Review of "Contribution of Lakes in Sustaining Greening of the Sahara during the Mid-Holocene" by Yuheng Li et al.

General comment. This paper adds to a body of studies of the effect of lakes during the North African "Green Sahara" mid-Holocene period. As it is rightly stated, there is still a discussion about which processes have enabled and sustained a relatively humid climate in that region during that period, and besides in particular vegetation, open water is one surface feature that has been proposed as a positive feedback mechanism involved in this interesting period of « recent » climate history. The manuscript does not add fundamentally new insights to this discussion, but as it stands, it is a basis for a useful contribution to this discussion, provided some necessary clarifications. These clarifications are needed in particular with respect to the model setup.

Specific and technical comments:

Line 20: Editorial - In several places in the introduction, reference is made to a recent review paper instead of older key papers. For example here in line 20 where only a (good and complete) review is cited, it might be interesting to expand the list of papers cited to include some preceding key papers. However, that’s an editorial question and it is also acceptable to only cite the review paper, for clarity.

Line 68: A general remark: This paper used an isotope-enabled version of a GCM. I expected some more isotope-related analyses in this paper, for example to provide insights into precipitation recycling in the various simulations. I was a bit frustrated not to see more on this, as this might add some rather unique information from this study.

Line 79: Nowadays, T42 is on the lower end of usual climate model resolutions. Is there a reason to think that the results might be sensitive to resolution? For example, are there higher-resolution studies of the West African Monsoon system with MIROC, and is the monsoon representation in MIROC sensitive to model resolution?

Line 92: "Figure S1a shows..." - Not very clear, figure hard to read. Can the procedure be explained in a bit more detail? I guess the main point is that the lake fraction in MH_ref and PI_ref (note typesetting error line 92, it should be subscript "ref") is weak, right? Because that provides a "almost no lake" reference for the other simulations. Can that be said more clearly?

Line 110: "1.48 x 108 km2" - please use superscripts correctly. What is the point to compare the lake area over NAf with the global land area? Lake fraction should be relative to the region you are looking at. Or is that the case here? Confusing. If it is relative to the entire land area of the Earth, it's huge...

Line 112: "In this study, wetlands are considered as lakes". What does that mean in the model world? How deep are the lakes? Does that simply mean that the water is present perennially? Please clarify how lakes are prescribed and treated.

Line 150: A r^2 of 0.33, is that really good?

Line 152: "Our simulations bias..." - Can this be clarified, e.g. by restricting the scatter plot to an area with, say, Africa, Southern Europe and Western Asia?

Figure 3: I appreciate that 200, 600 and 850 hPa winds and geopotential heights are also given, but it's unclear whether there is any reason why SM is associated with 200 hPa circulation, evap with 600 hPa, and t2m with 850 hPa. Is there a reason?

Figure 3a: Soil moisture changes. How much of that is prescribed? In the sense, does the prescribed lake water count here? Water quantities are huge, what does 1 m mean here (until what depth?)

Figure 3c: This is a strong cooling. What is the depth of these lakes? Is is thermal inertia due to depth or evaporative cooling?

Line 269: "Additionally..." - This sentence is not grammatically correct I think.

Figure 5a: Typo in the legend - should probably be "unitless" (as in the caption), not "uniteless".

Line 279: Here are the isotopes, but the explanation is hard to follow for non-specialist readers. This needs and deserves some more explanation.

Line 314: "Limited by..." - this is confusing, not well written. One wonders whether you have dynamical lakes and vegetation in the model (you don't, if I understand correctly). Please clarify - it would be good to provide a bit more detail in the methods section about this.

Line 321: "out components, such as orbital forcing and greenhouses..." - you mean "external forcings, such as orbital parameter and greenhouse gas changes" (or something similar)?

Line 327: "Limited by model dependency and module integration..." - this is unclear. Do you mean to say that the results are highly model-dependent (because the results from different studies are somewhat contradictory), and that they depend on the feedback mechanisms represented (e.g. dynamics lakes and vegetation included or not)?

Line 331: Full stop missing at the end.

Supplementary material:

Figure S2 - "Experiements" (typo). "(G)" missing in the lowest panel.