This paper describes a compilation of published 137Cs and 239,240Pu results in soils from the Southern Hemisphere in a database prepared under the AVATAR project. The aim of the database is to put together the scarce information on those radionuclides that is available for the southern hemisphere and use that data to understand the sources of anthropogenic radioactivity to that region using the isotopic composition of Pu. Besides, by applying a machine learning algorithm, the environmental factors that might influence the distribution of those radionuclides are identified.

Having such compilation and interpretation studies is necessary to understand reported results by different authors for a certain region and make the most of that data. Thus, the paper presents an interesting and useful study for the scientific community. The paper is correctly written and organized, and figures are representative and of high quality. However, key aspects are not properly addressed and it needs a major revision before being finally published.

General comments:

-Throughout the paper, the health risks of the artificial radionuclides are commented in different parts. I think such comments should not appear in this paper, which is focused on a compilation of reported data and assessment of the values. Besides, what is said about health risks of anthropogenic radioactivity is not rigorous and not realistc at all, and create a sense of drama. I refer here to the UNSCEAR 2000 report regarding the dose assessment for anthropogenic radioactivity. 

-When discussing the Pu ratios to study contamination sources, the heterogenous nature of the contamination in the Southern Hemisphere should be considered. There are two papers reporting Pu ratios and inventories in soils from all over the northern and Southern Hemispheres: Hardy et al., 1972 and Kelley et al., 1999. Both papers use the same set of soil samples from all over the world, but Hardy et al., analysed aliquots of 1 kg, and Kelley of 5 g by novel techniques. Thus, different effects are observed in both studies. Indeed, Kelley et al., observed heterogenous results (variable 240Pu/239Pu ratios) in samples from the Southern Hemisphere, far from testing sites, when analysing different aliquots from the same sample. Such effect was not observed by Hardy, since he processed 1 kg aliquots. The same heterogenous behaviour has been observed in Chamizo et al., NIMB, 2011, (doi:10.1016/j.nimb.2011.04.021) when analysing soil samples from Chile. And a similar effect was observed in peat bog cores from Madagascar in (Chamizo et al., 2020, doi.org/10.1016/j.scitotenv.2020.139993). So the fact that low ratios are not observed in a specific sample does not mean that the actual ratio is representative. It seems that in the Southern Hemisphere the Pu contamination is quite heterogenous because of the influence of low-yield French and British tests. This should be considered when discussing the sources of radionuclides and the reported results.

Specific comments:

1. Introduction.

Lines 52-60: The focus of the paper is not properly presented. The introduction should start in Line 58, where the use of anthropogenic radionuclides (only the ones with a long half-life like 239Pu and 240Pu, which should be also clarified) as golden spikes of the Anthropocene is stated. The health risks of artificial radioactivity should not be mentioned in the paper. I suggest removing the reference to such health risks from the introduction and the abstract, as stated before.

Line 68:  239+240Pu appears out of the blue… Please, name the Pu isotopes independently, 239Pu and 240Pu, and, at some point, state that they have been traditionally measured by alpha-spectrometry and that´s why the joined 239+240Pu activity is reported (since alpha spectrometry is not able to separate the alpha emission of both isotopes). This applied to the whole paper. Besides, when naming for the first time the different radionuclides, indicate the half-lives of them.

Line 74: What beta particles? Beta minus?

Line 83: The dispersion of radionuclides in the atmosphere depends on their physic-chemical properties. Caesium and Iodine are volatile and can be transported long distances. Transuranic are mainly attached to particles, and are dispersed more locally and regionally. This should be clearly explained.

Line 93: GF peaked beginning of the 1960s.

Line 100: 239Pu and 240Pu have been extensively studied. They are not “new” emerging tracers anymore.

Line 104: Note that the actual dispersion of actinides from the Chernobyl accident has not been clarified so far. There are papers reporting the presence of anthropogenic U and Pu from Chernobyl in the Baltic Sea. See, for instance, Lin et al., 2021. https://doi.org/10.1021/acs.est.1c02136

1. Origins of fallout 137Cs and 239+240Pu in Southern Hemisphere soils

Line 143: Comment on the local/regional impact of the Hiroshima and Nagasaki bombs.

Lines 144-145: It is mentioned that thermonuclear weapons introduced most of the FRN in the stratosphere. This should be further explained and illustrated with data from the UNSCEAR 2000 report.

Line 152: Differentiate between “pure fission” atomic weapons and “thermonuclear” weapons. Is it fussion (with double s)?

Line 179: 239+240Pu differs…

Line 180: Sr90, Cs137 and 239,240Pu have also different geochemical properties.

Line 192: Comment also on the fact that a significant fraction of the Southern Hemisphere is covered with water.

1. AVATAR-Soils Database overview

Line 306: Be careful when stating that Hardy et al., is the only paper reported overall Pu results. A comment on the subsequent study by Kelley et al.,  should be given also.

Line 355: Say “the 240Pu/239Pu atom ratio”. This applies to the whole paper.

1. Distribution and sources of fallout 137Cs and 239+240Pu inventories

Figure 4: I think quoting just the fission yield of the nuclear tests is not correct. I think the total yield of the detonations should be represented instead, which is the most reliable number we have.

Line 405: Be careful when stating that only 239+240Pu results were reported by Hardy et al., 1973. The work done in Kelly et al., 1999, should be also mentioned and put in context.

Table 1: How are the ratios for the pre-moratorium period estimated? Note that the Arctic and the Antarctic are very pristine areas that might not be comparable with soils from the general environment. You could also cite the values reported in (López-Lora et al., 2023, https://doi.org/10.1021/acs.est.2c07437) for a sediment core from the Baltic Sea where the two NWT periods are identified and studied independently.

Table 1: Please, check the 240Pu/239Pu ratios for the Franch and British tests. The given values are very low, and might not be truly representative. For the French tests, a value of 0.05 was reported in (Chamizo et al., 2015, http://dx.doi.org/10.1016/j.nimb.2015.05.008). For the British tests, similar values were reported in (Johansen et al., 2014).

Table 1: When reporting values for the fallout source “Global Fallout”, the period 1952-1963 should be named, since soils keep the integrated signal for the whole period, and not for 1958-1963. 

Line 505: Comment on the problem of the heterogenous nature of the Pu contamination in the southern hemisphere. See the work in (Chamizo et al., 2020, https://doi.org/10.1016/j.scitotenv.2020.139993) focused on peatbogs from Madagascar. This is an evidence that deviations in the Pu ratios have been also found in Africa.

“Expected”. “Assumed”. Readers would like to follow these authors but, absent clear uncertainty guidelines, we must remain suspicious. AVATAR database represents potentially the best state-of-the-art database of SH FRN, but authors have yet to convince this reader!