Hancock et al. present a comprehensive study that integrates satellite observations of CH₄ with an atmospheric transport model and prior emissions estimates to derive an optimal set of methane emissions for South American countries. This analysis further disaggregates the optimized emissions by sector and country, enhancing the atmospheric constraint on South American methane emissions. The topic is significant, and well-aligned with the scope of ACP. I am not an expert on inverse modeling but rather on in-situ measurements so I can’t comment deeply on the mathematical aspects of the inversion method and learnt a lot. Below, I outline a few questions and suggestions.

Line 93: what is the retrieval success rate specifically for South America?

Regarding Fig. 1: How would the seasonal variability of methane concentration affect these emission estimates? A spatial plot of the total number of samplings in each season during 2021 might be useful to include in the supplement. Is it possible to extend this inverse modeling setup to estimate total methane emissions on a monthly scale?

Line 115: "There are few observations over the mountainous Andes, affecting much of Chile and Peru, so the inversion for those countries relies significantly on glint observations offshore and on observations of transported methane." How does this affect the uncertainties in estimating emissions for this area?

How does the inverse model handle temporal variability in emissions? While the model optimizes the overall magnitude and spatial patterns in emissions, does it also optimize seasonal or year-to-year variability? In other words, does the model assume that the temporal distribution of emissions is known or fixed according to the prior temporal distribution?

Line 206: "We use 600 Gaussian functions as state vector elements to balance aggregation and smoothing errors." While a reference is provided, a brief explanation of why 600 Gaussian functions are used would be helpful.

Regarding Fig. 3: The ratios of the posterior/prior emissions in Fig. 3c show values close to zero or over 2 in many areas (e.g., Bolivia and Argentina). Does this imply that the inversion zeroed or doubled emissions? If so, are the resulting emissions reasonable?

Given the goal of the paper is to "evaluate the national inventories submitted to the United Nations Framework Convention on Climate Change (UNFCCC) under the Paris Agreement and to identify opportunities to improve countries' bottom-up reporting methods," including model-data comparisons against independent CH₄ observations is crucial for evaluating the inverse model. While there is a comparison with aircraft measurements in the Bolivian Amazon region, this is not enough. A supplement showing the bias of methane concentration in the prior and posterior run relative to in-situ observations in South America would be very informative.

Minor Point: Line 98: TCCON is first mentioned here. It should be referred to as the "Total Carbon Column Observing Network (TCCON)." Additionally, providing a brief explanation of TCCON would help readers unfamiliar with this field understand its purpose.

This paper presents a comprehensive analysis of methane emissions across South America. By employing high-resolution satellite data from TROPOMI and GOSAT, the authors present a comprehensive and spatially detailed estimation of methane sources, mainly anthropogenic sources, which constitutes a significant contribution to the comprehension of regional methane budgets. However, there are a few aspects of the paper that could be addressed to yield a more robust and comprehensive analysis.

The integration of data from two complementary satellite instruments helps to improve estimates through the use of inverse modelling. Nevertheless, a validation with independent CH4 observations available in South America would be beneficial for this study. For instance, the authors could undertake a comparison of the posterior mole fraction with data obtained from the ATTO tower in Brazil and vertical profiles in the Amazon region.

With regard to the regional budget, it would be beneficial to include a comparison with other previously published top-down estimates of total methane emissions for each country, such as those included in the Global Methane Budget.

The authors could additionally provide insight into the implications of their findings for policymaking strategies. This would be beneficial, as they evaluated national anthropogenic emissions inventories reported by individual countries to the UNFCCC. In particular, the authors could elucidate how these data could help local governments to mitigate methane emissions.

Specific comments are provided below.

Line 20: The term "correcting" may suggest that top-down estimates are inherently more accurate than bottom-up estimates, whereas both approaches are subject to their own sets of uncertainties. To prevent any potential misinterpretations, it would be helpful to use a term such as "adjusting" or "reconciling" when discussing the comparison or combination of these estimates. It is also important to discuss the limitations of both methods, top-down and bottom-up estimates, in the paper.

Lines 118-119: state that satellite observations are distributed throughout the year, but are most dense during the southern hemisphere dry season (June-September). How the lower density of observations during the wet season in comparison with the dry season could affect the posterior estimates. This is particularly relevant given that this region has extensive wetland areas, where the highest emissions are expected during the wet season.

Figure 1 illustrates the annual mean 2021 dry-column methane mixing ratios (XCH4) after subtraction of background, clearly demonstrating the absence of TROPOMI data in the Amazon region, which is compensated by GOSAT observations. However, an examination of the plot of the number of observations shows that there are fewer GOSAT observations during the months of April to June in comparison with other months. Please describe the extent of data coverage for South America during this period, with particular attention to the Amazon region.It would be beneficial to have a map as supplementary material that includes both TROPOMI and GOSAT dry-column methane mixing ratios (XCH4) for the initial period of the year, particularly April to June.

Lines 196-200: What is the lifetime of methane considering all the sinks, including oxidation by hydroxyl (OH) radicals and tropospheric chlorine (Cl), oxidation in the stratosphere, and uptake by soils?

Line 328: “Most of that increase is from anthropogenic emissions”. Does this imply that the prior estimated wetland emissions for the South American region are consistent with the atmospheric measurements? Alternatively, could the posterior wetland fluxes be more dependent on the prior estimates due to the limited observations in the Amazon region (which has larger methane emissions, as illustrated in Figure 2), as reflected in the low averaging kernel sensitivities? It would be beneficial to conduct a comparison with independent atmospheric observations to evaluate the posterior estimates.