Tropics-wide intraseasonal oscillations

Jiawei Bao, Sandrine Bony, Daisuke Takasuka, Caroline MüllerProceedings of the National Academy of Sciences 2025 · DOI

TL;DR

This paper identifies a previously underappreciated "Tropics-wide Intraseasonal Oscillation" (TWISO) that significantly impacts the tropical climate, operating on a 30-60 day cycle. This oscillation, driven by complex interactions between convection, radiation, and surface fluxes, challenges the notion that only slower climate modes significantly influence the tropical mean state.

What Problem Does This Paper Address?

The tropical climate system is known for its variability, but traditionally, the dominant influences on the tropical mean state were attributed to slower processes like the seasonal cycle or interannual variations (e.g., El Niño-Southern Oscillation). Faster, intraseasonal oscillations, like the well-known Madden-Julian Oscillation (MJO), were thought to primarily affect regional weather patterns without leaving a significant imprint on the tropical average. However, recent work suggested that convective organization might have radiative impacts at the tropical mean scale, raising the question: what kind of intraseasonal phenomenon could be responsible for such a widespread signal? This paper aims to identify and characterize such a phenomenon, explore its drivers, and understand its relationship with known tropical oscillations like the MJO and the fundamental instability of radiative-convective equilibrium (RCE).

Methods

The researchers employed a multi-pronged approach:

  • Data Analysis: They analyzed daily time series of key tropical climate variables (sea surface temperature, 500 hPa air temperature, and net top-of-atmosphere radiation) from satellite observations and reanalysis datasets.
  • Deseasonalization and Filtering: To isolate intraseasonal variability, they first removed the daily climatological seasonal cycle and then applied a bandpass filter to retain oscillations with periods between 30 and 60 days. This filtered signal was denoted with a subscript 'f'.
  • Correlation and Regression Analysis: They used lead-lag correlation and regression analyses to examine the relationships between different variables and to map the spatial patterns associated with the identified oscillation. This allowed them to understand how different components of the climate system covary and propagate.
  • MJO Indexing: They utilized established indices (RMM and OMI) to define active and inactive MJO periods for composite analysis.
  • RCE Instability Framework: They drew upon theoretical insights from radiative-convective equilibrium models, which predict inherent instabilities that can lead to oscillations.

Datasets

  • ERA5 Reanalysis Data: This was the primary dataset used for most analyses, covering the period 2001-2020.
  • Satellite Observations: Used for validation and comparison, including:
  • OISST (Optimum Interpolation Sea Surface Temperature)
    • MERRA2 (Modern-Era Retrospective analysis for Research and Applications, Version 2)
    • CERES (Clouds and the Earth's Radiant Energy System)
  • GPM IMERG: Used for daily precipitation data to quantify mesoscale convective organization.
  • MJO Indices:
    • Real-time Multivariate (RMM) Index (from the Bureau of Meteorology of Australia)
    • OLR-based MJO Index (OMI) (from NOAA)

Key Insights and Ideas

  • A "Tropics-Wide" Intraseasonal Oscillation (TWISO) Exists: The most significant finding is the clear detection of a coherent 30-60 day oscillation that spans the entire tropics, impacting not just regional phenomena but the tropical mean state itself. This challenges previous assumptions about the scale of intraseasonal influence.
  • Interactions are Key: TWISO isn't driven by a single factor but emerges from a complex interplay between moist convection, radiation, surface fluxes (heat exchange between ocean and atmosphere), and large-scale atmospheric circulation (Hadley and Walker cells).
  • MJO is a Modulator, Not a Prerequisite: While the MJO often coincides with and amplifies TWISO, the oscillation can occur independently. This suggests TWISO is a more fundamental process, potentially linked to the inherent instability of the tropical climate system.
  • RCE Instability as a Potential Driver: The findings align with theoretical predictions from radiative-convective equilibrium models, where fundamental interactions can lead to self-sustained oscillations even without external forcing like the MJO.
  • TWISO Impacts Tropical Energy Balance: The oscillation's influence on net radiation suggests it plays a role in modulating the tropical energy budget, with potential implications for exchanges with extratropical regions.

Key Results

  • Spectral Peaks: Power spectra of tropical-mean SST, T500, and Ntoa show pronounced peaks in the 30-60 day intraseasonal band.
  • Variance Contribution: Filtered intraseasonal oscillations account for a significant portion of the variability in net TOA radiation (31%) and 500 hPa temperature (26%).
  • Covariance: The filtered tropical mean overturning circulation index (If) is strongly correlated with T500,f (R ≈ 0.69, If leads by 8 days) and negatively correlated with SST,f (R ≈ -0.51, If leads by 10 days).
  • Spatial Propagation: Regression analysis reveals an eastward propagation of atmospheric warming anomalies originating from the Maritime Continent, followed by strengthening of large-scale circulation and surface winds.
  • MJO Connection: Significant correlations exist between TWISO and MJO indices, with RMM1 leading TWISO and RMM2 lagging. Peak absolute correlations reach 0.6.
  • Independence from MJO: Composite analysis shows that TWISO persists even during inactive MJO periods, albeit with potentially reduced amplitude.

Limitations and Future Work

  • Distinguishing Drivers: While the study strongly suggests TWISO is a fundamental process, disentangling the precise contributions of MJO influence versus inherent RCE instability from observational data alone is challenging. The authors recommend future work combining modeling and mechanism-denial experiments.
  • Model Representation: It remains an open question whether current climate models can reliably represent TWISO, given its complex interactions.
  • Extratropical Connections: The role of extratropical processes and their interaction with TWISO, particularly concerning atmospheric angular momentum, warrants further investigation.
  • Broader Impacts: The full implications of TWISO for tropical weather and climate, including its potential to increase extreme event likelihood and influence global weather patterns, require deeper exploration.

Summary updated

Abstract

The tropical climate variability is characterized by various oscillations across a range of timescales. Oscillations that imprint the tropical mean state are generally attributed to slow processes, such as the seasonal cycle or interannual variability. Here, we identify a pronounced tropics-wide intraseasonal oscillation (TWISO) in satellite observations and reanalyses. This oscillation, with a period of 30 to 60 d, is evident across multiple variables and involves interactions between convection, radiation, surface fluxes, and large-scale circulation. It is primarily manifested as convective perturbations in the tropical Indo-Pacific warm pool accompanied by oscillations in the large-scale tropical overturning circulation. Here, we examine the relationship between TWISO, the Madden-Julian Oscillation (MJO), and the instability of radiative-convective equilibrium. Certain phases of TWISO coincide with specific phases of the MJO, suggesting a potential connection between the two. However, although the MJO can amplify the oscillation amplitude of TWISO, it is not essential for TWISO to occur. Finally, due to its broad manifestation across the tropics, TWISO potentially exerts widespread influence on tropical weather and climate at regional scales.
BibTeX
@inproceedings{bao2025tropics,
  author = {Jiawei Bao and Sandrine Bony and Daisuke Takasuka and Caroline Müller},
  title = {{Tropics-wide intraseasonal oscillations}},
  booktitle = {Proceedings of the National Academy of Sciences},
  year = {2025},
  doi = {10.1073/pnas.2511549122},
}
Figure