The article by Hällberg et al. presents an interesting and well thought study of seasonal variability of rainfall and climate over the last deglaciation in the Island South-East Asia (ISEA) region. The authors focus on how rainfall has changed regionally above the main landmass, that have themselves considerably been reshaped and shrinked along the deglacial eustatic sea level rise, and compare their transient model output with landmark paleoclimate (speleothem) reconstructions from different localities in the indonesian archipelago. They detail, in particular, the seasonality which seems to have been greater at 12 ka as compared to nowadays, and manifest as the development of a dry austral summer dry period in the southernmost part of the ISEA. They discuss their results in light of peat deposits that started accumulating during the Holocene, leading to a potential savanna corridor connecting the main islands of Indonesia and plains in the Sunda and Sahul shelves prior to their exundation. 

The analysis is nice, and I believe it should be published after some corrections that may be easy to deal with but could improve the manuscript, as detailed below.

First, the PI temp and precip patterns should be compared to modern data in Figure 3 or elsewhere, to highlight the ability of the model to reproduce local-scale climate variability, and even perhaps at seasonal and interannual timescales. The model seems to perform well, but rainfall seasonal and interannual variability at equatorial latitudes is, I think, notoriously challenging to reproduce with fully coupled GCMs. The authors should do better to convince the reader that the model is able to reproduce with a decent precision and accuracy at least the modern climate variability prior to inferring the transient evolution of past regional rainfall in such a complex environmental setting, with coastlines constantly varying. I think you may also want to discuss in greater details the large-scale outputs of PMIP4 outputs to highlight such point: those recent results indicate extreme variability in LGM rainfall model output scattering in the ISEA region (Kageyama et al., 2021, their figure 6 is particularly instructive).

I also think that some important data available should be cited and discussed as follows:

- On ENSO, you cite Clement, Koutavas and Sadekov to suggest that ENSO was enhanced durng the LGM, without discussing the contrasting results of Leduc et al., 2009, Ford et al., 2015 and Liu et al., 2017 that point to a reduced ENSO variability during the LGM. Those results that demonstrate that the LGM ENSO state is still debated should be explicitely discussed along with the Clement, Koutavas and Sadekov articles.

-I wonder why you choose to not discuss leaf wax isotopes (the Niedermeier and Konecky papers), other speleothem records (Griffiths), etc. without attempting to highlight them in Figure 2, as you’ve done for the tree other speleothem records. Also, the original paper evidencing a dry and open vegetation during the LGM at the Konecky site is Russel et al., 2014, and should also be cited, as dD interpretation is puzzling at this site as evidenced in Konecky.

-I also think you should better discuss what has happened at 12 ka in more details in the data as long as you opt for focusing on that time period. The results obtained in the simulation does not always fit to the data you highlight in Figure 2 for that time period, and there is a kind of overshoot in precip seasonality after 12 ka, when then YD resumed. I understand you can’t describe everything in its full complexity but your statements for the 12 ka in particular are not always met in reconstructions, and a lot happens before and after 12 ka in both data and climate simulation, as shown in Figure 2.

Some other minor details:

-Fig 5b & d, I don’t really see stronger vs weaker trade winds in the lower branch of the Hadley cell in panels a and c, but rather an eastward displacement of the convection site above the western Pacific at 12 ka WRT nowadays

-Still on Figure 5, the situation shown in panel c looks quite like the LGM situation as seen in Holstein et al., 2018 (their figure S6), it is interesting to note that the LGM and YD have the same effect on the convective cell displacement during those two contrasting yet cold time periods 

-You state in the discussion and the conclusion that the deglacial interval « resembles to ENSO », but you show only variability and seasonal features. Is it possible to differentiate variability in rainfall that occurred at interannual timescale ? The article by Liu et al., 2014, may help.

Ford, H. L., Ravelo, A. C., & Polissar, P. J. (2015). Reduced El Niño–Southern oscillation during the last glacial maximum. Science, 347(6219), 255-258.

Hollstein, M., Mohtadi, M., Rosenthal, Y., Prange, M., Oppo, D. W., Méndez, G. M., ... & Hebbeln, D. (2018). Variations in Western Pacific Warm Pool surface and thermocline conditions over the past 110,000 years: Forcing mechanisms and implications for the glacial Walker circulation. Quaternary Science Reviews, 201, 429-445.

Kageyama, M., Harrison, S. P., Kapsch, M. L., Lofverstrom, M., Lora, J. M., Mikolajewicz, U., ... & Zhu, J. (2021). The PMIP4 Last Glacial Maximum experiments: preliminary results and comparison with the PMIP3 simulations. Climate of the Past, 17(3), 1065-1089.

Leduc, G., Vidal, L., Cartapanis, O., & Bard, E. (2009). Modes of eastern equatorial Pacific thermocline variability: Implications for ENSO dynamics over the last glacial period. Paleoceanography, 24(3).

Liu, Z., Lu, Z., Wen, X., Otto-Bliesner, B. L., Timmermann, A., & Cobb, K. M. (2014). Evolution and forcing mechanisms of El Niño over the past 21,000 years. Nature, 515(7528), 550-553.

Russell, J. M., Vogel, H., Konecky, B. L., Bijaksana, S., Huang, Y., Melles, M., ... & King, J. W. (2014). Glacial forcing of central Indonesian hydroclimate since 60,000 y BP. Proceedings of the National Academy of Sciences, 111(14), 5100-5105.

Zhu, J., Liu, Z., Brady, E., Otto‐Bliesner, B., Zhang, J., Noone, D., ... & Tabor, C. (2017). Reduced ENSO variability at the LGM revealed by an isotope‐enabled Earth system model. Geophysical Research Letters, 44(13), 6984-6992.

Review of Hallberg et al.

Overall, I like the concept of the manuscript: describing seasonal changes in western tropical precipitation during the Younger Dryas climate event. However, I cannot accept the manuscript in its current form, and I suggest major revisions to the manuscript. The primary cause for the revision is to clearly separate which forcings are responsible for differences between the Younger Dryas state and Pre-Industrial (PI).

In the intro (paragraph on lines 49-55): the jury is still out as to how ENSO changed during the glacial times, including during the deglaciation. This paragraph and the paper should depict the current state of ENSO research. Reviewer #1 provides many important references.

The major issue: by comparing 12 ka to PI, the authors are conflating several different forcings that could drive the precipitation changes, such that it is not clear if the conclusions hold as presented. The summary of the difference in forcings between 12 ka (Younger Dryas) and 0 ka are:

• Precession of the equinoxes.
• Sea Level changes and the presence of large ice sheets on land
• Freshwater forcing reducing AMOC

To separate the forcings, and hence attribute what is causing the differences between the paleo and PI runs, the authors should compare different timeslices, in addition to the 12 ka and 0 ka.

• Usually the early Holocene (~10 ka) or the mid-Holocene (~6 ka) compared to 0 ka is used to identify changes forced by precession of the equinoxes, like in the PMIP protocal. This is *especially important when considering changes in the seasonal cycle, as precession has little effect on mean annual changes in insolation, but instead causes very large impacts on the seasonal distribution of insolation (e.g. Huybers, Science, 2006 or Huybers, QSR, 2007). The insolation changes in turn lead to changes in precipitation. Therefore, the authors need to make major changes to the manuscript, as the seasonal cycle is a major point of focus for this manuscript.

• Comparing the time interval just before the Younger Dryas, i.e. the Bølling–Allerød interstadial (BA) at 13-14 ka, with 0 ka will address both orbital changes and sea level/ice sheets. By then using the 10 ka run, which addresses precession only, and the ~13 ka run, one can isolate the climate effects due to sea level and ice sheets.

• Comparing the run during the BA with the YD will isolate the climate effects due to AMOC changes.

I am not suggesting to do any additional runs, such as 10 ka. Rather, present what runs have been completed, and present the output accordingly. Schenk, et al., Nature Comm., 2018, ran the BA interval. So that should be readily available to the authors. The 0, 12, and 13 ka runs should be used to better describe which forcings are driving the seasonal changes in precipitation of the western tropical Pacific.

Plot the coast lines in Figure 3-7 using the boundary conditions in the model, not modern observations of coastlines or bathymetry. This greatly facilitates the interpretation of the precipitation signals. One should look at where the model has land – not the actual planet.

If you want to use the term “El Niño like” to describe the mean state, one should include changes in the thermocline and sea surface salinity (Di Nezio, Paleo., 2011). Sea surface temperatures alone are not sufficient to describe the mean state when discussing changes in ENSO. Furthermore, what is the ENSO signal in the runs? 150 years is not adequate enough to converge on the median of the ENSO response (Lawman, et al., Science Adv., 2022), but it would be nice to see how the Niño 3.4 SSTA at 12 and 13 ka responds in CESM1, as it has a realistic ENSO.

Please update the Borneo record to use the recently published version (Buckingham, et al., GRL, 2022). It has a clear YD signal. I realize that this manuscript was submitted before the GRL manuscript was published. This is not to penalize, but rather to update, this manuscript. Also, the interpretations in that manuscript may help strengthen this manuscript.