealed that on the Panaro River, at the Bomporto station, the UNSEEN surrogate series were able to generate peak discharges markedly higher than those recorded both in the historical dataset simulated by EFAS and in the historical observations from the ARPAE network. This demonstrates that, for this basin, the ensemble data extraction approach is a powerful tool for revealing physically possible but as-yet-unobserved flood scenarios. In contrast, on the Rhine, at the Casalecchio station, the surrogate series failed to exceed the historical maximums observed. Furthermore, analysis of the annual cycle showed that the surrogate series tends to systematically underestimate autumn extremes in this second basin. To understand this discrepancy, the study analyzed large-scale meteorological fields, extracted both from the ERA5 reanalysis for historical events and directly from the SEAS5 simulations for the UNSEEN scenarios. It emerged that the meteorological causes underlying critical floods in the two rivers are driven by different atmospheric dynamics. The most intense flooding events on the Panaro River are typically linked to the presence of a deep trough over the central-western Mediterranean or to the development of a well-structured cyclone over central Italy—configurations capable of triggering intense, moisture-laden easterly/northeasterly flows that impact the Apennine side. In contrast, critical floods on the Reno River appear to be linked to the prevalence of a predominantly zonal flow—that is, from the west—driven by a deep and extensive Atlantic-origin low-pressure system. This distinction explains the model’s varying performance: the ensemble weather models examined show differing abilities to reproduce the frequency and intensity of these specific synoptic configurations, highlighting that the UNSEEN approach must always be validated by considering the specific meteorological sensitivity of each individual watershed. 5. Final Products and Practical Applications The scientific results achieved within this PRIN research line translate into operational tools to support society and land-use planning: • Generation of Unprecedented Flood Catalogs: The framework enables the compilation of standardized databases of synthetic yet realistic extreme events. These catalogs provide a robust statistical basis for redefining the return periods of flood discharges, overcoming the limitations of traditional parametric methods applied to short time series. • Risk Scenarios for Stress Testing: High-impact simulations drawn from the catalog provide coherent meteorological and hydrological storylines. Basin authorities and managers of critical infrastructure—such as bridges, levees, and hydraulic structures—can use these scenarios to verify the resilience of these structures when faced with unprecedented stresses. Bianco, E., Davini, P., Zappa, G., Manzato, A., Giordani, A., & Ruggieri, P. (2026). A framework for generating catalogues of high-impact UNSEEN flood events. Climate Services, 42, Article 100504. https://doi.org/10.1016/j.cliser.2026.100504 Manzato A. et al. (in review). How strongly do seasonal precipitation biases in SEAS5 affect EFAS river discharge trends and upper-tail statistics? The case of two Italian mid-size catchmentsProject details
Unibo Team Leader: Paolo Ruggieri
Unibo involved Department/s:
Dipartimento delle Arti
Dipartimento di Fisica e Astronomia "Augusto Righi"
Coordinator:
ALMA MATER STUDIORUM - Università di Bologna(Italy)
Total Eu Contribution: Euro (EUR) 237.637,00
Total Unibo Contribution: Euro (EUR) 130.637,00
Project Duration in months: 24
Start Date:
30/11/2023
End Date:
28/02/2026
