Berton et al. (2026) Multi-scale dynamics and mesoscale drivers of the catastrophic September 2022 Marche (central Italy) flood

Identification

• Journal: Natural hazards and earth system sciences
• Year: 2026
• Date: 2026-11-03
• Authors: Matteo Berton, Elenio Avolio, Mario Marcello Miglietta
• DOI: 10.5194/nhess-26-5541-2026

Research Groups

• Department of Biological and Environmental Sciences and Technologies, University of Salento, Lecce, Italy
• Department of Science and Technology, University of Naples “Parthenope”, Naples, Italy
• Department of Geosciences, University of Padua, Padua, Italy
• National Research Council of Italy, Institute of Atmospheric Sciences and Climate (CNR-ISAC), Lamezia Terme, Italy
• National Research Council of Italy, Institute of Atmospheric Sciences and Climate (CNR-ISAC), Padua, Italy
• Department of Atmospheric and Cryospheric Sciences, University of Innsbruck, Innsbruck, Austria

Short Summary

This study used high-resolution numerical simulations to investigate the multi-scale dynamics and mesoscale drivers of the catastrophic September 2022 Marche flood in central Italy, revealing the critical role of Apennine orography in anchoring the storm and the extreme sensitivity of rainfall patterns to initial and boundary conditions.

Objective

• To investigate the atmospheric conditions and multi-scale dynamical mechanisms that drove the catastrophic September 2022 Marche flood in central Italy, focusing on physical precursors and dynamic mechanisms rather than operational predictive skill.

Study Configuration

• Spatial Scale: Three one-way nested domains with horizontal grid spacings of 10 km (200x200 grid points), 2 km (151x151 grid points), and 0.66 km (172x130 grid points), covering central Italy and the Marche region. The model atmosphere was discretized into 41 vertical levels, extending from the surface up to a top pressure of 50 hPa.
• Temporal Scale: The simulation was initialized at 00:00 UTC on 15 September 2022, with a duration of 24 hours, focusing on the event of 15 September 2022.

Methodology and Data

• Models used: Advanced Research WRF (WRF-ARW) model, version 4.6.1, with a convection-permitting configuration for the intermediate (2 km) and innermost (0.66 km) domains. Physics parameterizations included Thompson microphysics, RRTMG longwave and shortwave radiation, Noah-MP land surface model, YSU planetary boundary layer, and Revised MM5 surface layer. The Tiedtke cumulus scheme was used for the coarsest domain (10 km).
• Data sources:
  • Initial and Lateral Boundary Conditions (Control Run): NCEP GDAS/FNL 0.25° Global Tropospheric Analyses and Forecast Grids, updated every 3 hours.
  • Sensitivity Experiments I.C./B.C.: NCEP GFS historical archive, ECMWF operational high-resolution (HRES) data, and ERA5 reanalysis data (0.25° resolution).
  • Observational Data for Validation: H-SAF H03B satellite instantaneous precipitation rate, Italian Department of Civil Protection radar network (Vertical Maximum Intensity radar reflectivity), EUCLID/Siemens BLIDS network (lightning detection data), Pratica di Mare radiosoundings (WMO ID: 16245), and validated rain gauge observations from the Civil Protection Department of the Marche Region (SIRMIP) and the Italian Department of Civil Protection (Dewetra platform).

Main Results

• The catastrophic flood was driven by the interaction of warm, moist air masses with the central Apennines, which triggered and anchored quasi-stationary convective cells and rainbands.
• A persistent low-level convergence line, generated by the interaction of the low-level flow with local topography (extending northeastward from Mount Amiata), contributed significantly to the uplift of moist air.
• Atmospheric instability, initially suppressed by a capping inversion ("Loaded Gun" sounding profile), was released as humid air arrived at low levels, eroding the inhibition. Upper-level dynamics (jet streak) and mountain-induced gravity waves also played a role in sustaining vertical motion.
• The WRF model qualitatively reproduced the storm's spatial position, orientation, and quasi-stationary evolution, but underestimated peak rainfall accumulations (160–180 mm simulated vs. over 400 mm observed) and delayed the event by 1–2 hours.
• Simulated precipitation decay was linked to near-surface thermodynamic biases, with the model showing a warmer and drier environment that failed to maintain saturation.
• Sensitivity experiments demonstrated a strong dependence on initial and boundary conditions and physical parameterizations. The ECMWF-driven simulation failed to reproduce the correct rainfall due to a weaker surface pressure gradient, reduced moisture advection, and lower instability, causing a temporal mismatch between low-level moisture supply and upper-level forcing.
• Orography was crucial for anchoring the system and providing mechanical lift; sensitivity tests showed a significant reduction in rainfall intensity (up to 60%) with reduced orography.
• High Integrated Vapor Transport (IVT) and Precipitable Water (PW) values (exceeding 40–50 mm) were identified as critical for fueling convection.

Contributions

• Provides a detailed multi-scale dynamical analysis of the catastrophic September 2022 Marche flood using high-resolution WRF simulations, explicitly resolving meso-gamma scale terrain features.
• Identifies the critical role of the Apennine orography in anchoring stationary convective systems and facilitating low-level convergence, enhancing understanding of localized severe orographic precipitation.
• Highlights the extreme sensitivity of localized heavy rainfall predictability to initial and boundary conditions and model physics, particularly the timing and location of low-level moisture transport relative to topography and upper-level forcing.
• Demonstrates how subtle differences in low-level atmospheric dynamics (e.g., surface pressure gradients, moisture advection) in different forcing datasets can lead to significant discrepancies in simulated rainfall patterns and intensity for extreme events.

Funding

• Project "REDUCE METEO HAZARDS (RMH)", BRiC-2024 call.

Citation

@article{Berton2026Multiscale,
  author = {Berton, Matteo and Avolio, Elenio and Miglietta, Mario Marcello},
  title = {Multi-scale dynamics and mesoscale drivers of the catastrophic September 2022 Marche (central Italy) flood},
  journal = {Natural hazards and earth system sciences},
  year = {2026},
  doi = {10.5194/nhess-26-5541-2026},
  url = {https://doi.org/10.5194/nhess-26-5541-2026}
}