Caby et al. (2025) Dynamical properties of atmospheric and sea level extremes in the North-East Atlantic

Identification

• Journal: Climate Dynamics
• Year: 2025
• Date: 2025-12-19
• Authors: Théophile Caby, Lucia Pineau‐Guillou, Florian Sévellec, J.‐M. Delouis
• DOI: 10.1007/s00382-025-08008-9

Research Groups

• Laboratoire d’Océanographie Physique et Spatiale, Univ Brest CNRS Ifremer IRD, Brest, France
• ODYSSEY Project-Team, INRIA CNRS, Brest, France

Short Summary

This study characterizes the dynamical properties of atmospheric pressure and sea level storm surges in the North-East Atlantic using a novel phase space velocity metric (v) and local dimension (d), revealing that extreme events are associated with lower dimensions and higher instability.

Objective

• To characterize the dynamical properties of atmospheric pressure and sea level storm surges in the North-East Atlantic, focusing on extreme events, by applying dynamical systems theory tools and introducing a novel metric, the normalised phase space velocity (v).

Study Configuration

• Spatial Scale:
  • Atmosphere: North Atlantic (40° W to 10° E, 30° N to 70° N) on a 1° grid.
  • Ocean: Eastern North-East Atlantic (20° W to 10° E, 40° N to 60° N) on a 0.4° grid.
• Temporal Scale:
  • Period: 1960–2010 (50 years).
  • Resolution: 6 hours.

Methodology and Data

• Models used:
  • TUGO global ocean model (for generating surge hindcast data).
  • Dynamical systems theory approach, including local dimension (d), persistence index ((\theta)), and a newly introduced normalised phase space velocity (v).
• Data sources:
  • Atmospheric: 20th Century Reanalysis version 3 (20CR) for sea level pressure (SLP) and wind data.
  • Oceanic: ClimEx 1900-2015 sea level hindcast for storm surge data.

Main Results

• A novel metric, the normalised phase space velocity (v), was introduced to track the acceleration of dynamics near extreme events, providing an intuitive and easy-to-compute measure of state stability without relying on a radius.
• The mean local dimension (d) of the attractors for both atmospheric pressure (8.9) and oceanic surges (9.5) is significantly smaller than their respective phase space dimensions (2601 and 3876), indicating strong correlations between grid points.
• Extreme events, whether stormy (low SLP or high surges) or calm (high SLP or negative surges), are characterized by low local dimensions, suggesting they form from a limited set of predictors.
• Stormy states exhibit significantly higher instability (high v) compared to calm states (low v).
• High atmospheric pressure systems (anticyclones) generally last longer than low-pressure systems, consistent with their lower phase space velocity (v).
• Summer states tend to be more persistent (lower v) than winter states for both atmospheric pressure and oceanic surges.
• Intense stormy events (lowest atmospheric pressure, strongest winds, largest storm spatial extension) are associated with minimum local dimension and largest phase space velocity.
• The local dimension of the atmospheric system (d) is strongly negatively correlated with the spatial extension of low-pressure structures (Pearson correlation coefficient (r=-0.73)), with the largest storms reaching up to 2 x 10(^{12}) m(^2).
• Around extreme events, the dynamics show a sharp increase in velocity (v) before the peak of a storm, followed by a decrease in both local dimension (d) and v. For calm events, v remains relatively constant, while d decreases.

Contributions

• Introduction of a novel, dimensionless, and easily computable metric, the normalised phase space velocity (v), which quantifies state stability and allows for fair comparison across different states and systems without relying on a fixed or variable radius.
• Demonstrated that computing the inverse persistence index ((\theta)) with a fixed radius (as proposed) assigns low persistence to stormy states and high persistence to calm states, resolving issues associated with variable radius computations used in previous studies.
• First application of dynamical systems theory tools (local dimension and persistence/velocity) to the analysis of storm surges at a regional scale, providing new insights into their underlying dynamics.
• Characterization of the evolution of dynamical properties (d and v) around extreme atmospheric and oceanic events, revealing distinct patterns for stormy versus calm conditions.

Funding

• ISblue project, Interdisciplinary graduate school for the blue planet (ANR-17-EURE-0015)
• French government program “Investissements d’Avenir” (France 2030)
• French National Research Agency (ANR) Grant ClimEx (ANR-21-CE01-0004)
• U.S. Department of Energy, Office of Science Biological and Environmental Research (BER) (for 20CR data support)
• National Oceanic and Atmospheric Administration Climate Program Office (for 20CR data support)
• NOAA Physical Sciences Laboratory (for 20CR data support)

Citation

@article{Caby2025Dynamical,
  author = {Caby, Théophile and Pineau‐Guillou, Lucia and Sévellec, Florian and Delouis, J.‐M.},
  title = {Dynamical properties of atmospheric and sea level extremes in the North-East Atlantic},
  journal = {Climate Dynamics},
  year = {2025},
  doi = {10.1007/s00382-025-08008-9},
  url = {https://doi.org/10.1007/s00382-025-08008-9}
}