Miró et al. (2022) A component-based approximation for trend detection of intense rainfall in the Spanish Mediterranean coast

Identification

• Journal: Weather and Climate Extremes
• Year: 2022
• Date: 2022-09-28
• Authors: Juan Javier Miró, Marc Lemus‐Canovas, Roberto Serrano‐Notivoli, Jorge Olcina Cantos, María José Estrela, Javier Martín Vide, Pablo Sarricolea, Óliver Meseguer-Ruiz
• DOI: 10.1016/j.wace.2022.100513

Research Groups

• Departament de Geografia, Universitat de Val`encia, Spain
• Andorra Research + Innovation, Andorra
• Climatology Group, Department of Geography, University of Barcelona, Spain
• Departamento de Geografía, Universidad Aut´onoma de Madrid, Spain
• Interuniversitary Institute of Geography, University of Alicante, Spain
• Departamento de Geografía, Universidad de Chile, CITRID, Programa de Reducci´on de Riesgos y Desastres, Universidad de Chile, Chile
• Departamento de Ciencias Hist´oricas y Geogr´aficas, Universidad de Tarapac´a, Chile

Short Summary

This study identified 12 spatial patterns of 899 torrential precipitation events (≥150 mm in 24 h) along the Spanish Mediterranean coast and Balearic Islands from 1950–2020, revealing significant positive trends in accumulated volumes for three eastern/southeastern components linked to unstable atmospheric situations and moist advection.

Objective

• To identify the main torrential precipitation events (≥150 mm in 24 h) along the Spanish Mediterranean coast between 1950 and 2020, determine their spatial components and trends, and obtain their synoptic characterizations at mean sea-level pressure (slp) and 500 hPa geopotential height (z500) to provide new insights into associated surface and mid-tropospheric processes.

Study Configuration

• Spatial Scale: Spanish Mediterranean coast and Balearic Islands, including an additional area at least 100 km inland from the Iberian Peninsula. Synoptic analysis covered 25°N–70°N and 30°W–30°E.
• Temporal Scale: 1950–2020 (71 years) using daily precipitation data.

Methodology and Data

• Models used:
  • Non-negative matrix factorization (NNMF) for dimensionality reduction and spatial component identification.
  • Nonparametric Sen method for trend calculation (slope in mm per decade).
  • Non-parametric Mann–Kendall test for statistical significance of trends (p-value < 0.05).
  • Vertically integrated water-vapor transport (IVT) calculated between 1,000 hPa and 300 hPa.
• Data sources:
  • 3,537 daily precipitation series from CDRD-HR-EIP and SPREAD databases, sourced from Agencia Estatal de Meteorología (AEMet), Sistema Integral de Atenci´on al Regante (SIAR), and Fundaci´on Centro de Estudios Ambientales del Mediterr´aneo (CEAM).
  • Daily mean sea-level pressure (slp) and 500 hPa geopotential height (z500) from the fifth European Centre for Medium-range Weather Forecasts Re-analysis (ERA-5) at 0.25° horizontal resolution.

Main Results

• 899 torrential precipitation events (≥150 mm in 24 h) were identified between 1950 and 2020.
• 12 distinct spatial components for these events were identified using NNMF, with 10 describing mainland events and 2 characterizing events over the Balearic Islands.
• Three components (6, 7, and 8), located in the eastern and southeastern areas, exhibited statistically significant positive trends in accumulated volumes. Component 8 showed an average annual weight increase of 75.3% and over 100% for its 99.9th percentile between 1950 and 2020.
• Torrential rainfall events are strongly linked to unstable atmospheric situations in the mid-troposphere (troughs and cut-off lows) and SW–SE advection fluxes transporting moist air from the Western Mediterranean.
• Events in the Balearic Islands (Components 9 and 11) are more associated with E–NE advections and significant orographic components (e.g., Tramuntana mountain range).
• The majority of these events occurred in early autumn (September and October), correlating with the highest Mediterranean Sea surface temperatures.
• Wind direction is a crucial factor, with maximum precipitation often occurring where surface flow is perpendicular to the coastline or prominent mountain ranges.
• Significant water vapor transport is observed, particularly when moisture trajectories originate in the Western Mediterranean, an area characterized by high sea surface temperatures.

Contributions

• Introduces a novel component-based approximation (NNMF) for analyzing trends in intense rainfall, which is particularly effective for non-negative and sparse precipitation data.
• Provides a comprehensive classification and characterization of 12 distinct spatial patterns of torrential precipitation events across the Spanish Mediterranean coast and Balearic Islands over a 71-year period.
• Quantifies significant positive trends in accumulated volumes for specific eastern and southeastern regions, enhancing understanding of regional climate change impacts on extreme rainfall.
• Offers detailed synoptic characterizations (slp, z500, IVT) for each spatial component, establishing clear links between specific atmospheric circulation patterns, moisture sources, and regional extreme rainfall.
• Emphasizes the critical roles of mid-tropospheric instability, moist advection from the Western Mediterranean, and orographic forcing in triggering torrential rainfall, especially during autumn.
• The findings are expected to improve the understanding of processes leading to catastrophic situations and contribute to the development of more effective early alert systems and management plans.

Funding

• Proyecto UTA-Mayor N◦5807–22 from the Universidad de Tarapac´a, Chile.
• Climatology Group (2017SGR1362, Catalan Government).
• Universidad Aut´onoma de Madrid (UAM) and Comunidad de Madrid through project SI3-PJI-2021-00398.
• Natural Hazards and Global Change research group from UAM.
• Government of Arag´on through the “Program of research groups” (group H09_20R, “Climate, Water, Global Change, and Natural Systems”).
• Spanish Ministerio de Ciencia e Innovaci´on through research project PID2020-118797RB-I00 (MCIN/AEI/10.13039/501100011033).
• Generalitat Valenciana through research project PROMETEO/2021/016 (Conselleria d’Innovaci´o, Universitats, Ci`encia i Societat Digital).

Citation

@article{Miró2022componentbased,
  author = {Miró, Juan Javier and Lemus‐Canovas, Marc and Serrano‐Notivoli, Roberto and Cantos, Jorge Olcina and Estrela, María José and Vide, Javier Martín and Sarricolea, Pablo and Meseguer-Ruiz, Óliver},
  title = {A component-based approximation for trend detection of intense rainfall in the Spanish Mediterranean coast},
  journal = {Weather and Climate Extremes},
  year = {2022},
  doi = {10.1016/j.wace.2022.100513},
  url = {https://doi.org/10.1016/j.wace.2022.100513}
}