Beguerı́a et al. (2025) Evolution of extreme precipitation in Spain: contribution of atmospheric dynamics and long-term trends

Identification

• Journal: Stochastic Environmental Research and Risk Assessment
• Year: 2025
• Date: 2025-03-28
• Authors: Santiago Beguerı́a, Miquel Tomàs‐Burguera, Roberto Serrano‐Notivoli, David Barriopedro, Sergio M. Vicente‐Serrano
• DOI: 10.1007/s00477-025-02961-x

Research Groups

• Estación Experimental de Aula Dei, Consejo Superior de Investigaciones Científicas (EEAD-CSIC), Zaragoza, Spain
• Laboratorio de Clima y Servicios Climáticos (LCSC), Zaragoza, Spain
• University of Illes Balears, Palma de Mallorca, Spain
• Department of Geography and Regional Planning, Environmental Sciences Institute (IUCA), University of Zaragoza, Zaragoza, Spain
• Instituto de Geociencias (IGEO), Consejo Superior de Investigaciones Científicas - Universidad Complutense de Madrid (CSIC - UCM), Madrid, Spain
• Instituto Pirenaico de Ecología, Consejo Superior de Investigaciones Científicas (IPE-CSIC), Zaragoza, Spain

Short Summary

This study applies a non-stationary peaks-over-threshold model to 341 weather stations in Spain (1951-2020) to analyze the evolution and drivers of extreme precipitation, finding that atmospheric circulation indices significantly influence extreme event characteristics, while long-term temporal trends are less pronounced but predominantly negative when accounting for atmospheric variability.

Objective

• To assess the influence of atmospheric circulation patterns on the main characteristics of precipitation events (intensity and magnitude).
• To determine the existence of long-term trends in extreme precipitation.
• To investigate the interactions between atmospheric circulation and long-term trends.

Study Configuration

• Spatial Scale: Mainland Spain and the Balearic Islands, utilizing an observational network of 341 weather stations.
• Temporal Scale: 1 January 1951 to 31 December 2020 (70 years), using daily precipitation data.

Methodology and Data

• Models used: Non-stationary Peaks-Over-Threshold (NSPOT) model, specifically a Poisson-Generalized Pareto Distribution (P-GPD) model. Stationary P-GPD model (M0), univariate non-stationary models (M1), and a multivariate non-stationary model (M2) were developed and compared. Quantile regression (R package quantreg) was used for modeling the threshold-covariate relationship, and Maximum-Likelihood Estimation (MLE) (R package ismev) for GPD parameter estimation.
• Data sources:
  • Daily precipitation data: Observational records from the Spanish meteorological agency (AEMET), comprising 341 quality-checked and gap-filled time series.
  • Atmospheric circulation indices: Daily time series of the North Atlantic Oscillation (NAO), Mediterranean Oscillation (MO), and Western Mediterranean Oscillation (WEMO), computed as standardized differences of sea-level pressure from ERA5 reanalysis data.

Main Results

• A multivariate non-stationary model (M2), incorporating time and three atmospheric circulation indices (NAO, WEMO, MO), demonstrated significantly better performance than simpler univariate and stationary models in the majority of stations across Spain.
• Atmospheric circulation indices exert a strong and significant influence on the magnitude-frequency relationship of extreme precipitation events (daily peak intensity and accumulated event precipitation), with distinct spatial and seasonal patterns.
• The Western Mediterranean Oscillation (WEMO) shows the strongest influence, particularly a negative effect on event intensity and magnitude along the Mediterranean coast and Ebro River valley, and a positive effect towards the northwest of the study area.
• The Mediterranean Oscillation (MO) generally has a dampening (negative) effect across most of Spain, with an opposite effect on the Mediterranean coast.
• The North Atlantic Oscillation (NAO) has a comparatively smaller marginal effect, showing a dampening effect over central and southwest Spain, and an intensification in the northwest, particularly for event magnitude.
• The marginal influence of time on extreme precipitation attributes is relatively small and lacks a spatially coherent pattern.
• While no strong conclusive proof of overall temporal trends was found, the multivariate model revealed more stations with significant decreases (102 for intensity, 82 for magnitude) than increases (30 for intensity, 16 for magnitude) in extreme precipitation event attributes, especially for intensity, once atmospheric variability was accounted for.
• Highest extreme precipitation intensities (exceeding 300 mm per day) are found along the Mediterranean coast and in the Balearic Islands, while highest magnitudes (over 800 mm per event) are found in Mediterranean and Atlantic-influenced areas (North, West, and Southwest).

Contributions

• This study is the first to comprehensively consider the combined influence of large-scale atmospheric modes of variability and long-term temporal trends on extreme meteorological events across mainland Spain using Non-Stationary Extreme Value Theory (NSEVT).
• It demonstrates the superior performance of multivariate non-stationary models in disentangling the decoupled influence of atmospheric dynamics and long-term trends, thereby improving the signal-to-noise ratio for trend detection in extreme events.
• The research provides a refined spatial and seasonal analysis of the influence of NAO, WEMO, and MO on extreme precipitation intensity and magnitude across Spain.
• The developed framework offers new opportunities for improved statistical predictability of extreme events, statistical downscaling of climate model outputs, and extreme event attribution by effectively accounting for dynamic covariates.

Funding

• Spanish National Research Agency (MCIN/AEI) through project PID2020-116860RB-C22.
• R.S.-N. personal grant RYC2021-034330-I.
• European Commission - NextGenerationEU (Regulation EU 2020/2094) through CSIC’s Interdisciplinary Thematic Platform “Clima (PTI Clima) / Development of Operational Climate Services”.
• Aragón Government through grant E02-20R.
• Open Access funding provided thanks to the CRUE-CSIC agreement with Springer Nature.

Citation

@article{Beguerıa2025Evolution,
  author = {Beguerı́a, Santiago and Tomàs‐Burguera, Miquel and Serrano‐Notivoli, Roberto and Barriopedro, David and Vicente‐Serrano, Sergio M.},
  title = {Evolution of extreme precipitation in Spain: contribution of atmospheric dynamics and long-term trends},
  journal = {Stochastic Environmental Research and Risk Assessment},
  year = {2025},
  doi = {10.1007/s00477-025-02961-x},
  url = {https://doi.org/10.1007/s00477-025-02961-x}
}