Dommenget et al. (2026) Basic atmospheric dynamics control on ENSO and tropical basin interactions

Identification

• Journal: Climate Dynamics
• Year: 2026
• Date: 2026-04-01
• Authors: Dietmar Dommenget, Jie Wang
• DOI: 10.1007/s00382-026-08144-w

Research Groups

• Monash University, Clayton, VIC, Australia

Short Summary

This study uses a simplified atmospheric model to investigate how basic atmospheric dynamics, influenced by the shape and interaction of tropical ocean basins, control the growth rate (Bjerknes feedback) and period of the El Niño Southern Oscillation (ENSO). It finds that atmospheric processes, including an optimal zonal length for heat sources and the influence of remote basins, strongly regulate ENSO dynamics, challenging existing theories that often neglect these factors.

Objective

• To explore how simplified atmospheric dynamics, as represented by a Matsuno-Gill-type model, control the growth rate (Bjerknes feedback) and period of ENSO, considering the effects of heat source shape (zonal and meridional width) and interactions with remote tropical ocean basins.

Study Configuration

• Spatial Scale:
  • GFDL CM2.1: Global, low-resolution atmosphere (3.0° longitude × 3.75° latitude), coarse resolution ocean (1° longitude × 1.5° latitude).
  • Matsuno-Gill model: Tropical domain from 68° S to 68° N, with a resolution of 2.8125° longitude × 1.126° latitude.
  • Idealized heat sources: Zonal widths from 10° to 200° longitude, meridional widths from 4° to 20° latitude.
• Temporal Scale:
  • GFDL CM2.1 simulations: 250 years, with the last 200 years analyzed.
  • Analysis of variability: Monthly mean anomalies, seasonal to interannual timescales.
  • Lag-lead correlations: From -2 years to +2 years.

Methodology and Data

• Models used:
  • Geophysical Fluid Dynamics Laboratory (GFDL) Climate Model, version 2.1 (CM2.1) for idealized world simulations.
  • Shallow water atmosphere (Matsuno-Gill-type) model (based on Matsuno, 1966 and Gill, 1980) for idealized heat source experiments and forced simulations.
• Data sources:
  • Output from GFDL CM2.1 idealized world simulations (14 different configurations of land-sea distribution, ocean bathymetry, and land topography).
  • Output from Matsuno-Gill model simulations, forced by prescribed mass (heat) sources mimicking idealized patterns or SST regression patterns from GFDL simulations.

Main Results

• Equatorial heat sources have an optimal zonal length of approximately 40° longitude for the strongest equatorial zonal wind response (Bjerknes feedback).
• Meridional wind convergence towards the equator monotonically increases with the zonal length of the heat source.
• Basins larger than the Pacific can exhibit weaker ENSO variability due to atmospheric dynamics controlling wind stress, contrary to some existing theories.
• Increasing the meridional width of the heat source enhances the Bjerknes feedback, reduces meridional wind convergence, and shifts the maximum wind curl further away from the equator, which decreases oceanic Rossby wave speeds.
• These changes in wind patterns force ENSO patterns in larger basins to be more equatorially confined and shorten the ENSO period.
• Interactions with heat sources in remote tropical ocean basins can strongly enhance the Bjerknes feedback, exerting a stronger control on ENSO dynamics than basin size alone.
• Out-of-phase remote heat sources can further control the period of ENSO.
• The Matsuno-Gill model successfully reproduces the weakening of the Bjerknes feedback with increasing basin width and captures the spatial patterns of wind response in multi-basin setups, showing high correlations (0.7-0.8 for pattern, 0.8 for WBOX wind index).

Contributions

• Challenges the conventional ENSO theory that larger tropical ocean basins necessarily lead to stronger ENSO variability and that tropical basin interactions are irrelevant for ENSO strength and period.
• Provides a fundamental understanding of how the zonal and meridional extent of atmospheric heating patterns, as well as remote tropical basin interactions, modulate the Bjerknes feedback and oceanic Rossby wave dynamics, thereby controlling ENSO growth rate and period.
• Identifies an optimal basin size for ENSO strength (around 40° longitude) due to global atmospheric dynamics, an effect not previously highlighted in ENSO literature.
• Demonstrates that remote heat sources/sinks can significantly amplify the Bjerknes feedback and influence the ENSO period through out-of-phase forcing.
• Offers a simplified framework to guide paleoclimate research on ENSO variability in past Earth configurations with different land-ocean distributions.
• Lays a foundation for revising current ENSO theory to explicitly incorporate the detailed dynamics of tropical atmospheric response to heating patterns and the influence of remote tropical ocean heating.

Funding

• Australian Research Council (ARC), Centre of Excellence for Climate Extremes (Grant Number: CE170100023).
• China Scholarship Council (for Jie Wang's visit to Monash University).
• National Computational Infrastructure (NCI Australia) for model simulations.

Citation

@article{Dommenget2026Basic,
  author = {Dommenget, Dietmar and Wang, Jie},
  title = {Basic atmospheric dynamics control on ENSO and tropical basin interactions},
  journal = {Climate Dynamics},
  year = {2026},
  doi = {10.1007/s00382-026-08144-w},
  url = {https://doi.org/10.1007/s00382-026-08144-w}
}