Haruna et al. (2026) Regional hotspots and contrasts in the trends of mean and extreme daily precipitation in France

Identification

• Journal: Journal of Hydrology Regional Studies
• Year: 2026
• Date: 2026-04-01
• Authors: Abubakar Haruna, Juliette Blanchet, Guillaume Evin, Emmanuel Paquet
• DOI: 10.1016/j.ejrh.2026.103378

Research Groups

• Univ. Grenoble Alpes, CNRS, INRAE, IRD, Grenoble INP (Institute of Engineering and Management Univ. Grenoble Alpes), IGE, Grenoble, France
• EDF-DTG, Grenoble, France

Short Summary

This study analyzes the spatio-temporal trends of mean and extreme daily precipitation in metropolitan France from 1950 to 2022 using a non-stationary statistical framework, revealing complex, non-uniform changes including widespread summer drying, increased autumn wet-day frequency, and localized hotspots of increasing extreme precipitation.

Objective

• To quantify and map relative trends in four key precipitation metrics throughout metropolitan France on annual, seasonal, and monthly scales: (i) wet-day frequency, (ii) mean wet-day precipitation, (iii) mean all-day precipitation, and (iv) extreme precipitation, defined as the 20-year return level.

Study Configuration

• Spatial Scale: Metropolitan France, analyzed across 934 daily rain gauge stations.
• Temporal Scale: 1950 to 2022 (73 years), with analysis conducted at annual, seasonal (winter: Jan–Mar, spring: Apr–Jun, summer: Jul–Sep, autumn: Oct–Dec), and monthly scales.

Methodology and Data

• Models used: A non-stationary statistical framework comprising a mixed-distribution model. A logistic model was used for dry-day frequency, and a nonstationary generalized gamma distribution was used for wet-day precipitation amounts. A linear relationship between generalized gamma parameters and Sea Surface Temperature (SST) anomaly was assumed. Trends were assessed using maximum likelihood estimation and statistical significance was determined via a non-parametric residual bootstrap procedure (95% confidence level).
• Data sources: Daily precipitation data from a dense network of 934 rain gauges sourced from Météo-France and Électricité de France (EDF). Temporally smoothed annual Sea Surface Temperature (SST) anomaly, spatially averaged over the Mediterranean and the near Atlantic (longitude 28° W to 19° E and latitude 36° to 58° N), obtained from the NOAA-NCDC Extended Reconstructed Sea Surface Temperature Dataset Version 5, used as a covariate.

Main Results

• Wet-day frequency: A general annual increase was observed across most of France (68% of stations, 44% significant), primarily driven by strong increases in autumn (over 90% positive trend) and moderate increases in winter. Conversely, significant drying (decrease in frequency) was observed during late summer (August and September) at 80% of stations.
• Mean wet-day precipitation: A slight annual decrease was noted. Regionally, consistent decreases were found in the Rhône Valley from December to March, while increases occurred in the northern Alps and Jura mountains from March to May.
• Mean all-day precipitation: An overall positive trend was observed across France, predominantly driven by the increase in wet-day frequency rather than an increase in the amount of wet-day precipitation. Strong decreases were seen in summer, while increases were common in autumn and spring (except in the southwest), and a decrease in the southeast during winter with increases elsewhere.
• Extreme precipitation (20-year return level): Significant annual increases were concentrated in hotspots including the Île-de-France area, Brittany, the Rhône Valley, the Cévennes, and the southern Alps. Conversely, significant decreases were observed in parts of the northern Alps, the Vosges mountain range, and the western Massif Central. Seasonally, winter and summer generally experienced decreases in extreme precipitation, while spring and autumn showed increases.
• Spatio-temporal heterogeneity: The study revealed complex, non-uniform trends with profound spatio-temporal heterogeneity across France, highlighting that monthly and seasonal patterns often diverge from broader annual averages. High internal climate variability at the monthly scale was found to often compete with the long-term climate signal.

Contributions

• Provides a comprehensive, high-resolution analysis of observed trends across the entire spectrum of daily precipitation (frequency, mean intensity, and extremes) for all of metropolitan France, leveraging a dense network of 934 rain gauges over seven decades (1950–2022).
• Employs a unified non-stationary statistical framework based on the generalized gamma distribution, conditioned on Sea Surface Temperature (SST) as a thermodynamic covariate, allowing for statistically coherent trend assessment across different intensities.
• Identifies and characterizes regional hotspots and areas of significant change, underscoring the profound spatio-temporal heterogeneity and sub-seasonal variability of precipitation trends in France, which is crucial for regional climate adaptation, water resource management, and refined flood and drought risk assessments.
• Offers a high-resolution benchmark for evaluating Regional Climate Models' ability to capture observed precipitation trends in France.

Funding

• Agence Nationale de la Recherche - France 2030 (PEPR ‘‘Transformer la modélisation du climat pour les services climatiques’’ (TRACCS) program, grant number ANR-22-EXTR-0005).
• French electric company, Électricité de France (EDF), in collaboration with CNRS/IGE.

Citation

@article{Haruna2026Regional,
  author = {Haruna, Abubakar and Blanchet, Juliette and Evin, Guillaume and Paquet, Emmanuel},
  title = {Regional hotspots and contrasts in the trends of mean and extreme daily precipitation in France},
  journal = {Journal of Hydrology Regional Studies},
  year = {2026},
  doi = {10.1016/j.ejrh.2026.103378},
  url = {https://doi.org/10.1016/j.ejrh.2026.103378}
}