Please find my comments attached.

The objective of this study was to estimate the total magnitude and interannual variability of carbon fluxes over the Amazon basin through the use of inverse modeling techniques. The study utilized important vertical measurements of CO2 and CO over Amazonia and measurements from surface monitoring networks. Two sets of biospheric prior fluxes were used for inversion, and the study assessed whether the conclusions were influenced by prior emissions. The results of the study indicated that the Amazon region was a small carbon source to the atmosphere, mainly due to fire emissions, which corroborates the results reported by Gatti et al. (2021), albeit with varying magnitudes. The study also examined the contributions of different regions to the Amazonian carbon budget and found that emissions in the eastern Amazon were greater than those in the western region, primarily due to fires. Overall, this study is of great interest to the global carbon community. However, readers could benefit from additional details in certain areas. Furthermore, discussing the limitations of the study and potential avenues for future research would further enhance the impact of the study.

General Comments.

The model setup needs to be described in more detail to help readers visualize the experiment setup. The spin-up used for simulations, and any other relevant details should be provided. Additionally, since vertical profiles are being assimilated, discussing and providing validation plots regarding vertical transport would be beneficial. It is important to consider how much uncertainty in vertical transport might affect the results. The uncertainties and impacts associated with the coarse resolution of the inversion on the results should also be discussed. Additionally, it would be useful to know if all sites used in the inversion have full data coverage or if some have discontinuous data. The information on the data period should also be included in Table A1.

It is not clear whether the inversion is performed globally or regionally. If it is performed globally, it would be useful to compare the simulated growth rate with the observed global growth rate as a metric to test the modeled growth. (e.g., Fig. 5d in Chandra et al., 2022: https://acp.copernicus.org/articles/22/9215/2022/).

The author used the same profiles used in the inversion for the evaluation of atmospheric inversion. However, evaluating the inversion using independent observations not used in the inversion, e.g., HIPPO, AToM, CONTRAIL, or other regular aircraft measurements from other campaigns, would be more rigorous.

It is unclear which fire emissions are used in the calculation. For example, in line#165, it is mentioned that the fire emissions are optimized in the CO2 inversion estimate. Then, Section 2.2.3 discusses the optimization of carbon fire emissions from INVICAT using MOPITT CO. It is unclear whether these two are independent or the same. Additionally, it is unclear which fire emissions are used in the calculations of the Amazonia carbon budget. A comparison plot of prior and optimized BB will be helpful to visualize the correction in BB emissions.

Technical Comments

At line#231, the biomass burning emission ratios are given as 16 ppm CO/ppm CO2. IS the unit correct (or ppb/ppm)? And this number is not found in the cited reference.

The country boundaries in Figure 1 make it too messy. It would be better to remove the country boundaries and include the mask of the Amazon. Including the abbreviation of site names would also be helpful.