Thank you to the authors for this interesting paper. I am basically supportive of what you are doing here. I appreciate that your main focus is atmospheric chemistry, and there you produce some very interesting results. However, I have some clarifying questions about the human dimensions of your scenarios. I think their presentation should be tweaked to guard against potential misuse, e.g., in support of unsuitable savanna abatement schemes. 



(1) Human impact on the baseline scenario



I think you need to do more to recognise that in your study region, human controlled burning already plays a significant role in the fire regime. It is good you recognise that EDS burning is present in many parts of the study region. However, it is not always as simple as human EDS burning vs LDS wildfires. 



For example, in North-eastern Zambia, which has a significant impact on your results, there is widespread fire use for shifting cultivation^1,2,3. This often occurs very late in the dry season, just before the first rains. This then, is a very different form of fire use from early dry-season burning: it either occurs before the first rains, or it doesn't make sense from an agricultural point of view. Hence, it is not that surprising that pushing this into the early dry season has a substantial impact on RF, but I'm not sure of the real world applicability of this finding.



(2) Human implications of shifting fire use



I think this difficulty may arise from an overreliance on the Australian case study as a conceptual model. There, Aboriginal fire was systematically removed from the landscape, before its targeted reintroduction. As such, a targeted early burn was possible. This is not the case in your study region, where fire use is a fundamental component of peoples' livelihoods, and often their social/religious identities, and occurs at many points throughout the dry season for different reasons. Hence, intervening in the way implied by your scenarios is likely to be much more complicated than in the Australian case [see ⁴ for a real world example of what happens when this goes wrong].



It is good that you note the possibility of local/regional scale difficulties in implementing your scenarios. However, the section beginning at line 674 should be rewritten to reflect the different context in which your scenarios are being run to Northern Australia, and do more to highlight they are theoretical maximums that would have very substantial feasibility challenges for real world application. Some more engagement with the specifics of the study context would be useful. See ⁵ for a global synthesis / index of literature on human fire use, from where I located refs ^1,2,3. 

Thank you again for the interesting work.

Oliver Perkins

1 https://repository.kulib.kyoto-u.ac.jp/items/02ea39b5-5ea9-4571-a37c-12afc3ca5f0b

2 https://www.tandfonline.com/doi/full/10.1080/00380768.2014.883487

3 https://www.jstor.org/stable/30135835#metadata_info_tab_contents

4 https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2023EF003552

5 https://figshare.com/collections/DAFI_a_global_database_of_Anthropogenic_Fire/5290792/4