Papex

Open problem · Earth science · Open

Dynamics of the El Niño–Southern Oscillation

Is the irregular El Niño–Southern Oscillation driven by random forcing, internal chaos or a regular external cause?

Summary

The El Niño–Southern Oscillation alternately warms and cools the equatorial Pacific and affects weather worldwide, but it is hard to predict more than a few months ahead; whether its irregularity is stochastic, chaotic or deterministically forced is unresolved.

Why it is open

The El Niño–Southern Oscillation (ENSO) is the alternation of warm El Niño and cold La Niña conditions in the tropical Pacific every few years, and the strongest year-to-year fluctuation of the global climate system [11]. At its core is the Bjerknes feedback: a weaker east–west temperature difference along the equator weakens the trade winds, which weakens the difference further [11].

The open question is what makes the cycle irregular. George Philander and Alexey Fedorov framed it as a choice between El Niño as one phase of "a continual, self-sustaining natural mode" of the ocean and atmosphere, and El Niño as a temporary departure from normal conditions "triggered" by a random disturbance such as a burst of westerly winds [6]. Theory allows a third answer: in simple models, nonlinear resonance with the seasonal cycle can make the oscillation chaotic [1, 2]. External forcing, such as volcanic eruptions, may also pace events [7].

Some El Niño events warm mainly the eastern Pacific and others the central Pacific; how far this diversity is predictable relates to whether ENSO's complexity comes mainly from random processes or slower deterministic dynamics, and it matters for regional effects such as those on tropical cyclones [11] (see Tropical cyclone frequency). Forecasts started before and during boreal spring are more uncertain, the spring predictability barrier [10]. Its response to greenhouse warming also remains uncertain [11].

Long simulations show why century-long observational records may be too short to constrain ENSO variability [8].

What is known

Oscillator theory (1994–1997). In Fei-Fei Jin's recharge oscillator, the Bjerknes feedback makes anomalies grow, and the recharge and discharge of heat in the upper equatorial ocean turns El Niño into La Niña and back; over a wide range of coupling strengths the oscillation "can be either self-excited or stochastically sustained", with a period of 3–5 years [4]. In 1994 Jin, David Neelin and Michael Ghil showed that an El Niño model becomes chaotic as nonlinear resonances with the annual cycle overlap [1]; in the same issue of Science, Eli Tziperman and colleagues modelled ENSO as "a low-order chaotic process driven by the seasonal cycle" [2].

Noise-driven views (1995–2003). From sea surface temperatures of 1950–90, Cécile Penland and Prashant Sardeshmukh argued that the tropical Indo-Pacific can be modelled as a stable linear system driven by spatially coherent Gaussian white noise [3]. In 22 years of observations, William Kessler found that the system can stay in a weak La Niña state for up to two years, so that the memory of earlier influences would be lost, and read El Niños as "event-like disturbances to a stable basic state" [5]. In 2003 Philander and Fedorov concluded from stability analyses, energetics and simulations that the observed oscillation is "a weakly damped mode that is sustained by random disturbances" [6].

Weather noise (2018 review). Timmermann and colleagues describe wind events as triggers: ocean heat content alone does not guarantee a strong El Niño [11].

Long records (2003–2013). Using palaeoclimate reconstructions and volcanic chronologies from AD 1649, Brad Adams, Michael Mann and Caspar Ammann found a multi-year El Niño-like response to explosive tropical eruptions, roughly doubling the probability of an El Niño in the following winter, a link they note "has remained controversial" [7]. Andrew Wittenberg's analysis of his 2,000-year model run suggests that such slow modulation "need not require multidecadal memory" and can arise from the statistics of random events [8]. Fossil corals from the Northern Line Islands spanning 7,000 years show highly variable ENSO activity with no systematic trend; twentieth-century variance is higher than the fossil average but not unprecedented [9].

Seasons and diversity (2015–2018). El Niño events usually start in boreal spring, peak in winter and decay in late winter and spring [11]. Adding the annual cycle of ENSO's growth rate to a damped, noise-driven recharge oscillator produces a spring predictability barrier, which noise that depends on the ocean state makes more realistic, Aaron Levine and Michael McPhaden found [10]. The 2018 review reports that central-Pacific events became more common after 1998; coupled models can generate such changes without external forcing, but underestimate ENSO diversity [11].

Response to warming (2019–2021). Richard Seager and colleagues noted that state-of-the-art models predict a weaker west-to-east temperature gradient across the equatorial Pacific as greenhouse gases rise, whereas the observed gradient has strengthened in recent decades; they attribute the models' error to the cold bias of their equatorial cold tongues [12]. A 2021 review led by Wenju Cai found that models that best capture key ENSO dynamics tend to project stronger ENSO sea surface temperature variability, an increase "strongly influenced by internal variability" [13]. The IPCC's 2021 assessment found "no model consensus" for a systematic change in that variability over the 21st century (medium confidence), but judged it very likely that ENSO rainfall variability will increase by the second half of the century under intermediate, high and highest emissions scenarios [14].

Forecasting practice (2018–2024). In 2024 Sen Zhao and colleagues reported that an extended nonlinear recharge oscillator (XRO), which adds ENSO's seasonally varying interactions with other climate modes, forecasts ENSO skilfully up to 16–18 months ahead, comparable to the most skilful artificial-intelligence (AI) forecasts, whose skill cannot yet be linked to specific physical processes [15].

Main approaches

A damped oscillator kept going by weather. ENSO is a stable or weakly damped mode sustained by random disturbances [3, 6]. Strongest evidence: an empirical model of observed sea temperatures with stable modes and noise [3]; observed breaks in the cycle [5]; the wind-triggered 2015/16 event and the expected events that failed to grow [11]; and a realistic spring barrier in such a model [10]. Main difficulty: the observational evidence covers a few decades [3, 5], short compared with the modulation seen in long model runs [8], and the larger amplitude of El Niño than of La Niña indicates nonlinear processes, which a linear noise-driven model lacks [11].

A self-sustained or chaotic oscillation paced by the seasons. The coupled system oscillates on its own, and its interaction with the annual cycle makes it irregular [1, 2]. Strongest evidence: the chaos scenario explains several climate model results and gives spectra consistent with the data [1]; a prediction model's output is consistent with low-order chaos [2]; and Jin's recharge oscillator can be self-excited, not only sustained by noise [4]. Main difficulty: statistical tests on observed sea temperatures point to stable, forced dynamics [3], and observed breaks in the cycle suggest that memory is lost between events [5].

External pacing. Forcing from outside the ocean–atmosphere system sets the timing of some events [7]. Strongest evidence: the roughly doubled chance of El Niño in the winter after explosive tropical eruptions since 1649 [7]. Main difficulty: the volcanic link has been controversial [7]; a model with constant forcing produces large changes in ENSO by itself [8]; and the coral record shows no clear dependence on changes in insolation [9].

Several modes and their interactions. Timmermann and colleagues propose two natural modes of the coupled system, with periods of about four and two years and patterns resembling eastern- and central-Pacific events, both close to the threshold of growth and easily excited by wind events and other climate modes [11]. Strongest evidence: the modes resemble their observed counterparts [11], and the XRO traces the enhancement of its long-range skill to the initial conditions of other climate modes [15]. Main difficulty: whether the diversity in coupled climate models really comes from this pair of modes still "needs to be studied", the authors write [11].

Open sub-questions

  1. Telling noise from chaos. Which statistic of observed or coral-based ENSO records could separate a noise-driven damped oscillator from a seasonally paced chaotic one, given that records of a century or less may not show ENSO's slow modulation [1, 3, 8]?
  2. What makes the spring barrier. In forecasts started from observed ocean heat content, how much of the spring predictability barrier comes from the seasonal cycle of growth rate and how much from state-dependent wind noise [10, 11]?
  3. Two modes in climate models. In coupled models with reduced atmospheric noise, do modes with periods near four and two years still produce eastern- and central-Pacific events [11]?
  4. Volcanic pacing. Does the higher chance of El Niño after explosive tropical eruptions hold in reconstructions and eruption records independent of those Adams and colleagues used [7]?
  5. Gradient and variability. If the observed strengthening of the east–west temperature gradient continues, does ENSO sea surface temperature variability change as the models that best simulate ENSO project [12, 13, 14]?
  6. What AI forecasts use. Is the long-lead skill of AI forecasts traceable to the same climate-mode interactions that give the XRO its 16–18-month skill [15]?

Work on this in Papex

If you have a testable answer to one of the sub-questions above, you can draft it as a private hypothesis in Papex, or prepare the draft from Claude or ChatGPT with the Papex connector. You review the exact version, sign it and decide whether to publish it; a public record that names this problem is listed on this page. A Papex record shows who signed which version and when. It is not peer review and does not mean the hypothesis is correct.

Key references

  1. Jin, F.-F., Neelin, J. D. & Ghil, M. (1994). El Niño on the devil's staircase: annual subharmonic steps to chaos. Science 264(5155), 70–72. 10.1126/science.264.5155.70 · Checked 15 Oct 2026
  2. Tziperman, E., Stone, L., Cane, M. A. & Jarosh, H. (1994). El Niño chaos: overlapping of resonances between the seasonal cycle and the Pacific ocean-atmosphere oscillator. Science 264(5155), 72–74. 10.1126/science.264.5155.72 · Checked 15 Oct 2026
  3. Penland, C. & Sardeshmukh, P. D. (1995). The optimal growth of tropical sea surface temperature anomalies. Journal of Climate 8(8), 1999–2024. 10.1175/1520-0442(1995)008<1999:TOGOTS>2.0.CO;2 · Checked 15 Oct 2026
  4. Jin, F.-F. (1997). An equatorial ocean recharge paradigm for ENSO. Part I: conceptual model. Journal of the Atmospheric Sciences 54(7), 811–829. 10.1175/1520-0469(1997)054<0811:AEORPF>2.0.CO;2 · Checked 15 Oct 2026
  5. Kessler, W. S. (2002). Is ENSO a cycle or a series of events? Geophysical Research Letters 29(23). 10.1029/2002GL015924 · Checked 15 Oct 2026
  6. Philander, S. G. & Fedorov, A. (2003). Is El Niño sporadic or cyclic? Annual Review of Earth and Planetary Sciences 31, 579–594. 10.1146/annurev.earth.31.100901.141255 · Checked 15 Oct 2026
  7. Adams, J. B., Mann, M. E. & Ammann, C. M. (2003). Proxy evidence for an El Niño-like response to volcanic forcing. Nature 426(6964), 274–278. 10.1038/nature02101 · Checked 15 Oct 2026
  8. Wittenberg, A. T. (2009). Are historical records sufficient to constrain ENSO simulations? Geophysical Research Letters 36(12). 10.1029/2009GL038710 · Checked 15 Oct 2026
  9. Cobb, K. M., Westphal, N., Sayani, H. R., Watson, J. T., Di Lorenzo, E., Cheng, H., Edwards, R. L. & Charles, C. D. (2013). Highly variable El Niño–Southern Oscillation throughout the Holocene. Science 339(6115), 67–70. 10.1126/science.1228246 · Checked 15 Oct 2026
  10. Levine, A. F. Z. & McPhaden, M. J. (2015). The annual cycle in ENSO growth rate as a cause of the spring predictability barrier. Geophysical Research Letters 42(12), 5034–5041. 10.1002/2015GL064309 · Checked 15 Oct 2026
  11. Timmermann, A., An, S.-I., Kug, J.-S., Jin, F.-F., Cai, W., Capotondi, A. et al. (2018). El Niño–Southern Oscillation complexity. Nature 559(7715), 535–545; author correction 567, E3 (2019), a missing middle initial. 10.1038/s41586-018-0252-6 · Checked 15 Oct 2026
  12. Seager, R., Cane, M., Henderson, N., Lee, D.-E., Abernathey, R. & Zhang, H. (2019). Strengthening tropical Pacific zonal sea surface temperature gradient consistent with rising greenhouse gases. Nature Climate Change 9(7), 517–522. 10.1038/s41558-019-0505-x · Checked 15 Oct 2026
  13. Cai, W., Santoso, A., Collins, M., Dewitte, B., Karamperidou, C., Kug, J.-S. et al. (2021). Changing El Niño–Southern Oscillation in a warming climate. Nature Reviews Earth & Environment 2(9), 628–644. 10.1038/s43017-021-00199-z · Checked 15 Oct 2026
  14. Lee, J.-Y., Marotzke, J., Bala, G., Cao, L., Corti, S., Dunne, J. P. et al. (2021). Future global climate: scenario-based projections and near-term information. In V. Masson-Delmotte et al. (eds.), Climate Change 2021: The Physical Science Basis. Contribution of Working Group I to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press, 553–672 (Crossref lists the 2023 printed edition, without authors). 10.1017/9781009157896.006 · Checked 15 Oct 2026
  15. Zhao, S., Jin, F.-F., Stuecker, M. F., Thompson, P. R., Kug, J.-S., McPhaden, M. J., Cane, M. A., Wittenberg, A. T. & Cai, W. (2024). Explainable El Niño predictability from climate mode interactions. Nature 630(8018), 891–898. 10.1038/s41586-024-07534-6 · Checked 15 Oct 2026

Page reviewed 15 Oct 2026

Source

Open, as listed by Wikipedia · Unsolved problems in geoscience. Checked 08 Oct 2026.

The status is the source’s. Papex does not decide whether a problem is solved.

Records on this problem

No public record addresses this problem yet.

More in Earth science