Paine and coauthors present a new Hg record in a unique sedimentological archive from Lake Bosumtwi that records West African hydroclimate over the last 96 thousand years. The study is partly honing and improving upon the Hg proxy for paleoclimate reconstructions and partly discussing hydroclimate history of the region at various timescales. The study is well executed and very well written, but the impact of the new record is a bit limited by the uncertainty in Hg driving mechanisms, but also the pre-existing reconstructions of lake level that may be more powerful for the paleoclimate framing. Please find comments, technical edits, and suggestions for improving the figures below.

Moderate comments:

• The split at 73 ka should be tested with statistical tools such as changepoint analysis to see how significant it is, rather than a visual analysis. Much of the discussion hinges on the differences before and after this time, so it should be bolstered by significance testing.
• The pattern in Hg and pollen look a lot like ice volume, and perhaps this should be discussed in the context of the age model.
• The record should be presented using the newest published version of the core’s age model from Vinnepand 2024. This will provide more detail throughout, and despite the lower resolution Hg data around this 73 ka time, may help pin down when this shift is, potentially in the broader context of drivers like ice volume.

Minor comments:

• I suggest adding a sentence or rephrasing the last sentence of the abstract to have a broader outlook on Hg as a proxy for hydroclimate moving forward
• In the same vein, I may suggest starting the introduction more focused on the importance of reconstructing hydroclimate, rather than on Hg – and the second paragraph of the intro, too, means that this paper is focusing on the Hg proxy development, rather than understanding Bosumtwi hydroclimate. The mix of both directions in this paper is certainly a strength, but the goal of pinning down the Hg proxy is ultimately to understand hydroclimate
• Line 65: is it precipitation pattern? Or precipitation amount/strength? Or something else?
• Line 94: perhaps start with a sentence that is a bit more focused for this study
• Line 120: what is its domain?
• Line 122: perhaps add some more citations for orbital control of the WAM seen by leaf waxes, like from O’Mara and Kuechler
• Line 124: I suggest changing drought events to arid periods, if these records are focused on orbital-scale climate variability
• Line 129: over what time period?
• Line 161: maybe add one sentence or phrase about how this makes the Lake Bosumtwi archive ideal for refining the Hg-paleoclimate method or generating paleoclimate reconstructions in general
• Line 163: perhaps just leave out this short-term definition, as only seasonal changes are described. Unless these vary on 10 year timescales? The description goes from less than 10 to over 1000 year variability.
• Section 2.1.1: is the ITCZ also expanding and contracting? Rather than just moving and strengthening/weakening? Also, there is more evidence that the insolation gradient, rather than just precession, is controlling the WAM during the Pleistocene (see O’Mara et al 2024)
• Lines 221-224: Nobody will disagree with this statement, but an additional clause indicating why lake level alone is not a perfect measure of precipitation will lead well into the study of Hg.
• I recommend reporting correlation as r, rather than r2, so we can see numerically the direction of correlation. And when correlations are presented, please include number (n) and significance (p).
• Line 517: some records show a drying trend towards the LGM, particularly in eastern Africa (Garelick, Baxter, Lupien, Tierney). Please discuss how this fits into your discussion, if the focus is as trans-continental as it is now.
• Lines 526-527: this is a narrow statement – see a nice explainer of fuel-limited and moisture-limited by Karp 2023

Technical:

• Lines 63-64: remove one of the ‘direct’s
• Line 148: change Ma to Myr-old, or “dated to 1.08 Ma”
• Line 230: cite the map in Fig 1
• Line 240: changed generated to described or similar
• Line 256: might help to include the maximum spacing between samples too
• Line 261-262: can simply cite the supplemental data, rather than including a sentence
• Lines 383-385: instead of talking directly about Figure 2, try removing the first part of this sentence and then citing Figure 2 at the end.
• Line 386: I don’t think this last sentence of the paragraph is necessary – there are a lot of these statements throughout that are a bit redundant that can be cut down
• Line 516-517: “these records” needs clarification aside from the figure citation.
• Line 526: citation
• Sections 5.2.1 and 5.2.2: I suggest renaming these sections into more descriptive titles
• Line 539: change ka to kyr

Figures

• Figure 3: as mentioned above, if presented as r rather than r2, there won’t be a need for the italics and it’ll be clearer.
• Figure 4: no need to show precession, and consider including key insolation curves such as 20 or 30 N and perhaps the insolation gradient (23 N – 23 S) and 65 N to test for drivers of hydroclimate throughout the late Pleistocene

This manuscript uses a sediment core collected that dates to 96k to link Hg to climatic events in West Africa.

The manuscript is well written, and the methods and QA/QC are clearly explained.

I would remove the word new from the title.

There do not appear to be any keywords. These should not be the same as words in your title.

Line 111 There is a colon in this line that should not be there.

Line 128 focusses is not correct use focus

The section 1.3 titled West African Monsoon should be changed from this section basically introduces the reader to the research objectives. Do you have a research hypothesis?

Line 139 I would get rid of aims to assess and just use “assessed”

Line 142 explore should be explored. Note: I am a strong believer that what was done should be described in past tense.

WAM and ITCZ should be written out the first time used with the abbreviation afterwards in parenthesis.

Line 217 which should be that, and there should be no comma preceding this

Line 230 which should be that. Which is used when indicating something general, while that is used when discussing something specific.

Line 260 run should be analyzed

Please explain in the methods how the core was stored and sampled. It is very easy to contaminate samples with Hg. This must be clearly addressed.

Line 276 which should be that. I am not going to mention any more of these and will leave this up to the authors to correct.

Line 364 “broadly track” is subjective. It would help if you did some statistical analyses to make this statement more quantitative. Same with the discussion line 365.

Line 369 some statistical analyses would help quantify the discussion.

Line 389 ah here is the statistical analyses result. Please move up to the results section. Note: r2 values are meaningless without a p-value. Is it worth putting in the equation for the correlation? This might be of interest to others.

Don’t forget to put your statistical analyses method in the methods section.

Are there any Hg analyses of the crater walls? I think it would be a good idea to add a description of the geology associated with the crater area in the site description.

Figure 3 is blurry. Please identify the p-values associated with the r2 values.

Can you do some regression analyses for data presented in Figure 4? Would this make your discussion more robust?

Line 501 please include the Outridge reference since he was the first to state this.

Line 508 can you provide information on documented fire events in this area that would directly impact water in the crater?

Again, would it be worth doing some correlation analyses for data show in Figure 5 to make your paper more quantitative?

Line 537 seeks? How about this? This work combined geochemical data obtained for a sediment core collected from Lake Bosumtwi, Ghana, with climate data in order to explain measured trends in Hg concentrations.

I like the schematic, Figure 6.