The paper provides an analysis of greenhouse gas (GHG) emissions and trends in Africa over the past three decades, focusing on evaluating different datasets and their potential for verifying official country-reported data. The study examines emissions of carbon dioxide (CO2), methane (CH4), and nitrous oxide (N2O) using both bottom-up approaches (such as national inventories and ecosystem models) and top-down methods (including atmospheric inversions).

The findings contribute to understanding emission trends and uncertainties in Africa, which is crucial for climate policy and the goals of the Paris Agreement. Overall, the topic is interesting. However, I have some concerns as follows:

1. In addition to providing the datasets (https://doi.org/10.5281/zenodo.7347077), this paper needs to include datasets usage (quality control method, datasets limitation, etc.).
2. The method for calculating trends needs to be described and the impact of different trend calculation methods on trend results needs to be discussed.
3. As statistics play a crucial role in this study, it is important to provide further details, such as confidence intervals, to ensure its robustness.
4. When employing in situ surface networks for dataset validation, are there specific factors, such as latitude, longitude, climate zones, etc., that exhibit correlations with the product's quality?
5. Lines 89-90. BU methods and TD methods need to be further explained.
6. Line 124 and Line 162. Redefinition for the abbreviation "bottom-up (BU)". You have already defined it in line 89. Please recheck the manuscript to ensure that the same error does not occur.

Review of “Greenhouse gas emissions and their trends over the last three decades across Africa”, by Mostefaoui et al.

Overall this is a well written and comprehensive analysis of multiple datasets which make a really useful compendium of information that brings together pan-African emissions trends of different greenhouse gases. It makes a valuable contribution to the literature documenting and assessing the merits of different data sources and showing them all in one place.

Of particular value is the careful assessment of, and treatment of, uncertainty arising from UN reporting of land-use and the issues around this. The manuscript is very thorough and this really helps the reader understand the issues.

I have a few minor comments which may help presentation, but recommend publication once these are addressed.

My main comment is that due to its length, I think there may be a cleaner way to summarise and finish the paper. When I got to figure 10 (and also fig 11) I thought “this is a nice synthesis plot” (i.e. it feels like a really good place to finish and summarise everything you’ve found), but then the paper goes back into more details and uncertainties – I was then left not knowing whether to believe figures 10 and 11 any more…. It would be nice maybe to have the discussion of uncertainties first, and then finish with a clear synthesis figure like 10/11 which show your final best estimates.

Chris Jones

Minor comments

-       You note that land use CO2 emissions have the greatest uncertainty. It is worth noting that this is true globally and infact land-use CO2 are the most uncertain of all GHGs (at least when expressed as CO2e) according to IPCC

-       For N2O, You note the difference between bottom-up and top-down estimates. I’m curious how this compares with the global situation – is this similarly affected by the potential contribution of natural sources? Or is this just a problem for Africa (and if so why?). 

-       Given the importance of this uncertainty – could the paper recommend which estimate to use? Do you believe the top-down or bottom-up ones better? At the moment it just elaves the reader to see two very different possible answers….

-       Table 1: why do you list all 3 N2O inversions as different lines here, but for CO2 you just have “GCB ensemble” (which is itself multiple models)?

-       Line 291 – Pongratz is a good reference here for the issues of “loss of sink capacity”. https://esd.copernicus.org/articles/5/177/2014/esd-5-177-2014.pdf

-       Line 339 – why do you use GWP100 numbers from AR4 (16 years old?) – can you update these to AR6?