the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 16 Apr 2026
Precession-forced asymmetric continental heating shapes ENSO variability
Abstract. Geologic and modelling evidence reveals that the ENSO is strongly affected by the precession of the Earth’s rotation axis, yet the mechanisms remain unclear due to interactions among multiple forcings. Using high-resolution model simulations reconstructing the ENSO activity across a full precessional cycle, we find that ENSO is strongest during austral summer perihelion, as today. This behavior arises from asymmetric continental heating: austral summer perihelion introduces strong warming on Australia, east of the Indo-Pacific ITCZ. Because deep convection favors the warmest areas, this causes the ITCZ and the Warm Pool to shift eastwards. As a result, the Pacific’s east-west thermal contrast is reduced, lowering the threshold for oscillations of convection and amplifying ENSO activity. In contrast, boreal summer perihelion warms Afro-Eurasia, shifts the ITCZ westward and weakens ENSO. Proxy records across the Indo-Pacific support this changes in climate state. Understanding asymmetric continental heating helps us link astronomical modulation to ENSO behavior and improve long-term predictions of tropical climate change.
Status: final response (author comments only)
-
RC1: 'Comment on egusphere-2026-1183', Anonymous Referee #1, 11 May 2026
-
AC2: 'Reply on RC1', Yufei Liu, 10 Aug 2026
We sincerely thank the reviewer for this insightful and constructive suggestion. Following your suggestion, we have conducted additional analyses to further evaluate the role of upper-ocean stratification in modulating ENSO variability. These analyses are organized into four steps. First, we diagnose upper-ocean stratification differences between the two end-member experiments (270° and 90° precession phases; Fig. R1; also included in Fig. S7). Second, we examine the relationship between upper-ocean stratification and ENSO amplitude across all 24 precession experiments (Fig. R2). Third, we further evaluate the thermocline feedback (α), which represents an important oceanic coupling process through which changes in upper-ocean stratification can influence ENSO growth (Fig. R3). Finally, these diagnostics have been incorporated into the revised Fig. 4, together with other mean-state and Bjerknes feedback indicators, to provide a more comprehensive assessment of the processes linking precessional forcing to ENSO variability. Taken together, these additional analyses suggest that upper-ocean stratification represents an important intermediate oceanic pathway within the proposed framework. Specifically, asymmetric continental heating reorganizes the tropical Indo-Pacific mean state, which subsequently strengthens upper-ocean stratification and thermocline feedback. Together with changes in wind feedback, these oceanic adjustments enhance Bjerknes feedback efficiency and contribute to stronger ENSO variability under the present-day-like precessional configuration.
- AC4: 'Reply on RC1', Yufei Liu, 10 Aug 2026
-
AC2: 'Reply on RC1', Yufei Liu, 10 Aug 2026
-
RC2: 'Comment on egusphere-2026-1183', Anonymous Referee #2, 22 May 2026
Liu and colleagues document results from 24 timeslice simulations with AWI-ESM over a full precession cycle, in which boundary conditions change the seasonal and hemispheric distribution of insolation. They focus on investigating how ENSO variability responds to global mean-state shifts driven by orbital forcing. I think that these simulations and investigations therein are valuable. However, several questions arise while reading this version of the manuscript, so I do not think the central claims are supported by the analyses presented. I detail my claims below:
Major Comments:
- The manuscript does not convincingly demonstrate a mechanistic link between asymmetric land heating and ENSO variability.
- The central claim in this paper is that precession-forced asymmetric continental heating shifts the Indo-Pacific ITCZ and warm pool, modifies the Pacific east-west thermal contrast, and thereby changes ENSO amplitude. However, the analyses presented do not mechanistically demonstrate this proposed causal chain. What is shown more clearly is that Northern Hemisphere and Southern Hemisphere moist static energy, land heating, and rainfall vary differently (and perhaps somewhat anti-phased acc to Fig. 4a–b) over a precession cycle. That result, while valuable in its own right,is not sufficient to establish that these changes mechanistically control interannual ENSO variability.
- The manuscript should first distinguish how orbital forcing reorganizes the seasonal-mean tropical climate and second, how such a mean-state reorganization changes the coupled ocean-atmosphere feedbacks that generate ENSO on interannual timescales (see e.g., DiNezio et al. 2013—J.Clim; Fedorov et al. 2020—Chapter 8, ENSO in a changing climate; Pontes et al. 2022—Nat.Geosci; Thirumalai & DiNezio et al. 2024—Nature; Molina et al. 2026—J Clim; such relevant papers need to be cited and discussed, especially in the context of the mechanism proposed in the manuscript). The current manuscript largely supports the first statement, but the second remains very underexplored and is a critical limitation of the analysis presented.
- A more compelling mechanistic demonstration would necessitate showing that the proposed heating asymmetry directly influences quantities involved in the time-transient and seasonally-phase-locked growth and decay of ENSO, such as Bjerknes feedback strength, surface heat-flux damping, mixed-layer depth, recharge/discharge behavior, westerly wind-burst statistics, or the seasonality of coupled instability (etc.) Recent ENSO literature has emphasized these specific process-level diagnostics, including the roles of upper-ocean stratification (Tuckman and Yang, 2026—arXiv), Walker circulation strength (Fedorov et al. 2020), warm-pool westward extent (Thirumalai, DiNezio et al. 2024—Nature), and ocean-atmosphere coupling (Okumura et al. 2010—J. Clim.) in modulating ENSO amplitude and the frequency of extreme El Niño events.
- The manuscript’s key point relies on Fig. 3, which shows a correlation between the east–west MSE “gap” and the Niño3 index. This is used as evidence that the zonal MSE contrast triggers a positive feedback, and essentially that this correlation outlines causation. During El Niño, the warm pool shifts eastward, increasing eastern-Pacific MSE and collapsing the MSE gap, making these two measures expressions of the same event. In order to showcase causality, what is the evolution across an interannual ENSO event acc. to the model and is there evidence of the gradient leading ENSO onset over time? What forcings affect such changes? Can this be isolated to mean state changes and then precessional changes? These questions remain unanswered.
- The proposed mechanism is not clearly separated from large-scale Indo-Pacific rainfall synchronizatoin.
- The manuscript suggests that continental heating alters convection and the warm pool, thereby amplifying ENSO. However, the analyses do not rule out a simpler explanation: both the MSE field and Niño3 variability could be responding to a larger-scale, precession-driven reorganization of rainfall in the Indo-Pacific, the ITCZ position, and the Walker circulation. In this scenario, MSE and ENSO variability would be related but not necessarily causally connected as proposed. To address this, the authors should demonstrate lead-lag relationships among MSE anomalies, convection, zonal wind stress, thermocline depth, and Niño3 SST on interannual timescales. If MSE anomalies are mechanistically important, they should precede or modulate the atmospheric wind response that initiates ENSO growth, rather than merely covary with mature ENSO SST anomalies. Composite analyses of the El Niño onset, development, peak, and decay phases would be far more convincing than correlations across broad spatial domains. For instance, composites of precipitation, MSE, surface winds, SST, and thermocline depth during El Niño onset would help test their proposed pathway in a more direct manner.
Minor Comments:
- It is unclear why Niño3 is the chosen index of focus, and not Niño3.4, as is standard for monitoring modern ENSO variability. The authors should show that this choice does not affect their conclusions.
- The authors cite Yang et al. 2024 for model “experimental settings”, but it would be helpful to include pCO2 and other greenhouse gas levels, alongside information related to obliquirt and other boundary conditions as well.
- The authors should specify the duration of each timeslice used for ENSO calculations, the number of independent ENSO events sampled, and include explicit uncertainty estimates for ENSO variance.
Citation: https://doi.org/10.5194/egusphere-2026-1183-RC2 -
AC3: 'Reply on RC2', Yufei Liu, 10 Aug 2026
We sincerely thank the reviewer for this thoughtful and constructive suggestion. Following the reviewer’s recommendation, we have systematically restructured the manuscript and incorporated a series of additional diagnostic analyses to provide a clearer and more comprehensive understanding of the underlying mechanisms.
Specifically, we have reorganized the overall framework of the manuscript into two closely connected stages. The first stage focuses on how precessional forcing, through asymmetric continental heating, reorganizes the tropical Indo–Pacific mean state. The second stage investigates how these mean-state changes regulate ENSO variability by modifying the efficiency of air–sea coupling feedback.
To strengthen the mechanistic interpretation, we have added a comprehensive set of diagnostics in the revised manuscript (Fig. 4), including precession-induced changes in key tropical Indo–Pacific mean-state characteristics, such as the Walker circulation (Fig. R1), surface wind stress (Fig. R2), Indo-Pacific warm pool position (Fig. R3), mixed-layer depth (Fig. R4), and upper-ocean stratification (Fig. R5). In addition, we have introduced several diagnostics directly related to air–sea coupling processes, including westerly wind burst intensity (Fig. R6), warm water volume variability in the western Pacific as an indicator of oceanic heat transport (Fig. R7), and the Bjerknes feedback index (Fig. R8).
Furthermore, to evaluate the robustness of our results, we performed additional ENSO event statistics, amplitude analyses, and sensitivity tests based on all 24 precession experiments using different Niño index definitions (Fig. 1, Fig. S3, and Fig. R19–R20). We also added lead–lag analyses comparing the 90° and 270° precession phases (Fig. R12), as well as composite analyses of El Niño events (Fig. R13–R18), to further clarify the temporal evolution and physical mechanisms linking mean-state changes to ENSO variability.
Overall, we sincerely appreciate the reviewer’s insightful and valuable comments. These revisions have substantially improved the organization, clarity, and scientific depth of the manuscript, allowing us to present the underlying physical mechanisms in a more coherent and comprehensive manner.
- AC5: 'Reply on RC2', Yufei Liu, 10 Aug 2026
-
EC1: 'Editor Comment on egusphere-2026-1183', Christo Buizert, 12 Jun 2026
Dear authors,
Your manuscript has now been seen by two reviewers. Both reviewers believe your work has merit. However, as you can see, particularly reviewer #2 has some substantial concerns about your work, and they do not think the central claims are supported by the analyses presented in the current version of the manuscript. Their concern rises above the level of a typical manuscript revision, and I anticipate you will need to make substantial changes in response.
As the next step in the review process, please provide detailed responses to all the points raised by both reviewers. Where appropriate, please indicate how you would address the comments in a revised manuscript. Again, providing detail is critical here: rather than a general statement like "we will address this issue in a revised manuscript", please indicate the exact changes and proposed revised text.
I look forward to reading your responses, and please do not hesitate to reach out in case of any questions.
All the best, Christo Buizert (CP editor)Citation: https://doi.org/10.5194/egusphere-2026-1183-EC1 -
AC1: 'Reply on EC1', Yufei Liu, 10 Aug 2026
We sincerely appreciate the reviewers’ insightful comments, which have helped us identify important limitations in the original manuscript and substantially improve the clarity, rigor, and mechanistic interpretation of our study. In response to the reviewers’ concerns, we have undertaken a substantial revision and reorganized the manuscript around a two-step physical framework: precessional forcing first reorganizes the tropical Indo-Pacific mean state, and this mean-state reorganization subsequently modifies the efficiency of coupled ocean–atmosphere feedback, thereby regulating ENSO variability.
Specifically, we have expanded our analyses to address the reviewers’ concerns regarding the characterization and mechanisms of ENSO variability. We quantified both ENSO frequency and amplitude across all 24 precession experiments (Fig. S3 and Figs. R19–R20 in the responses to Reviewer #2). We further added diagnostics of the Walker circulation, equatorial zonal wind stress, Indo-Pacific warm-pool position, mixed-layer depth, and upper-ocean stratification (Fig. S7, Fig. 4, and R1 in the response to Reviewer #1), together with analyses of westerly wind bursts, warm-water volume, the individual components of the Bjerknes feedback (Fig. 4 and Figs. R1–R11 in the response to Reviewer #2). In addition, we performed lead–lag analyses (Fig. R12 in the response to Reviewer #2) and composite analyses of representative 90° and 270° experiments (Figs. R13–R18 in the response to Reviewer #2) to further evaluate the temporal evolution of ENSO events.
Based on these additional analyses, we have revised the Methods, Results, Discussion, and Conclusions sections. Importantly, we removed the interpretation of MSE as an ENSO trigger to avoid overinterpretation and refined the mechanistic framework accordingly. The revised manuscript now emphasizes that precessional forcing induces asymmetric continental heating between Afro–Eurasia and Australia, reorganizes the tropical Indo-Pacific mean state, modifies the efficiency of coupled ocean–atmosphere feedback, and thereby regulates ENSO variability across the precession cycle. We believe that these revisions have substantially strengthened the manuscript and provide a more comprehensive and physically grounded explanation of the proposed mechanism.
Detailed responses to each reviewer comment, together with the corresponding figures, manuscript locations, and revised text, are provided in the accompanying responses to Reviewers #1 and #2.
-
AC1: 'Reply on EC1', Yufei Liu, 10 Aug 2026
Viewed
Since the preprint corresponding to this journal article was posted outside of Copernicus Publications, the preprint-related metrics are limited to HTML views.
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 280 | 0 | 3 | 283 | 0 | 0 |
- HTML: 280
- PDF: 0
- XML: 3
- Total: 283
- BibTeX: 0
- EndNote: 0
Viewed (geographical distribution)
Since the preprint corresponding to this journal article was posted outside of Copernicus Publications, the preprint-related metrics are limited to HTML views.
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Precession affects ENSO variability, which tends to be strong during periods of austral summer perihelion; however, the underlying mechanisms remain contentious. The authors propose that asymmetric continental heating between Afro–Eurasia and Australia at perihelion and aphelion drives the migration of the ITCZ, alters the Pacific east-west thermal contrast, and modulates the threshold for triggering ENSO events. For instance, summer perihelion induces strong warming over Australia (east of the Indo-Pacific ITCZ), shifting the ITCZ and the warm pool eastward, thereby facilitating the development of strong El Niño events. This mechanism bears resemblance to that of extreme El Niño under greenhouse warming, characterized by a reduced west-minus-east equatorial SST gradient.
Nevertheless, the proposed mechanism must be evaluated alongside other suggested processes. While these factors are not necessarily independent, they may be mutually inclusive, at least in part. Furthermore, the current results should be interpreted within the context of ongoing greenhouse warming, particularly regarding its impact on upper-ocean stratification (Cai et al., 2018, Nature). According to this study, we should be experiencing a high ENSO activity period, which aligns with observations. An attribution study (Cai et al., 2021, Nature Reviews Earth & Environment) indicates that ENSO amplitude has increased by over 30% since the 1960s, yet greenhouse warming accounts for only about 10% of this trend. This suggests that other processes likely play a significant role. The 10% increase is primarily attributed to enhanced upper-ocean stratification, despite the authors noting a La Niña-like mean state change. Under greenhouse warming, although rapid warming occurs over the Afro-Eurasian landmass, the effect of oceanic stratification appears to dominate.
Given these competing factors, I recommend that the authors investigate whether differences in upper-ocean stratification between periods of strong Australian warming and periods of strong Afro-Eurasian warming play a contributory role.