Through model simulations, Zhou et al. found that although dust emission is greatly reduced during the mid-Holocene Green Sahara periods compared to PI, the emission of isoprene and monoterpenes increased substantially (~4 fold). The increased emission of these biogenic volatile organic compounds (BVOCs) enhanced the formation of secondary organic aerosol (SOA), eventually caused a 17% increase in cloud condensation nuclei. Such processes have been overlooked in almost all previous studies. The study is thus quite novel and the results have important implications to the climate impact of a green Sahara, and should be a useful contribution to the journal Climate of Past. The manuscript is well organized and written and the methods and results are reasonable, I suggest a minor revision.

Comments:

1. In paragraph between lines 36-47, the influence of lakes on precipitation and west African monsoon should be mentioned. A few studies have worked on this, for example, Specht et al. (2022; https://doi.org/10.5194/cp-18-1035-2022) and Chandan and Peltier (2020; https://doi.org/10.1029/2020GL088728).
2. L94: “formextremely” misses a space.
3. The model was run for only two years, have the authors tested and checked that the results were stable already at the second year?
4. L162: “a-1” -> a^-1
5. L191: I don’t understand what “the latter one” refers to.
6. Figure 2, are the simulated emissions in PI reasonable compared to the observations? It would be useful if the authors can show the comparison between the pi_orig and the modern observations in the supplementary material.
7. L270-282: Table 3 should cited in this paragraph
8. L323: “significant” -> “significantly”
9. In the captions of Figures 5&6, do the authors mean “same as Figure 4”?
10. L125: It will be better if the authors directly state here that the present-day meteorological data were used in all 5 experiments. Although the sentence implies this but I assume it was only for the three PI experiments, and did not realize this until the end of the manuscript. This is a caveat of the study because impact of BVOC on CCN might be different if the MH meteorological data (some of the authors clearly have such data from model simulations) had been used.

Dr. Zhang,

Thank you for assigning me to deal with this manuscript entitled” Simulating the dust emissions and SOA formation over Northern Africa during the mid-Holocene Green Sahara period”. This is an interesting work and is proposed to publish on the Climate of the Past after minor revisions.

Best,

Yong

Comments to the authors

Please delete the following texts in the section 2.2, as I do not think it is really relevant to the scientific issues discussed, but rather to the technical details of the model setup.

“The model was installed on Puhti supercomputer at CSC (IT Center for Science, Finland), and 90 CPU (Central Processing Unit) cores were utilized for each parallel simulation run. One simulation year cost about 10 hours in real life.”

Here The authors conduct two years run of TM5-MP：one year is used for spin up and the other is used for analysis. I am not sure if it is too short for one year simulation to calculate the mean state of variables.