Gambini et al. (2025) Uncertainty Quantification and Spatial Biases Assessment in Precipitation Forecasts: A Methodology for Real-Time Flood Forecasting Applications

⚠️ Warning: This summary was generated from the abstract only, as the full text was not available.

Identification

• Journal: Journal of Hydrometeorology
• Year: 2025
• Date: 2025-09-10
• Authors: Enrico Gambini, Giovanni Ravazzani, Marco Mancini, Ismaele Quinto Valsecchi, Alessandro Cucchi, Alberto Negretti, Silvio Davolio, Oxana Drofa, Gabriele Lombardi, Alessandro Ceppi
• DOI: 10.1175/jhm-d-24-0140.1

Research Groups

Not explicitly mentioned in the provided abstract.

Short Summary

This study proposes a methodology to assess and account for spatial biases in high-resolution convective rainfall forecasts to improve flood predictions in small watersheds. It identifies a systematic 20 km northeastward displacement error in the MOLOCH model's forecasts for northern Italy and suggests using a derived displacement probability density function to generate rainfall ensembles for hydrological uncertainty quantification.

Objective

• To propose a methodology for assessing spatial biases in rainfall forecasts from a convection-permitting meteorological model.
• To identify if the model exhibits "preferential" misplacement directions for convective events.
• To suggest a method to incorporate this spatial uncertainty into hydrological predictions.

Study Configuration

• Spatial Scale: Small watersheds, a large portion of the Lombardy Region (northern Italy), specifically the "hydraulic node of Milan" study area.
• Temporal Scale: Analysis of 64 significant convective rainfall events.

Methodology and Data

• Models used: Modello Locale in Hybrid Coordinates (MOLOCH) meteorological model, kernel density estimation (KDE) for deriving probability density functions.
• Data sources: Quantitative precipitation forecasts (QPF) from MOLOCH, observed rainfall fields.

Main Results

• The MOLOCH model exhibits an average displacement error of 20 km for convective rainfall forecasts, quantified using the fractions skill score (FSS).
• A systematic misplacement tendency was identified, with forecasts consistently shifted toward the northeast direction in the "hydraulic node of Milan" study area.
• A bidimensional rainfall displacement probability density function was obtained through KDE.
• This distribution can be used to generate an ensemble of shifted rainfall forecasts from a high-resolution deterministic model, thereby accounting for QPF misplacement uncertainty in hydrological predictions.

Contributions

• Proposes a novel methodology to quantify and characterize spatial biases in high-resolution meteorological model rainfall forecasts.
• Identifies systematic and preferential misplacement directions for convective events in a specific region.
• Introduces a method to transform a deterministic high-resolution rainfall forecast into an ensemble by considering spatial displacement uncertainty, enhancing flood forecasting.
• The methodology is generalizable and applicable to other river basins and limited-area meteorological models.

Funding

Not explicitly mentioned in the provided abstract.

Citation

@article{Gambini2025Uncertainty,
  author = {Gambini, Enrico and Ravazzani, Giovanni and Mancini, Marco and Valsecchi, Ismaele Quinto and Cucchi, Alessandro and Negretti, Alberto and Davolio, Silvio and Drofa, Oxana and Lombardi, Gabriele and Ceppi, Alessandro},
  title = {Uncertainty Quantification and Spatial Biases Assessment in Precipitation Forecasts: A Methodology for Real-Time Flood Forecasting Applications},
  journal = {Journal of Hydrometeorology},
  year = {2025},
  doi = {10.1175/jhm-d-24-0140.1},
  url = {https://doi.org/10.1175/jhm-d-24-0140.1}
}