A review of “Satellite Observations of Smoke-Cloud-Radiation Interactions Over the Amazon Rainforest” by Herbert and Stier

The authors use a suite of satellite and reanalysis measurements to explore the effect of smoke on clouds and the overall radiative forcing. They do it for the dry season over the Amazon for 18 years of data. They describe part of the diurnal cycle by comparing associations between smoke AOD and cloud properties during TERRA and AQUA passing times.

The methods are well described, and the analysis for such a time series shows convincing results. However, the interpretation of the results is not clear enough, and there are a few important points that the author should explore more:

• These is an inherent problem in the cloud detection results in cases of high AOD. The higher the AOD, the more likely that smoke areas will be classified as clouds. Such a problem dictates a positive bias of the reported cloud fraction with AOD. Moreover, an average AOD of 0.8 in one-degree resolution might result from much higher AOD pixels in the core retrieval resolution. I encourage the authors to explore days with high AOD and to look at both aerosol and cloud properties in a few level 2 data to understand the inputs to the level 3 averages. It is easy to see the above problem. It influences their reported results on AOD-cloud associations in the high AOD regimes.
• On the same note – Overland in cloudy conditions, the AOD retrievals could be sparse. Therefore many level 3, one-degree pixels could report an average AOD that originates from only a few true retrieved pixels. In particular for cases with high cloud fractions. Again creating a bias in the data reliability as a function of CF. The author should explore it and avoid using level 3 pixels that are the outcome of aerosol retrievals from a small fraction of the domain.
• Another related front that is not covered here is related to problems of measuring aerosols in the vicinity of clouds. The authors should describe the potential problems (both in the aerosol and cloud domains) and explain why they trust the reported results.
• A significant part of the reported aerosol effect is linked to aerosol absorption. Absorption is hardly mentioned in the text (it is referred to vaguely as ARI, which in my opinion, is not clear enough). If indeed the authors think that aerosol absorption plays an important role in the reported effects, they should explain it more, and they should provide some measures of the importance of absorption. For example, they can use newer satellite measurements to explore the optical properties of the Amazonian smoke. As well as Aeronet measurements for the SSA. I am aware that some radiative effects calculations are presented in the SI, but it is not explaining the above points.

The manuscript ACP-2022-796 entitled “Satellite Observations of Smoke-Cloud-Radiation Interactions Over the Amazon Rainforest” by Herbert et al. presents a comprehensive analysis of 18-year time series from multiple satellite sensors over the Amazon basin. The study focuses particularly on the influence of (heavy) biomass burning smoke during the peak of the Amazonian burning season on a spectrum of atmospheric parameters, including cloud properties, top-of-atmosphere radiative flux, rainfall, and column water vapor content. The smoke influence is represented here by the satellite-measured aerosol optical depth (AOD). The study concludes that the cloud and radiative properties of the Amazonian atmosphere are strongly modified by the smoke. Among other observations, the extent of smoke influence defines if convective activity is either enhanced or suppressed. Further, an interesting contrast between the morning and afternoon cloud responses has been found. The results are thoroughly discussed/integrated in the context of previous studies and existing knowledge.    

My overall recommendation is that the manuscript should be published after some relatively minor changes. From a scientific perspective, it clearly fits the scope of ACP. It is a thorough and interesting study, relevant for the field of Amazonian research and beyond. I am convinced that the study will become a useful resource and reference for researchers studying the smoke influence of atmospheric radiative transfer and aerosol-cloud interactions. The conclusions are important for discussions and predictions of the future development of the Amazon ecosystem, which is under increasing pressure from progressing climate and land use change. From a formal perspective, the quality of the manuscript is high - it is well-written and all arguments and aspects are presented clearly.

Below, you will find some general comments that I recommend to consider to further improve the quality of the manuscript:  

• AOD is the main parameter representing the atmospheric smoke/aerosol concentration in this study. I suggest adding a targeted paragraph, which outlines and discusses the relationship between AOD and the actual aerosol optical properties as well as their ability to act as CCN. In other words: To what extent does AOD (as a simplified parameter) represent the relevant microphysical properties of the aerosol and what are the limitations of this?
• The influx of African smoke, which appears to occur in elevated layers, is discussed briefly on page 2. This particular aspect might profit from some more discussion in the introduction and/or results and discussion section. This relates to the particular questions (i) whether the assumption that the smoke is only/mostly in the PBL is justified and (ii) whether the current analysis provides any indications for the extent of influx of African smoke into the central and southern Amazon basin.  
• The manuscript summarized a broad spectrum of aspects and observations, with manifold links to previous studies. Because of the density of information and the length of the manuscript, it is hard to read in some sections. I have got the impression that a (slightly) different overall structure might make the study more accessible. The current structure separates a ‘results section’ and a ‘discussion and conclusions section’. First, I think that the results section already comprises quite some discussion of the results already. Therefore, parts of the discussion section appear somewhat redundancy. I wonder if for this particular study a merged ‘results and discussion section’ might work better. Second, I think that the study needs a concise and short conclusion paragraph – separated from the longer discussion sections. This would make it easier for readers that just would like to grasp the essence and main conclusions.