the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 18 Mar 2026
Dominance of Obliquity over Precession in Polar Temperature Variability: Insights from an Energy Balance Model
Abstract. The sensitivity of a Zonally Averaged Energy and Moisture BAlance Climate Model (ZEMBA) to changes in the Earth’s orbit is investigated. The model is intended to explore the dynamics of Quaternary glacial-interglacial cycles, particularly the dominance of 41-kyr obliquity cycles in ice volume and ocean temperature during the Early Pleistocene, despite summer insolation being primarily influenced by 19- and 23-kyr precession cycles. Through equilibrium simulations for the Pre-Industrial and Last Interglacial Period, we demonstrate that ZEMBA's response to strong orbital forcing qualitatively matches the behavior of climate models from the Coupled Model Intercomparison Project Phase 6 (CMIP6). Transient simulations of ZEMBA over the Early Pleistocene reveal a pronounced 41-kyr cyclicity in surface temperatures at the polar latitudes, in correspondence to variations in the Earth’s obliquity. Sensitivity experiments underscore the essential role of sea ice in driving temperature variability in the polar regions. The dominant 41-kyr cyclicity in surface air temperature is attributed to obliquity’s influence on winter sea ice extent, which governs the release of substantial ocean heat to the atmosphere. The more subdued effect of precession on surface air temperature is linked to the counterbalancing relationship between insolation intensity and summertime duration, which constrains variability in both winter sea ice and ocean heat fluxes.
- Preprint
(9961 KB) - Metadata XML
-
Supplement
(6633 KB) - BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on egusphere-2026-1151', Anonymous Referee #1, 29 Apr 2026
-
AC1: 'Reply on RC1', Daniel Gunning, 17 Aug 2026
We would like to thank RC1 for their comments on the manuscript, which has allowed us to improve the robustness of our results and address issues with the model analysis. Following comments from all RCs, we are in the process of re-running these experiments for a lower CO2 level (in keeping with the Early Pleistocene) and at a higher resolution, prior to revising the manuscript with these updated results. Here, we present our responses, with RC1’s comments highlighted in bold, and our replies in normal font weight.
Review of Gunning et al. submitted to Climate of the Past: “Dominance of Obliquity over Precession in Polar Temperature Variability: Insights from an Energy Balance Model”
In this manuscript, the authors investigate the climate and atmospheric response to the 41-kyr glacial cycles in the Early Pleistocene using their ZEMBA model (Gunning et al., 2025). ZEMBA is a one-dimensional energy-balance climate model that considers atmosphere, land, and ocean at each latitudinal grid point, as well as hydrological cycles. The performance of the model was evaluated in their previous study (Gunning et al., 2025) for the preindustrial and the Last Glacial Maximum (LGM) periods. In this study, they first present the model's performance in reproducing the Last Interglacial (LIG) period, evaluating its performance under different astronomical forcing. I like the concept of ZEMBA, which is a useful tool for understanding long-term climate change. However, the present manuscript lacks context on the 41-kyr problem, and the absence of ice-sheet modeling limits the model's applicability. My comments are summarized below, with capital Ls representing the line number.
Gunning, D. F., Nisancioglu, K. H., Capron, E., & van de Wal, R. S. (2025). ZEMBA v1. 0: an energy and moisture balance climate model to investigate Quaternary climate. Geoscientific Model Development, 18(9), 2479-2508.
Major Comments
The authors elegantly summarize the so-called 41-kyr problem in the Introduction. They clearly explained two standing hypotheses (Huybers, 2006; Raymo et al., 2006) and their issues in accounting for the lack of precession periodicity in δ18O; however, they did not explain how the ZEMBA investigation can help resolve the 41-kyr problem, which makes the contribution of the present study very unclear. Explaining the 41-kyr problem ultimately requires ice-sheet modeling, which is not included in the current ZEMBA model, thereby limiting its contribution to the 41-kyr problem. When I first read through the manuscript, I thought it extended the hypothesis presented by Raymo and Nisancioglu (2003), who suggested that the latitudinal insolation gradient, dominated by 41-kyr cyclicity, affects heat and moisture fluxes to high-latitude regions and hence the 41-kyr periodicity in ice-sheet mass. In this sense, I think this study contributes to the 41-kyr problem even without ice-sheet modeling, but the linkage between what this study did and the 41-kyr problem is very unclear in the present manuscript. I recommend clearly situating this study within the context of the 41-kyr problem in the Introduction and enriching the Discussion.
The original intention of the manuscript was to isolate a feedback between obliquity and winter sea ice that heightens the sensitivity of high-latitude temperature to obliquity cycles, which we proposed as a potential contributing factor to the dominant 41-kyr cycles of the Early Pleistocene. More broadly, following the development of ZEMBA as a tool for studying long-term climate change under orbital forcing (Gunning et al. 2025), we sought to document how the model's fast climate feedbacks, associated with sea-ice and snow-cover change, respond to the obliquity and precession cycles, and whether these feedbacks inform on the '41-kyr world' despite the dominant precession forcing of summer insolation.
We agree the introduction would benefit from a more explicit linkage between the ’41-kyr problem’ and the orbital sensitivity experiments performed in ZEMBA. To address this, we will revise the introduction to (1) ground the study explicitly in the context of the orbital cycles of obliquity, precession, and the '41-kyr problem; (2) motivate the use ZEMBA, without an interactive ice sheet, as a means of isolating how sea ice and snow cover respond to obliquity and precession forcing separately; and (3) state clearly what this configuration can and cannot address, in terms of missing feedbacks from vegetation, clouds, and ice-sheet dynamics. The revised discussion will expand the comparison with related orbital-sensitivity studies (Tabor et al. 2014; Erb et al. 2014).
As the RCs have brought to our attention that the study of Tabor et al. (2014) documents a similar obliquity-sea ice surface feedback, we acknowledge that the scope and emphasis of the manuscript should be adjusted, which we address in detail in response to the second major comment below.
Huybers, P. (2006). Early Pleistocene glacial cycles and the integrated summer insolation forcing. Science, 313(5786), 508-511.
Raymo, M. E., & Nisancioglu, K. H. (2003). The 41 kyr world: Milankovitch's other unsolved mystery. Paleoceanography, 18(1).
Raymo, M. E., Lisiecki, L. E., & Nisancioglu, K. H. (2006). Plio-Pleistocene ice volume, Antarctic climate, and the global δ18O record. Science, 313(5786), 492-495.
As I mentioned in the comment above, the current ZEMBA model does not couple to the ice-sheet model. The authors discuss temperature responses in high-latitude regions, particularly the dominance of the 41-kyr periodicity, accounting for the effects of variations in ice-sheet mass. However, the formation of ice sheets affects the climate via many feedbacks, including changes in surface albedo, altitude, and atmospheric circulation. Without ice-sheet variations, it is difficult to assess the robustness of their results, especially their discussion of snow cover and sea-ice distributions. I think ice-sheet modeling is highly recommended to discuss the 41-kyr problem in this study, as was done using both a simplified model (Huybers and Tziperman, 2008) and more sophisticated models (Tabor et al., 2015; Tabor and Poulsen, 2016; Willeit et al., 2019; Watanabe et al., 2023). Even if it is difficult, careful justification, mention of the robustness of their results, and the uncertainty associated with the lack of feedback, especially the ice-sheet variability, and the potential impact of the snow cover change and heat-moisture budgets discussed in this study on ice-sheet mass balance are required.
Huybers, P., & Tziperman, E. (2008). Integrated summer insolation forcing and 40,000‐year glacial cycles: The perspective from an ice‐sheet/energy‐balance model. Paleoceanography, 23(1).
Tabor, C. R., Poulsen, C. J., & Pollard, D. (2015). How obliquity cycles powered early Pleistocene global ice‐volume variability. Geophysical Research Letters, 42(6), 1871-1879.
Tabor, C. R., & Poulsen, C. J. (2016). Simulating the mid-Pleistocene transition through regolith removal. Earth and Planetary Science Letters, 434, 231-240.
Watanabe, Y., Abe-Ouchi, A., Saito, F., Kino, K., O’ishi, R., Ito, T., ... & Chan, W. L. (2023). Astronomical forcing shaped the timing of early Pleistocene glacial cycles. Communications Earth & Environment, 4(1), 113.
Willeit, M., Ganopolski, A., Calov, R., & Brovkin, V. (2019). Mid-Pleistocene transition in glacial cycles explained by declining CO2 and regolith removal. Science Advances, 5(4), eaav7337.
As noted above, the original intention for the manuscript was to outline a mechanism for an obliquity amplification of high latitude temperature related to sea ice, prior to ice-coupling. The RCs have made us aware that similar findings were documented by Tabor et al. (2014). While we are encouraged that ZEMBA reproduces these surface feedbacks captured in a full general circulation model, we agree the manuscript would benefit from a shift in emphasis: confirming these findings with a simplified, independent model whose efficiency allows for targeted sensitivity experiments which are impractical for a full GCM. Thereafter, we isolate the individual roles of sea ice and snow cover on high-latitude temperature, to further underscore the importance of this sea-ice mechanism. In the revised manuscript, following suggestions from RC2, we will provide further sensitivity analysis on these fixed-sea ice and snow cover runs in response the orbital cycles of obliquity and precession, which would provide more detail on how obliquity influences sea ice and temperature relative to precession. In addition, the robustness of the obliquity sea ice feedback will be shown at two different CO2 levels (pre-industrial and Early Pleistocene levels)
For the present manuscript, we agree that using the high-latitude temperature response alone as a marker of ice-sheet mass balance is not sufficient to address the ’41-kyr’ problem. The obliquity control on (winter) sea ice only serves to heighten the sensitivity of surface air temperature to this orbital parameter relative to precession, despite a much-reduced summer insolation forcing (as in Tabor et al. 2014).
The revised manuscript will dedicate a section specifically to the analysis of metrics in ZEMBA more relevant for ice sheet mass-balance, which includes but is not limited to surface temperature. Specifically, these include summer average land temperature, surface melt, perennial snow cover, snowfall, and the net surface mass balance. Surface mass balance, in particular, may address some of the concerns, as it incorporates the effects of many feedbacks, including temperature, albedo, moisture transport and snowfall over land. Initial results from our higher resolution experiments show both a strong obliquity and precession influence on these variables. While the sea ice mechanism is important for heightening the sensitivity to the obliquity cycle, there is nonetheless a strong precession component not in keeping with the 41-kyr glacial cycle. This points to the importance of these other feedbacks, with the obliquity-sea ice feedback potentially helping dictate the timing of glacier termination every ~41-kyr in the Early Pleistocene.
In fact, we have performed simulations of ZEMBA coupled to a simplified NH ice-sheet model (as in Huybers and Tziperman, 2008), which we are ready to submit in a subsequent publication. For constant CO2 levels of 220 ppm, there is 41-kyr cycles of ice volume, ice sheet surface mass balance and high-latitude temperature. Given mass balance related variables in the ZEMBA only experiments show a strong obliquity and precession component, we interpret the dominant 41-kyr cyclicity in ice volume slightly differently. Ice-sheet related feedbacks for growth, including elevation-temperature, likely overwhelm these faster surface feedbacks and fuel ice sheet growth. The 41-kyr cyclicity is instead driven by a threshold response for termination which occurs when the summer insolation forcing from obliquity and precession align, which occurs approximately every second precession cycle when obliquity is above-average, and thus every ~41-kyr (as in Watanabe et al. 2025). Ice-coupled simulations with fixed sea ice cover still show the importance of the sea ice mechanism for the triggering of this threshold response, but it affects the periodicity of climate change in the high northern latitudes differently from the faster feedbacks in the ZEMBA only runs. We have chosen to keep the coupling to an interactive ice sheet as a separate paper - as it is a comprehensive study which would overwhelm the focus of the current manuscript.
For the current manuscript, we propose that together with our findings related to the obliquity-sea ice mechanism, which causes a dominant 41-kyr cycle in annual-mean temperature and a strong amplification of summer-mean temperature, a closer inspection of surface mass balance, perennial snow cover, snowfall and surface melt, together with the fixed sea ice and snow cover experiments, provide a robust analysis of how obliquity and precession cycles influence these fast climate feedbacks. The discussion will be more explicit in the feedbacks lacking in the simulations that are very important for glacial-interglacial periodicity.
Tabor, C. R., Poulsen, C. J., and Pollard, D.: Mending Milankovitch's theory: obliquity amplification by surface feedbacks, Clim. Past, 10, 41–50, https://doi.org/10.5194/cp-10-41-2014, 2014.
Specific Comments
L129–135: According to Table 1, it seems that the authors conducted transient simulations from 2.45 to 1.2 Ma using a constant atmospheric CO2 level at the preindustrial value (284 ppm). This value is too high for the mean value representing the early Pleistocene. This should be explained and justified in the main text of the Methods section. The potential impact of the atmospheric CO2 level may be discussed in the latter part.
We agree that running the transient simulations at a pre-industrial CO2 level of 284 ppm is not representative of the Early Pleistocene, and we have re-run the experiments for a constant atmospheric CO2 level of 235 ppm, more in keeping with Early Pleistocene levels measured from the Beyond EPICA ice core (Krauss et al. 2026). The results at 235 ppm are broadly consistent with those presented in the original manuscript, with high northern latitude surface air temperature continuing to show a dominant 41-kyr (obliquity) cycle. Together with runs at pre-industrial and a lower CO2 level, we fully assess the robustness of the result to the greenhouse gas forcing.
However, at this lower CO2 level, we identified an artefact arising from the model's coarse (5°) latitudinal resolution: lower-latitude bands with large surface areas sit close to a sea-ice threshold. As a result, small orbitally-paced changes in insolation can switch entire latitudinal bands between sea ice-covered and sea ice-free conditions, producing discrete step changes in temperature in the time series and violin plots (e.g., Figure 5) rather than a smooth response. This behaviour was not apparent under the warmer 284 ppm forcing as the model’s mean climate was not sufficiently close to this threshold. This limitation complicates the sensitivity analyses comparing the influence of obliquity and precession. To rectify this, we have optimised the model code to run substantially faster at finer latitudinal resolutions. We are currently generating the full set of transient simulations at this finer (1°) resolution and will incorporate the updated results in the revised manuscript.
Krauss, F., Schmitt, J., Bauska, T., Capron, E., Grilli, R., Heiserer, R., Silva, L., Stocker, T., Fischer, H., and Beyond EPICA community, T. E.: The first coarse-resolution Beyond EPICA CO2 record covering the Mid-Pleistocene Transition: Insights from long-term carbon-cycle dynamics, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-13891, https://doi.org/10.5194/egusphere-egu26-13891, 2026.
L144–146: The peak month of the isolation anomaly will differ if the authors apply the calendar-effect adjustment (Bartlein and Shafer, 2019). I recommend that the authors apply the calendar-effect adjustment, which will allow readers to better understand and compare the results with those of previous studies, such as Otto-Bliesner et al. (2021).
Bartlein, P. J., & Shafer, S. L. (2019). Paleo calendar-effect adjustments in time-slice and transient climate-model simulations (PaleoCalAdjust v1. 0): Impact and strategies for data analysis. Geoscientific Model Development, 12(9), 3889-3913.
Otto-Bliesner, B. L., Brady, E. C., Zhao, A., Brierley, C. M., Axford, Y., Capron, E., ... & Zheng, W. (2021). Large-scale features of Last Interglacial climate: results from evaluating the lig127k simulations for the Coupled Model Intercomparison Project (CMIP6)–Paleoclimate Modeling Intercomparison Project (PMIP4). Climate of the Past, 17(1), 63-94.
We thank the reviewer for pointing out the oversight of not applying the calendar-effect adjustment (Bartlein and Shafer, 2019) in our manuscript. We have now converted daily fields into monthly averages, avoiding the pitfall of comparing daily fields that correspond to different solar angles. This involved converting the Fortran programmes from https://github.com/pjbartlein/PaleoCalAdjust into Python equivalents. In the revised manuscript, all contour plots and seasonal time-series and spectral amplitudes have been replaced, moving from daily fixed-calendar fields to paleo-corrected monthly fields (e.g., winter sea ice, ocean heat fluxes).
Indeed, we find that including the paleo calendar-effect adjustment is important for the analysis. Our original contour plots using daily fixed-calendar fields (e.g., Figure 6) overemphasized peak insolation and surface warming in early summer and underemphasized it in late summer, relative to the updated version using paleo-calendar-adjusted monthly fields (Fig. R1). Nevertheless, we still find that obliquity generates stronger year-round warming at high northern latitudes relative to precession, with peak winter warming coinciding with positive ocean heat flux anomalies.
L187–193:
Are the changes of atmospheric and oceanic heat transport rates associated with different astronomical forcing compatible with previous results obtained using a general circulation model (e.g., Mantsis et al., 2011; 2014; Erb et al., 2013; Tabor et al., 2014)?
Erb, M. P., Broccoli, A. J., & Clement, A. C. (2013). The contribution of radiative feedbacks to orbitally driven climate change. Journal of climate, 26(16), 5897-5914.
Mantsis, D. F., Clement, A. C., Broccoli, A. J., & Erb, M. P. (2011). Climate feedbacks in response to changes in obliquity. Journal of Climate, 24(11), 2830-2845.
Mantsis, D. F., Lintner, B. R., Broccoli, A. J., Erb, M. P., Clement, A. C., & Park, H. S. (2014). The response of large-scale circulation to obliquity-induced changes in meridional heating gradients. Journal of Climate, 27(14), 5504-5516.
Tabor, C. R., Poulsen, C. J., & Pollard, D. (2014). Mending Milankovitch's theory: obliquity amplification by surface feedbacks. Climate of the Past, 10(1), 41-50.
Atmospheric and ocean heat transport closely follow the 41-kyr obliquity cycle in our standard (TBE) experiment, varying by less than 0.1 PW. The obliquity- and precession- only experiments also show that obliquity generates larger variations in atmospheric heat transport related to larger gradients in MSE between the equator and the pole, but this is still less than <3% of total atmospheric heat transport. Using the GFDL CM2.1 GCM, Mantsis et al. (2014) report a 9–26% increase in mid-latitude atmospheric heat transport in response to a switch from high to low obliquity; no equivalent precession experiment exists in that study. Erb et al. (2013), using the same model, compared the climate response to extremes in both obliquity and precession, but did not report changes in the strength of atmospheric circulation for either orbital parameter. However, Erb et al. (2013) suggest that the vertical structure of atmospheric heat transport- not resolved in ZEMBA- can affect radiative feedbacks for extremes in obliquity. For a decrease in obliquity, positive MSE anomalies in the tropics are transported poleward in the upper troposphere, while negative MSE anomalies are transported equatorward at the surface, which serve to decrease vertical lapse rates, increase atmospheric stability, and increase cloud cover. Finally, Tabor et al. (2014) find little difference in NH high-latitude eddy heat and moisture flux between their obliquity and precession forcing experiments and conclude that transport does not significantly contribute to the greater climate sensitivity to obliquity relative to precession. Given the introduction mentions theories for the 41-kyr world linked to atmospheric heat and moisture fluxes, the seemingly small influence of these orbital parameters on heat transport will be clarified in the results section for the fixed orbital experiments.
L199: There should be a period at the end of the sentence.
This will be corrected in the revised manuscript.
Figure 5: Consider using a color set that is accessible for people with color vision deficiencies. Same for Figure S4.
We will update the colour scheme to one that is more accessible.
L317–326: Same as my comment for L187–193.
See above.
Citation: https://doi.org/10.5194/egusphere-2026-1151-AC1
-
AC1: 'Reply on RC1', Daniel Gunning, 17 Aug 2026
-
RC2: 'Comment on egusphere-2026-1151', Anonymous Referee #2, 07 May 2026
Review of Gunning et al.
Gunning et al. utilize the zonally averaged energy balance model (ZEMBA) to evaluate the impact of orbital forcing on surface air temperature (SAT) and snow/ice during the Early Pleistocene. The authors first validate ZEMBA’s performance by comparing modern and Last Interglacial (LIG) simulations with PMIP models, demonstrating that ZEMBA yields comparable SAT and sea-ice changes. They subsequently perform transient and sensitivity experiments for the Early Pleistocene. Their results indicate that changes in obliquity strongly influence annual mean polar temperatures by affecting winter sea ice and ocean-atmosphere heat exchange, leading to a climate response synchronized with the 41-kyr obliquity cycle. Regarding climatic precession, the authors show a strong impact on summer snow/ice, however the net effect on annual mean temperature remains limited due to the seasonal compensation of insolation anomalies between winter and summer.
I find the experiments interesting, the manuscript is clearly written, and the topic is well-suited for the readership of Climate of the Past. Based on the points outlined below, however, I recommend a moderate-major revision.
Major Points
- Several studies have pointed out the seasonal impacts of obliquity and precession on temperature (Erb et al. 2013) and ocean heat flux (Tabor et al. 2014). The authors should discuss their findings in the context of existing literature. That said, the sensitivity experiments conducted by fixing sea ice and snow cover are novel. I suggest including further analysis (somewhat similar to Fig. 6) on these sensitivity runs to discuss how snow and sea ice and their interactions influence the seasonality of temperature anomalies. This would effectively differentiate this work from previous studies.
- A large portion of the Discussion section currently repeats the Results and could be streamlined. Instead, I suggest the authors expand the discussion on factors not currently accounted for in the model, such as vegetation feedbacks (e.g., Tabor et al. 2014; O’ishi et al. 2021), cloud feedbacks (e.g., Erb et al. 2013; Arima et al. 2026) and the potential impact of CO2 level on the results.
Specific Comments
L57: Please add a reference here.
L103: The spatial extent of the mixed layer depth seems unusual. Please provide a justification for why this specific location was set.
L165-167: What is the reasoning here? Is it because the sea-ice albedo is set to 0.7?
L168-169: I’m asking out of curiosity, but if the model was re-tuned to allow less Arctic sea ice during the LIG (127ka) by lowering the sea ice albedo, would this result in a model where precession exerts a stronger influence?
L181: Is this a typo for 71%?
L185: Please refer to the specific Figure being discussed to improve readability.
Section 4.4: There appears to be some confusion between Fig. 8 and Fig. 5. Please clarify.
L282-305: While informative, this paragraph does not seem directly relevant to the results and could be removed. If retained, please provide a concise explanation of how ZEMBA differs from other EBMs to justify the differing results.
L306-317: This is a repetition of the Methods and Results sections and should be removed.
L329: Should the large thermal inertia of the ocean also be highlighted as a key factor here?
L355-373: These lines repeat the Results section and should be condensed.
Figures
Fig. 6: Please include an explanation of the horizontal dashed lines (mx and mn) in the caption.
Fig. 7: Please specify in the caption which experiment is being analyzed.
Reference
Arima, N., Yoshimori, M., Abe-Ouchi, A., O'ishi, R., Chan, W.-L., Sherriff-Tadano, S., and Ogura, T.: Impact of the temperature-cloud phase relationship on the simulated Arctic warming during the Last Interglacial, Clim. Past, 22, 891–913, https://doi.org/10.5194/cp-22-891-2026, 2026.
Erb, M. P., Broccoli, A. J., and Clement, A. C.: The Contribution of Radiative Feedbacks to Orbitally Driven Climate Change, J. Climate, 26, 5897–5914, https://doi.org/10.1175/JCLI-D-12-00419.1, 2013.
O'ishi, R., Chan, W.-L., Abe-Ouchi, A., Sherriff-Tadano, S., Ohgaito, R., and Yoshimori, M.: PMIP4/CMIP6 last interglacial simulations using three different versions of MIROC: importance of vegetation, Clim. Past, 17, 21–36, https://doi.org/10.5194/cp-17-21-2021, 2021.
Tabor, C. R., Poulsen, C. J., and Pollard, D.: Mending Milankovitch's theory: obliquity amplification by surface feedbacks, Clim. Past, 10, 41–50, https://doi.org/10.5194/cp-10-41-2014, 2014.
Citation: https://doi.org/10.5194/egusphere-2026-1151-RC2 -
AC2: 'Reply on RC2', Daniel Gunning, 17 Aug 2026
We would like to thank RC2 for their comments on the manuscript, which has allowed us to improve the robustness of our results and address issues with the model analysis. Following comments from all RCs, we are re-running the experiments with a lower CO2 level (in keeping with the Early Pleistocene) and at a higher resolution. These updated experiments will be included in the revised manuscript. Here, we present our responses, with RC2’s comments highlighted in bold, and our replies in normal font weight.
Review of Gunning et al.
Gunning et al. utilize the zonally averaged energy balance model (ZEMBA) to evaluate the impact of orbital forcing on surface air temperature (SAT) and snow/ice during the Early Pleistocene. The authors first validate ZEMBA’s performance by comparing modern and Last Interglacial (LIG) simulations with PMIP models, demonstrating that ZEMBA yields comparable SAT and sea-ice changes. They subsequently perform transient and sensitivity experiments for the Early Pleistocene. Their results indicate that changes in obliquity strongly influence annual mean polar temperatures by affecting winter sea ice and ocean-atmosphere heat exchange, leading to a climate response synchronized with the 41-kyr obliquity cycle. Regarding climatic precession, the authors show a strong impact on summer snow/ice, however the net effect on annual mean temperature remains limited due to the seasonal compensation of insolation anomalies between winter and summer.
I find the experiments interesting, the manuscript is clearly written, and the topic is well-suited for the readership of Climate of the Past. Based on the points outlined below, however, I recommend a moderate-major revision.
Major Points
- Several studies have pointed out the seasonal impacts of obliquity and precession on temperature (Erb et al. 2013) and ocean heat flux (Tabor et al. 2014). The authors should discuss their findings in the context of existing literature. That said, the sensitivity experiments conducted by fixing sea ice and snow cover are novel. I suggest including further analysis (somewhat similar to Fig. 6) on these sensitivity runs to discuss how snow and sea ice and their interactions influence the seasonality of temperature anomalies. This would effectively differentiate this work from previous studies.
We thank the reviewer for raising our attention to these studies. The discussion of the revised manuscript will be supplemented with comparison to these orbital sensitivity experiments using more complex climate models (Tabor et al. 2014; Erb et al. 2014). A major focus of the manuscript was placed on the obliquity-sea ice surface feedback, which we were unaware had been reported on by Tabor et al. (2014). As stated in response to RC1, we are encouraged that ZEMBA reproduces these surface feedbacks captured in a full general circulation model and agree the manuscript would benefit from a shift in emphasis: (1) confirming these findings with simplified and independent model; and (2) building on this via targeted sensitivity experiments that the computational efficiency of ZEMBA allows.
By isolating the individual impact of sea ice and snow cover on high-latitude temperature, we further underscore the importance of sea-ice feedbacks. We agree that further analysis on the individual sensitivity of sea ice and snow cover to the orbital cycles of obliquity and precession is warranted, which would further elucidate the role of sea ice in amplifying the obliquity cycle versus precession. This will require four additional simulations: the fixed precession (TFX_PR) and fixed obliquity (TFX_OB) will have to be repeated for fixed sea ice cover and fixed snow-cover. The revised manuscript will include these experiments in a figure similar to Fig. 6 as suggested.
- A large portion of the Discussion section currently repeats the Results and could be streamlined. Instead, I suggest the authors expand the discussion on factors not currently accounted for in the model, such as vegetation feedbacks (e.g., Tabor et al. 2014; O’ishi et al. 2021), cloud feedbacks (e.g., Erb et al. 2013; Arima et al. 2026) and the potential impact of CO2 level on the results.
We agree that the discussion could benefit from a commentary on missing feedbacks including those from clouds and vegetation. This is especially true considering that while ZEMBA agrees reasonably well with the lig127k ensemble, it still underestimates the warming inferred from paleoclimate datasets, pointing to the importance of these missing feedbacks. We discuss these points regarding vegetation, clouds and CO2 below, that will be integrated in the revised discussion section.
On vegetation:
In the current configuration of ZEMBA, albedo over land is set to a uniform value of 0.15, independent of vegetation type. We therefore neglect vegetation-related feedbacks and consider only snow cover and sea ice feedbacks. We acknowledge, however, that vegetation feedbacks play an important role in amplifying the response to orbital forcing. O’ishi et al. (2021) show that the inclusion of dynamic vegetation feedbacks for LIG insolation anomalies- namely the replacement of tundra with boreal forests- is important for generating an annual-mean warming in the high northern latitudes. Reductions in albedo over land lead to additional energy input that is transported to the Arctic Ocean to reduce sea-ice cover, resulting in increased ocean heat fluxes in the autumn (O’ishi et al. 2021). Therefore, the inclusion of vegetation feedbacks in ZEMBA would be expected to increase the strength of sea ice and snow cover feedbacks in the orbital sensitivity runs. Furthermore, the inclusion of vegetation feedbacks could alter the relative influence of obliquity and precession on surface air temperature and other climate variables. Using an Earth System Model coupled to a vegetation model, Tabor et al. (2014) found that the stronger influence of obliquity on annual-mean temperature (relative to precession) drives a greater exchange between tundra and boreal forest in the high northern latitudes, producing larger changes in annual-mean albedo that amplifies the initial temperature response to obliquity. We might therefore expect the inclusion of vegetation feedbacks in ZEMBA to further amplify the dominance of obliquity over precession in high-latitude, annual-mean temperature. It should be noted that other studies report a greater sensitivity of vegetation to precession (Crucifix and Loutre, 2002), and the relative sensitivity of vegetation to each orbital parameter may depend on the vegetation model used. Nevertheless, given the strong influence of obliquity on annual-mean temperature, we might expect vegetation feedbacks to amplify the response to obliquity over precession. We are unaware of any zonally averaged climate models that have implemented vegetation feedbacks, but this would be an interesting addition to a future iteration of ZEMBA.
On clouds:
A limitation of ZEMBA is the prescription of a single cloud-cover fraction corresponding to a pre-industrial simulation from NorESM2. Sensitivity experiments with different zonal-mean cloud cover distributions demonstrate its strong impact on ZEMBA’s pre-industrial surface temperature (Gunning et al. 2025). We have not included interactive cloud feedbacks in ZEMBA, given the risk that a highly uncertain parameterization would detract from feedbacks that model intends to isolate. Nonetheless, cloud feedbacks may interact with the sea-ice mechanism discussed here. Arima et al. (2026) show, in an atmosphere-ocean GCM, that reduced Arctic sea ice at the LIG is associated with an increase in low-level cloud cover in autumn, enhancing the downward longwave cloud radiative effect and contributing to winter warming. This generates positive feedback between sea ice retreat and low-level cloud that can further reduce summer sea ice and enhance warming, which is sensitive to the cloud-phase representation (Arima et al. 2026). Sea ice retreat, which is more strongly influenced by obliquity in our experiments, could therefore conceivably be strengthened by such cloud feedbacks. When performing transient simulations of the GENESIS global climate model in response to obliquity and precession cycles, Tabor et al. (2014, 2015) report minor changes in cloud albedo in response to the orbital cycles, without reporting on the longwave effect.
On CO2:
Following the comments from RC1, we have re-run the experiments for a lower CO2 level of 235 ppm which is more in keeping with Early Pleistocene levels. We intend to document the results for both this lower Early Pleistocene CO2 level (235 ppm) and PI level (280 ppm) in an Appendix section, which we will reference in the discussion to show the robustness of the findings for different CO2 levels.
Specific Comments
L57: Please add a reference here.
We have added the reference for Huybers (2006) and Huybers and Tziperman (2008) to support the statement that ISI is considered more representative for the ablation of NH ice sheets over the duration of melt season.
L103: The spatial extent of the mixed layer depth seems unusual. Please provide a justification for why this specific location was set.
The zonal-mean and prescribed ocean circulation model is taken from Bintanja (1997), which largely forms the basis for ZEMBA. It comprises an ocean basin spanning 70°S to 80°N, together with a passive mixed layer of 100 m depth north of 80°N, which we retain. In the original publication, the choice was justified by observations indicating negligible meridional ocean heat transport north of 80°N. We find this to hold largely true when comparing ZEMBA’s ocean heat transport against that inferred from a NorESM2 pre-industrial run (Gunning et al. 2025) and other estimates derived from reanalysis products (Trenberth and Fasullo, 2017).
Bintanja, R. Sensitivity experiments performed with an energy balance atmosphere model coupled to an advection-diffusion ocean model. Theor Appl Climatol 56, 1–24 (1997). https://doi.org/10.1007/BF00863779
Gunning, D. F. J., Nisancioglu, K. H., Capron, E., and van de Wal, R. S. W.: ZEMBA v1.0: an energy and moisture balance climate model to investigate Quaternary climate, Geosci. Model Dev., 18, 2479–2508, https://doi.org/10.5194/gmd-18-2479-2025, 2025.
Trenberth, K. E., and J. T. Fasullo (2017), Atlantic meridional heat transports computed from balancing Earth's energy locally, Geophys. Res. Lett., 44, 1919–1927, doi:10.1002/2016GL072475.
L165-167: What is the reasoning here? Is it because the sea-ice albedo is set to 0.7?
The sea ice implementation in ZEMBA involves sea ice of a prescribed thickness forming when surface ocean temperatures drop below the freezing temperature. It is difficult to diagnose the exact reason for the discrepancy between this simplistic representation and the far more complex representation in CMIP6 models. For one, the sea ice implementation does not involve a reduction in sea ice albedo following melt, which might weaken the albedo feedback, nor does it contain leads. Another potential factor is the coarse resolution of the model (5°). Following comments from RC1, we reran experiments for CO2 levels of 235 ppm in keeping with Early Pleistocene levels. We found latitudinal bands that sat close to the sea ice threshold can switch rapidly from sea ice cover to no sea ice cover in discrete steps. We are now performing the same simulations at a finer resolution (following optimisation of the code) which may improve the comparison.
L168-169: I’m asking out of curiosity, but if the model was re-tuned to allow less Arctic sea ice during the LIG (127ka) by lowering the sea ice albedo, would this result in a model where precession exerts a stronger influence?
It is difficult to predict the exact response to a reduction in sea ice albedo. If the sea ice albedo was reduced and the model re-tuned to reproduce the same pre-industrial climate state under the new albedo, this may weaken the albedo feedback due to a weaker contrast between the open-ocean and sea-ice albedo. As the orbital experiments here, and those of Tabor et al., (2014), suggest that obliquity generates a stronger albedo feedback than precession through its influence on annual-mean insolation, it is possible the obliquity influence would be slightly weakened, though with limited certainty.
L181: Is this a typo for 71%?
Yes, this is a typo. In the new runs with lower CO2 levels these numbers are slightly different, but we have ensured the number reported in the text is the same as shown in the figure.
L185: Please refer to the specific Figure being discussed to improve readability.
The revised manuscript will address this.
Section 4.4: There appears to be some confusion between Fig. 8 and Fig. 5. Please clarify.
Yes, Section 4.4 erroneously references Figure 5 instead of Figure 8, which will be addressed in the revised manuscript.
L282-305: While informative, this paragraph does not seem directly relevant to the results and could be removed. If retained, please provide a concise explanation of how ZEMBA differs from other EBMs to justify the differing results.
We will condense this paragraph on the differences with other EBMs.
L306-317: This is a repetition of the Methods and Results sections and should be removed.
The revised manuscript will condense these repetitions of the results section and focus on enriching the discussion with the results from similar orbital sensitivity experiments.
L329: Should the large thermal inertia of the ocean also be highlighted as a key factor here?
Yes, we agree that both the fixed ice sheet cover south of 75S, and the large proportion of open ocean immediately south of this, with a large thermal inertia, will both contribute to the reduced variability in surface temperature in the high southern latitudes relative to the north, which we will clarify in the revised manuscript.
L355-373: These lines repeat the Results section and should be condensed.
As before, the revised manuscript will condense these repetitions of the results section and focus on enriching the discussion with the results from similar orbital sensitivity experiments.
Figures
Fig. 6: Please include an explanation of the horizontal dashed lines (mx and mn) in the caption.
The horizontal dashed lines represent the average value of summer insolation at 65N across all maxima (mx) and minima (mn), which will be clarified in the caption.
Fig. 7: Please specify in the caption which experiment is being analyzed.
This will be addressed in the revised manuscript.
Reference
Arima, N., Yoshimori, M., Abe-Ouchi, A., O'ishi, R., Chan, W.-L., Sherriff-Tadano, S., and Ogura, T.: Impact of the temperature-cloud phase relationship on the simulated Arctic warming during the Last Interglacial, Clim. Past, 22, 891–913, https://doi.org/10.5194/cp-22-891-2026, 2026.
Erb, M. P., Broccoli, A. J., and Clement, A. C.: The Contribution of Radiative Feedbacks to Orbitally Driven Climate Change, J. Climate, 26, 5897–5914, https://doi.org/10.1175/JCLI-D-12-00419.1, 2013.
O'ishi, R., Chan, W.-L., Abe-Ouchi, A., Sherriff-Tadano, S., Ohgaito, R., and Yoshimori, M.: PMIP4/CMIP6 last interglacial simulations using three different versions of MIROC: importance of vegetation, Clim. Past, 17, 21–36, https://doi.org/10.5194/cp-17-21-2021, 2021.
Tabor, C. R., Poulsen, C. J., and Pollard, D.: Mending Milankovitch's theory: obliquity amplification by surface feedbacks, Clim. Past, 10, 41–50, https://doi.org/10.5194/cp-10-41-2014, 2014.
Citation: https://doi.org/10.5194/egusphere-2026-1151-RC2
Citation: https://doi.org/10.5194/egusphere-2026-1151-AC2
-
RC3: 'Comment on egusphere-2026-1151', Anonymous Referee #3, 09 May 2026
This paper uses an energy balance model to examine the response to orbital parameters during the early Pleistocene. I tend to believe that the integrated insolation idea provides a satisfactory explanation for the 41 kyr cycles, but it is always a good idea to question accepted dogmas. One specific motivation that is mentioned here and that I found interesting is a possible cancellation of 21-kyr cycles between the NH ice sheets and Antarctica (Raymo et al. 2006). I therefore find the problem and motivation interesting and relevant. The tool of an inexpensive and well-tuned energy balance model seems appropriate for such a task. The analysis is careful and interesting, but I do feel some additional experiments and analyses are necessary to make a solid case, and could turn this into an even more meaningful contribution. I therefore recommend a major revision.
The first major issue is that to be relevant to the 41 kyr ice ages, the authors should, in my opinion, diagnose the surface mass balance (SMB) of the ice sheets rather than the temperature. The SMB depends on the temperature, radiation, accumulation, etc., and only it can provide the needed link to the ice age problem. This needs to be added as another panel to something like the current Figure 5.
The other major issue is that to test alternative ideas for the 41 kyr cycles, it seems to me that the paper needs to explicitly compare the results (specifically the SMB) to those of the integrated insolation idea. This requires adding another curve to Figure 5 (including the SMB panel), showing the results of an integrated insolation calculation. This requires finding the optimal insolation threshold to be used for calculating this diagnostic. The optimal threshold would give an SMB time series that is closely correlated with the time rate of change of the observed ice volume record.
A third and final major issue is that while the introduction discusses the cancellation idea, etc., the results do not address that explicitly. By adding an SMB diagnostic, the authors could actually address the problem they pose as their motivation.
Finally, I find it interesting that sea ice plays an important role here (sections 4.3, 4.4, and conclusions). There is some literature on the role of sea ice in glacial cycles, and it may be helpful to compare the current results to such related papers and discuss similarities and differences.
Citation: https://doi.org/10.5194/egusphere-2026-1151-RC3 -
AC3: 'Reply on RC3', Daniel Gunning, 17 Aug 2026
We would like to thank RC3 for their comments on the manuscript, which has allowed us to improve the robustness of our results and address issues with the model analysis. Following comments from all RCs, we are re-running the experiments for a lower CO2 level (in keeping with the Early Pleistocene) and at a higher resolution, prior to revising the manuscript with these updated results. Here, we present our responses, with RC3’s comments highlighted in bold, and our replies in normal font weight.
This paper uses an energy balance model to examine the response to orbital parameters during the early Pleistocene. I tend to believe that the integrated insolation idea provides a satisfactory explanation for the 41 kyr cycles, but it is always a good idea to question accepted dogmas. One specific motivation that is mentioned here and that I found interesting is a possible cancellation of 21-kyr cycles between the NH ice sheets and Antarctica (Raymo et al. 2006). I therefore find the problem and motivation interesting and relevant. The tool of an inexpensive and well-tuned energy balance model seems appropriate for such a task. The analysis is careful and interesting, but I do feel some additional experiments and analyses are necessary to make a solid case, and could turn this into an even more meaningful contribution. I therefore recommend a major revision.
The first major issue is that to be relevant to the 41 kyr ice ages, the authors should, in my opinion, diagnose the surface mass balance (SMB) of the ice sheets rather than the temperature. The SMB depends on the temperature, radiation, accumulation, etc., and only it can provide the needed link to the ice age problem. This needs to be added as another panel to something like the current Figure 5.
As with RC1, we agree that an analysis of high-latitude temperature alone is not sufficient to inform on the ’41-kyr’ problem. The revised manuscript will dedicate a section specifically to the analysis of metrics in ZEMBA more relevant for glacial-cycle periodicities, namely the surface mass balance over land. Initial results from our higher resolutions show both a strong obliquity and precession influence on these variables. The obliquity component is stronger than expected given its much smaller influence on summer insolation, due to the obliquity-sea ice mechanism discussed in the manuscript. Nonetheless, there is a strong obliquity and precession component, not consistent with the 41-kyr cycles, which will be made explicit in the revised manuscript.
The other major issue is that to test alternative ideas for the 41 kyr cycles, it seems to me that the paper needs to explicitly compare the results (specifically the SMB) to those of the integrated insolation idea. This requires adding another curve to Figure 5 (including the SMB panel), showing the results of an integrated insolation calculation. This requires finding the optimal insolation threshold to be used for calculating this diagnostic. The optimal threshold would give an SMB time series that is closely correlated with the time rate of change of the observed ice volume record.
In the supplementary to the manuscript, we compared the annual-mean temperature to the various metrics of summer insolation, including the integrated summer insolation (ISI) metric. Following the comments here and from RC1, comparing the SMB over land to these insolation metrics is more appropriate, which we can carry out for the standard and fixed-orbital experiments in the revised manuscript. The addition of these SMB fields alongside ISI and other metrics of summer insolation would provide a useful addition to the analysis. As mentioned above, the SMB fields do exhibit a stronger precession component than the rate of change of ice volume, pointing to other missing feedbacks not included in the ZEMBA experiments.
A third and final major issue is that while the introduction discusses the cancellation idea, etc., the results do not address that explicitly. By adding an SMB diagnostic, the authors could actually address the problem they pose as their motivation.
We agree the current manuscript mentions the cancellation idea but does not properly address it in the discussion or results, other than a comparison of surface temperature. The main emphasis of the manuscript was the obliquity sea ice feedback (noted by Tabor et al. 2014) and the revised manuscript will explore this further with the fixed sea ice and snow cover experiment. Therefore, the revised manuscript will condense the sections talking specifically about the cancellation theory. However, given that our initial results from higher resolution runs of ZEMBA point to strong precession component in SMB at the high northern latitudes, its phase relationship to SMB in high southern latitudes will be explored and included in the supplementary.
Finally, I find it interesting that sea ice plays an important role here (sections 4.3, 4.4, and conclusions). There is some literature on the role of sea ice in glacial cycles, and it may be helpful to compare the current results to such related papers and discuss similarities and differences.
A major emphasis on the manuscript was placed on the amplification of obliquity on air temperature via sea ice feedbacks. As noted by RC2, similar findings were documented by Tabor et al. (2014) using a general circulation model. The revised manuscript will compare the results from ZEMBA to this and other highly relevant studies to compare similarities and differences. Additionally, we intend to explore this feedback further in the revised manuscript with the added benefit of being able to perform experiments with fixed sea ice and fixed snow cover.
Citation: https://doi.org/10.5194/egusphere-2026-1151-AC3
-
AC3: 'Reply on RC3', Daniel Gunning, 17 Aug 2026
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 704 | 419 | 57 | 1,180 | 127 | 50 | 69 |
- HTML: 704
- PDF: 419
- XML: 57
- Total: 1,180
- Supplement: 127
- BibTeX: 50
- EndNote: 69
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Review of Gunning et al. submitted to Climate of the Past: “Dominance of Obliquity over Precession in Polar Temperature Variability: Insights from an Energy Balance Model”
In this manuscript, the authors investigate the climate and atmospheric response to the 41-kyr glacial cycles in the Early Pleistocene using their ZEMBA model (Gunning et al., 2025). ZEMBA is a one-dimensional energy-balance climate model that considers atmosphere, land, and ocean at each latitudinal grid point, as well as hydrological cycles. The performance of the model was evaluated in their previous study (Gunning et al., 2025) for the preindustrial and the Last Glacial Maximum (LGM) periods. In this study, they first present the model's performance in reproducing the Last Interglacial (LIG) period, evaluating its performance under different astronomical forcing. I like the concept of ZEMBA, which is a useful tool for understanding long-term climate change. However, the present manuscript lacks context on the 41-kyr problem, and the absence of ice-sheet modeling limits the model's applicability. My comments are summarized below, with capital Ls representing the line number.
Gunning, D. F., Nisancioglu, K. H., Capron, E., & van de Wal, R. S. (2025). ZEMBA v1. 0: an energy and moisture balance climate model to investigate Quaternary climate. Geoscientific Model Development, 18(9), 2479-2508.
Major Comments
1.
The authors elegantly summarize the so-called 41-kyr problem in the Introduction. They clearly explained two standing hypotheses (Huybers, 2006; Raymo et al., 2006) and their issues in accounting for the lack of precession periodicity in δ18O; however, they did not explain how the ZEMBA investigation can help resolve the 41-kyr problem, which makes the contribution of the present study very unclear. Explaining the 41-kyr problem ultimately requires ice-sheet modeling, which is not included in the current ZEMBA model, thereby limiting its contribution to the 41-kyr problem. When I first read through the manuscript, I thought it extended the hypothesis presented by Raymo and Nisancioglu (2003), who suggested that the latitudinal insolation gradient, dominated by 41-kyr cyclicity, affects heat and moisture fluxes to high-latitude regions and hence the 41-kyr periodicity in ice-sheet mass. In this sense, I think this study contributes to the 41-kyr problem even without ice-sheet modeling, but the linkage between what this study did and the 41-kyr problem is very unclear in the present manuscript. I recommend clearly situating this study within the context of the 41-kyr problem in the Introduction and enriching the Discussion.
Huybers, P. (2006). Early Pleistocene glacial cycles and the integrated summer insolation forcing. Science, 313(5786), 508-511.
Raymo, M. E., & Nisancioglu, K. H. (2003). The 41 kyr world: Milankovitch's other unsolved mystery. Paleoceanography, 18(1).
Raymo, M. E., Lisiecki, L. E., & Nisancioglu, K. H. (2006). Plio-Pleistocene ice volume, Antarctic climate, and the global δ18O record. Science, 313(5786), 492-495.
2.
As I mentioned in the comment above, the current ZEMBA model does not couple to the ice-sheet model. The authors discuss temperature responses in high-latitude regions, particularly the dominance of the 41-kyr periodicity, accounting for the effects of variations in ice-sheet mass. However, the formation of ice sheets affects the climate via many feedbacks, including changes in surface albedo, altitude, and atmospheric circulation. Without ice-sheet variations, it is difficult to assess the robustness of their results, especially their discussion of snow cover and sea-ice distributions. I think ice-sheet modeling is highly recommended to discuss the 41-kyr problem in this study, as was done using both a simplified model (Huybers and Tziperman, 2008) and more sophisticated models (Tabor et al., 2015; Tabor and Poulsen, 2016; Willeit et al., 2019; Watanabe et al., 2023). Even if it is difficult, careful justification, mention of the robustness of their results, and the uncertainty associated with the lack of feedback, especially the ice-sheet variability, and the potential impact of the snow cover change and heat-moisture budgets discussed in this study on ice-sheet mass balance are required.
Huybers, P., & Tziperman, E. (2008). Integrated summer insolation forcing and 40,000‐year glacial cycles: The perspective from an ice‐sheet/energy‐balance model. Paleoceanography, 23(1).
Tabor, C. R., Poulsen, C. J., & Pollard, D. (2015). How obliquity cycles powered early Pleistocene global ice‐volume variability. Geophysical Research Letters, 42(6), 1871-1879.
Tabor, C. R., & Poulsen, C. J. (2016). Simulating the mid-Pleistocene transition through regolith removal. Earth and Planetary Science Letters, 434, 231-240.
Watanabe, Y., Abe-Ouchi, A., Saito, F., Kino, K., O’ishi, R., Ito, T., ... & Chan, W. L. (2023). Astronomical forcing shaped the timing of early Pleistocene glacial cycles. Communications Earth & Environment, 4(1), 113.
Willeit, M., Ganopolski, A., Calov, R., & Brovkin, V. (2019). Mid-Pleistocene transition in glacial cycles explained by declining CO2 and regolith removal. Science Advances, 5(4), eaav7337.
Specific Comments
L129–135: According to Table 1, it seems that the authors conducted transient simulations from 2.45 to 1.2 Ma using a constant atmospheric CO2 level at the preindustrial value (284 ppm). This value is too high for the mean value representing the early Pleistocene. This should be explained and justified in the main text of the Methods section. The potential impact of the atmospheric CO2 level may be discussed in the latter part.
L144–146: The peak month of the isolation anomaly will differ if the authors apply the calendar-effect adjustment (Bartlein and Shafer, 2019). I recommend that the authors apply the calendar-effect adjustment, which will allow readers to better understand and compare the results with those of previous studies, such as Otto-Bliesner et al. (2021).
Bartlein, P. J., & Shafer, S. L. (2019). Paleo calendar-effect adjustments in time-slice and transient climate-model simulations (PaleoCalAdjust v1. 0): Impact and strategies for data analysis. Geoscientific Model Development, 12(9), 3889-3913.
Otto-Bliesner, B. L., Brady, E. C., Zhao, A., Brierley, C. M., Axford, Y., Capron, E., ... & Zheng, W. (2021). Large-scale features of Last Interglacial climate: results from evaluating the lig127k simulations for the Coupled Model Intercomparison Project (CMIP6)–Paleoclimate Modeling Intercomparison Project (PMIP4). Climate of the Past, 17(1), 63-94.
L187–193:
Are the changes of atmospheric and oceanic heat transport rates associated with different astronomical forcing compatible with previous results obtained using a general circulation model (e.g., Mantsis et al., 2011; 2014; Erb et al., 2013; Tabor et al., 2014)?
Erb, M. P., Broccoli, A. J., & Clement, A. C. (2013). The contribution of radiative feedbacks to orbitally driven climate change. Journal of climate, 26(16), 5897-5914.
Mantsis, D. F., Clement, A. C., Broccoli, A. J., & Erb, M. P. (2011). Climate feedbacks in response to changes in obliquity. Journal of Climate, 24(11), 2830-2845.
Mantsis, D. F., Lintner, B. R., Broccoli, A. J., Erb, M. P., Clement, A. C., & Park, H. S. (2014). The response of large-scale circulation to obliquity-induced changes in meridional heating gradients. Journal of Climate, 27(14), 5504-5516.
Tabor, C. R., Poulsen, C. J., & Pollard, D. (2014). Mending Milankovitch's theory: obliquity amplification by surface feedbacks. Climate of the Past, 10(1), 41-50.
L199: There should be a period at the end of the sentence.
Figure 5: Consider using a color set that is accessible for people with color vision deficiencies. Same for Figure S4.
L317–326: Same as my comment for L187–193.