Reviewer Comments

Nonlinear Sensitivity of El Niño-Southern Oscillation across Climate States

By Pontes et al.

Pontes and coauthors present a compilation of climate model experiment results from past and future warm states (the mid-Holocene, the Last Interglacial, the mid-Pliocene, SSP5-8.5, and 4xCO₂) to investigate the mechanisms that drive El Niño-Southern Oscillation (ENSO) variability across changing background climate conditions. Taken together, the authors suggest the models support a common underlying mechanism influencing ENSO behavior: the latitudinal position of the Intertropical and South Pacific Convergence Zones (ITCZ and SPCZ, respectively). This response is tied to changes in the surface wind field and its associated impacts on the upper ocean, which can either enhance or weaken the ocean-atmosphere coupling strength that underpins ENSO development. Importantly, the authors find that this response to ITCZ and SPCZ position is non-linear. Peak ENSO suppression occurs when both convection centers are shifted equatorward by ~4 – 5° latitude. Further equatorward migrations, such as in the 4xCO₂ experiment, lead to permanent “El Niño-like” conditions that weaken ocean-atmosphere coupling and ENSO intensity. Likewise, poleward displacements by >8° latitude weaken the convective feedback, which also reduces ENSO intensity. The authors also find that this response is stronger for east Pacific ENSO than central Pacific ENSO. Altogether, the modeling results reveal negative quadratic relationships between convective center displacement and ENSO behavior, variability in easterly wind bursts, and wind-thermocline coupling, supporting a non-linear dynamic between ENSO and background climate.

Overall, I think the analysis presented in this paper is interesting and certainly a worthy contribution to the literature. I do have some major comments regarding the background climate conditions represented in the compilation and whether the authors’ proposed mechanism finds support in the paleo record. Otherwise, most of my suggestions are minor/editorial and could be triaged by the authors with ease.

Major Remarks

What immediately stood out to me was that the modeling experiments selected for the analysis exclusively represent warmer climates. Can the authors explain why they chose not to include model results from colder climate states (e.g., the PMIP LGM experiments)? I think the community’s constraints on glacial boundary conditions is particularly strong (perhaps better than, say, the mid-Pliocene) and there is an emerging consensus that ENSO was reduced at this time (e.g., Thirumalai et al., 2024 and references therein). The authors’ nonlinearity argument could be substantially bolstered by including the LGM simulation, which I expect would extend the right-hand “poleward displacement” side of the quadratic relationship in Figs. 2 and 3. But currently, without a colder state estimate, it’s hard to know if the nonlinearity presented here is a universal ENSO response or if it’s true only for warm states.

A side note to the comment above: I realized mid-way through this review that I might be mistaking what “equatorward” and “poleward” migrations are in reference to. Are they in reference to the equator, or the mean position of each convection center, or something else? It may be useful to quickly define what cardinal direction the authors are referring to in the discussion (e.g., “equatorward” might be northward for the SPCZ but southward for the ITCZ, according to Fig. 4). This is important because including glacial output might instead improve the left-hand “equatorward displacement” side of the quadratic, which is currently data sparse (as I note below).

Regarding the experiments included by the authors, the 4xCO₂ simulations does a lot of work in establishing the nonlinearity discussed in the paper. Indeed, three datapoints from this simulation anchor the left-hand “equatorward displacement” side of the quadratic function (Figs. 2b-d and 3b,c). Without these three points, I can see the remaining data forming more of a negative linear relationship, rather than a nonlinear one. Despite its importance, the 4xCO₂ simulation results receive little attention (e.g., it is not mentioned in section 3.1 and not presented in Fig. 1). I think these results should feature at least as prominently as the others, considering their importance for the nonlinearity argument.

Can the authors speculate on the applicability of their ITCZ response to millennial-scale climate change? I’m familiar with modeling results for the deglaciation (particularly cold stadial events) where a southward displacement of the ITCZ amplifies, rather than dampens, ENSO behavior (e.g., Liu et al., 2014; Timmerman et al., 2007). This is generally attributed to ENSO’s “frequency entrainment” to the annual cycle (Chang et al., 1994, 1995; Liu et al., 2002), which is weakened when the winds shift southward. Although this mechanism is still debated, it does find support in the paleo record, where studies have noted an increase in ENSO variability from the LGM to Heinrich Stadial 1 (Leduc et al., 2009; Glaubke et al., 2024; Sadekov et al., 2013) in response to a disruption of AMOC and southward shift of the ITCZ (e.g., Mosblech et al., 2012). Is there something unique about the meltwater-induced ITCZ migrations of the deglaciation that might represent a special case to the universal mechanism proposed here? Could it have something to do with the ITCZ and SPCZ moving in one direction together as opposed to moving closer or farther apart, as implied in Fig. 4?

Minor/Editorial Comments

(Note: [] represents a deletion; [words] represents added text.)

Abstract

Lines 19-21: The authors may want to consider a stronger and more specific problem statement. Perhaps something like: “However, [a common mechanism that can predict ENSO variability under a range of background conditions remains elusive.”]

Introduction

Throughout the introduction and later in the paper, the authors switch between using ka and Ma depending on the time period under discussion. For consistency, it might be worth sticking to one of these.

Line 62: Change “mid-Holocene” to “MH” here and throughout, as this is how you define it in the sentence prior.

Line 63: These are just two of many papers you can cite here. I would either add more for completeness (e.g., for the mid-Holocene, Conroy et al., 2008; Chen et al., 2016; White et al., 2018) or cite a review like Lu et al. (2018) and the references therein.

Data and Methods

Line 72: “…key [] periods in Earth’s history. Here, we analyze [three past] climate scenarios…”

Lines 72-74: You have already introduced and defined these climate periods in the introduction. You do not have to do this again here.

Lines 88-89: “… removing the monthly annual cycle.” Can you expand on this? What I think you’re saying is that you computed the mean annual cycle at a monthly resolution and subtracted that from each monthly value, but it’s not immediately clear.

Line 93: “… is projected approximately 45° between…” Is this 45° longitude?

Line 98: “To avoid being [misled]…”

Results and Discussion

Line 105: I would change “Results” to “Results and Discussion”

Section 3.1: I would add the 4xCO₂ results here. I would also suggest bringing in LGM simulations (as discussed above) and include them here as well.

Line 131: “…ITCZ and SPCZ…” As someone who works on the deglaciation, I think about ITCZ much more and have a better grasp of how it influences ENSO (i.e., changing the position of the trades). It might be helpful for those like me who don’t think about the SPCZ much to briefly mention how it relates to ENSO. Perhaps add it to the introduction? As of right now, the SPCZ isn’t mentioned until the results, so it would be nice to mention it sooner in the paper.

Lines 132-135: “Firstly…” Consider rephrasing. Perhaps “ENSO growth and extreme rainfall events are reduced the farther tropical convection centers are from the equator (Pontes et al., 2022). This occurs since the position of the convection centers determines the equatorial wind field and its associated upper ocean response.”

Line 141: “…[which underpins] ENSO development.”

Line 151: I would delete “preliminary”.

Line 152: Add (Fig. 2) to the end of the sentence.

Lines 155-158: Could the authors clarify what the equatorward and poleward shifts are with respect to? Is it the physical equator (0°)? The mean position of each convective center?

Line 163, 165, 173, etc: Consider italicizing parameters of an equation: “… (a = 0.XX)…”

Lines 182-183: Consider simplifying. “To investigate the [rainfall response to convection center migration], we…”

Conclusions

Line 229-230: Eliminate redundancy. “…linking meridional shifts of the [atmospheric] convection centers, [] ocean stratification, and zonal thermocline oscillations.”

Line 236: “…(<9°)…” Did the authors mean <8°, as mentioned in line 157?

Line 247-252: Is there room in an appendix of supplementary material to elaborate on these model biases and how they might influence the results? This could be helpful for non-modelers (such as myself) reading the paper. For example, I imagine the double-ITCZ problem is a relevant bias to dig into. If the SPCZ is defined as the region where precipitation is >50% the zonal average between 0° and 20°S (Line 530-531), then how might the excess precipitation south of the equator from the artificial southern ITCZ (mentioned in Line 540-541) influence that estimation?

Figures

Figure 1: Excellent figure. Of course, I think adding the 4xCO₂ results and some data from a glacial simulation would make it perfect.

Figure 2: Be sure to add what “equatorward” and “poleward” are in reference to. Consider also adding cardinal directions, if that helps.

Figure 3a: Consider using unique markers for each climate simulation like in Figs. 2b-d and 3b,c. It would be useful to see, for example, what proportion of the “strong equatorial shift” category of model results is from the 4xCO₂ experiment.

Figure 4: I like this figure! I was initially confused by how it was arranged, but once I was oriented, it clicked. Maybe make the X and Y arrows a bit larger to draw the eye?

References

1. Thirumalai, K. et al. Future increase in extreme El Niño supported by past glacial changes. Nature 1–7 (2024) doi:10.1038/s41586-024-07984-y.
2. Liu, Z. et al. Evolution and forcing mechanisms of El Niño over the past 21,000 years. Nature 515, 550–553 (2014).
3. Timmermann, A. et al. The Influence of a Weakening of the Atlantic Meridional Overturning Circulation on ENSO. J Climate 20, 4899–4919 (2007).
4. Chang, P., Wang, B., Li, T. & Ji, L. Interactions between the seasonal cycle and the Southern Oscillation - Frequency entrainment and chaos in a coupled ocean-atmosphere model. Geophys Res Lett 21, 2817–2820 (1994).
5. Chang, P., Ji, L., Wang, B. & Li, T. Interactions between the Seasonal Cycle and El Niño-Southern Oscillation in an Intermediate Coupled Ocean-Atmosphere Model. J Atmos Sci 52, 2353–2372 (1995).
6. Liu, Z. A Simple Model Study of ENSO Suppression by External Periodic Forcing*. J Climate 15, 1088–1098 (2002).
7. Leduc, G., Vidal, L., Tachikawa, K. & Bard, E. ITCZ rather than ENSO signature for abrupt climate changes across the tropical Pacific? Quaternary Res 72, 123–131 (2009).
8. Glaubke, R. H. et al. An Inconsistent ENSO Response to Northern Hemisphere Stadials Over the Last Deglaciation. Res. Lett. 51, (2024).
9. Sadekov, A. Y. et al. Palaeoclimate reconstructions reveal a strong link between El Niño-Southern Oscillation and Tropical Pacific mean state. Nat Commun 4, 2692 (2013).
10. Mosblech, N. A. S. et al. North Atlantic forcing of Amazonian precipitation during the last ice age. Nat Geosci 5, 817–820 (2012).
11. Conroy, J. L., Overpeck, J. T., Cole, J. E., Shanahan, T. M. & Steinitz-Kannan, M. Holocene changes in eastern tropical Pacific climate inferred from a Galápagos lake sediment record. Quaternary Sci Rev 27, 1166–1180 (2008).
12. Chen, S. et al. A high-resolution speleothem record of western equatorial Pacific rainfall: Implications for Holocene ENSO evolution. Earth Planet Sc Lett 442, 61–71 (2016).
13. White, S. M., Ravelo, A. C. & Polissar, P. J. Dampened El Niño in the Early and Mid-Holocene Due To Insolation-Forced Warming/Deepening of the Thermocline. Geophys Res Lett 45, 316–326 (2018).
14. Lu, Z., Liu, Z., Zhu, J. & Cobb, K. M. A Review of Paleo El Niño-Southern Oscillation. Atmosphere-basel 9, 130 (2018).

In this manuscript, the authors aim to clarify the relationship between properties of ENSO variability and the mean background climate in a range of different past and future warmer climate states. The authors employ different past climate simulations (covering the mid-Pliocene, last interglacial and mid-Holocene) and future simulations (ssp585 and abrupt 4x CO2) and present their results with respect to pre-industrial climate simulations. The results show a common mechanism linking changes to the mean atmosphere-ocean circulation to an increase or decrease in ENSO variability. In fact, the authors reveal a non-linear relationship between a change in ENSO variability and the change in the meridional positions of the intertropical and South Pacific convergence zones. Three distinct regimes of ENSO variability change are identified and the processes in the tropical Pacific mean climate leading to the existence of these regimes are explained.

I believe this is a very interesting study with relevant findings. The scientific content is substantial and provides a significant contribution to the field of ENSO complexity. The methods are sound, and the figures are very clear. I do have a few issues, related to the data selection and to the Introduction, that I would consider major (but should not require too much effort to resolve). Apart from those, I present minor comments and a few technical corrections. I recommend this manuscript to be published in Climate of the Past subject to revisions.

My comments are included in the attached PDF.

This study combines state-of-the-art model simulations of past climates and future warming to evaluate ENSO activity throughout a wide range of climate states. The authors find that the sensitivity of ENSO to the background climate is nonlinear and tied to the climatological position of the tropical Pacific convection centers, the ITCZ and SPCZ. Results of this study provide a comprehensive mechanism of how tropical Pacific mean state modulates ENSO activity.

I find this is a novel and interesting study and the analysis is basically sound. Nevertheless, as previously mentioned by the other two reviewers, the writing can be further refined so that the paper will be more readable.

Please see my specific comments below.

1. As seen in the dispersion diagrams and discussed in the manuscript, abrupt 4xCO2 simulation results show strong equatorward shift of convection centers, which is distinct from the past climates. I was wondering how the location of convection centers and ENSO-convection centers relationship change in other less aggressive CO2 emission scenarios. It will be helpful to show these results since “across climate states” is mentioned in the title of this study.

1. It’s interesting to see that the left branch of the quadratic regression curve is due to mid-Pliocene scatters in Fig. 3c. In contrast, there are scatters mainly from the abrupt 4xCO2 simulations at the left branch of the quadratic regression curve in Fig. 3b. Could you provide some explanation/discussion on the reason of this? In addition, why there are not many abrupt 4xCO2 scatters in Fig. 3c?

1. Fig. 1: It would be helpful if the color bars in Fig. 1a-1d could be revised to make positive values in warm colors (yellow and red) and negative values in cold colors (blue).

Technical corrections:

Line 30: Central El Niño events -> Central Pacific El Niño events

Line 58: ~129 116 thousand years ago -> ~129-116 thousand years ago

Line 77: CO2to -> CO2 to

Line 167: Figure 2a -> Figure 2b

Line 176: Figure 2b -> Figure 2d

Line 477: boreal spring-summer -> austral spring-summer