The paper is based on an impressive amount of research work, and provides important, systematically collected, early information regarding droughts or dry conditions reported in documentary evidence, from the earliest period of European colonisation in different parts of North America. The reviews also emphasise the importance of this work, appreciate the ideas, and present an overall positive opinion about the paper. However, there are some very useful major and minor suggestions in the reviews that the author should consider in the revised version of his manuscript. Additionally, I also have some further comments, suggestions to the paper – partly independent, but mostly related to the reviewers’ comments and suggestions.

Independent of the reviewers’ comments, I have some doubts regarding the very general use of “archives of societies” and “archives of nature” in the title. Based on the content of the paper, under “archives of societies” you exclusively mean European colonial documentary evidence, and this is compared in the paper to the “archives of nature”, and under the later term you predominantly mean tree-ring based SPDI evidence and, in two cases, you also refer to stalagmite evidence. Based on the title, it is not very clear what evidence you will include in the paper. In fact, the term “archives of societies” in comparison with “archives of nature” would suggest that you deal with all source evidence primarily related to human activities (i.e. documentary and archaeological evidence) on the one hand, compared to all the available natural scientific evidence regarding this period on the other. However, “archives of nature” in this sense should include, for example, sedimentary evidence, too: the paper examines a period of over hundred years, and this period is long enough to apply the available archaeological and, especially, sedimentary evidence, but no such evidence is applied in the paper. To overcome this problem, I would suggest to modify the title so that it more adequately reflects the content (e.g. “a comparison of documentary and high-resolution natural-scientific evidence” or “with tree-ring and stalagmite evidence”; essentially, with including the term “high-resolution”, one does not miss so much any more the testimony of archaeological and sedimentary evidence, although in both cases sometimes it is possible to find high-resolution information).
Another comment reflects on the use of source quality references throughout the paper. For example, although you do not mention (and would be very beneficial to do so), it indirectly becomes clear that you predominantly applied contemporary sources, descriptions of eye-witnesses. It would more clearly demonstrate the quality of the applied source evidence if you reflect on this question in one or two sentences (and some actual examples), in Sect. 2.1, and also in each case study. Moreover, as suggested in the Introduction, you also applied in two cases stalagmite evidence – its brief presentation, however, is missing from Sect. 2.2.
Regarding the Results (Chapter 3), a further comment is related to the case studies: whereas presentation of tree-ring and speleothem evidence always accompanied by very adequate source/literature references, the presentation of documentary evidence usually lacks any direct reference within the text on the sources, but sometimes one-one reference, occasionally, is included. Reference practice should be consequent within the case studies: i.e. please, include the relevant references (i.e. the source reference on drought) also in the description of documentary findings. In relation to the applied source evidence, you provided an Appendix of impressive length on source references. However, now the source references are in a simply alphabetic order, but – as you did not provide the related references within the documentary-based results - it is now difficult to use, and practically impossible to connect to the case studies by simple reading. I would suggest to group the full literature references of the applied sources, provided in the Appendix, to group according to the expeditions, described in Chapter 3. In this way the reader can easily connect the references, now added in Chapter 3, to the actual entries, listed in the Appendix.
Furthermore, in the case studies sometimes such terms are used as “second-hand” or “unreliable” source: this is not enough – in these cases, please, clearly define what you mean, and refer to the source (is it a non-contemporary source? Or contemporary, but only heard the information from eye-witnesses or others?). As your sources predominantly originate from contemporary eye-witnesses (i.e. high source quality evidence), it is also your interest to define those exceptional cases when you consider less reliable evidence.
Another comment concerns the affected regions, the approximate itinerary and locations mentioned in the case studies, based on documentary evidence. While a notable amount of figures and maps are included in the paper from the NADA database there is no map that shows the locations/regions mentioned in the documentary case studies. Please, provide either at least an overview map of the locations, regions and itineraries or include this evidence on the related NADA maps (where it would be anyway beneficial to have some orientation points, it is especially true for the New Mexico NADA maps).
A final comment concerns the Discussion and Conclusion: here it seems that the Discussion contains the conclusions, while in the Conclusion some of the conclusions, already mentioned in the Discussion, are repeated. In the Discussion there is a possibility to discuss certain significant questions, possible contradictions, large-scale comparisons etc. in more details. For example, you raise a rather interesting problem with mentioning the 1770-1772 crisis in Central Europe, and the conclusions on the OWDA-documentary comparison – this question, with using other previous comparisons of OWDA (e.g. Pfister et al., Bauch, Dobrovolny et al. Kiss etc.) – or especially if there is any other comparison with NADA – and documentary evidence, would be a rather interesting and useful topic for further discussion. And, what you now included in the Discussion chapter, could mainly go into the Conclusion.

Some reflections related to the comments of the Reviewers and the Author’s responses:
Reviewer No. 1 addresses many important points that can help improving the manuscript. Dear author, please, take into consideration the suggestions while reworking the paper. Moreover, some of the problems, addressed in the general comments of the reviewer could be rather interesting, further topics for the Discussion chapter. Regarding the highly relevant question of drought definitions, it would be useful to provide a short description of what is considered in the discussed regions as drought, and – if there is any information on that – what drought meant in the referred historical documentation, in the Methodology chapter. This means the inclusion of, for example, the modern understanding what drought is in the relevant areas (i.e. case studies), including the different drought types, and also its historical understanding. Here you could also add what type of dryness or drought-related information types are known from the sources you used in the present work. The interpretation of drought-related documentary information types is of crucial importance to the understanding of the case studies in Sections 3.2.
While the NADA database provides continuous drought-related reconstruction, the available documentary evidence is a fragmentary set of information, an “assembly of case studies”. Therefore, based on this comparison, no general conclusion can be drawn on whether or not in all cases the tree-ring evidence fit the documentary evidence. Based on your answers to the reviewer, I would like to ask, please, take the time, and read after what exactly the NADA reconstruction is based on, what information it provides, and what not. In your paper, and especially in your reflections on the reviews you refer to tree-ring based reconstructions like the NADA and the OWDA as they would provide the ultimate truth about droughts, their timing and intensity, and if the documentary evidence does not “fit” them, then they provide false information. This view also reflects in the conclusion when you are criticizing another work for suggesting a contradiction between documentary evidence and the OWDA reconstruction.
Tree-ring evidence is already in itself dependent on a number of environmental factors (i.e. not only precipitation, and not only of specific seasons), and even if it shows relatively high correlation with precipitation (those which shows) this relationship is usually not consistent – changing in time. Tree ring evidence, when primarily related to precipitation, is not available everywhere: for large regions no any evidence is available at al. To overcome this problem, the NADA, OWDA reconstructions contain extrapolated information, which can provide information even in large areas without any tree-ring evidence; nevertheless, the accuracy of the reconstruction (e.g. how accurately they capture the intensity of drought) might be lower than in the areas with high density of information. Concerning Europe, already a number of studies (not only Collet) suggested clear differences between documentary-based drought reconstruction and the OWDA reconstruction. In most cases these studies discussed periods and areas with very extensive, detailed and case-sensitive documentation – with a coverage much denser than the availability of tree-ring evidence. Thus, on the example of Europe, if the tree-ring based reconstruction shows a somewhat different picture than a documentary-based reconstruction, this does not automatically mean that the documentary evidence is “wrong”. Not talking about the fact that documentary evidence often provides local information, while tree-ring based overviews mainly concentrate on large-scale general information. Naturally, the situation might be different in North America; however, if you think this is the case, you have demonstrate it first. I kindly suggest to look after how drought information is derived from tree-ring evidence, and also how e.g. the NADA reconstruction was made (you can also find there references on individual tree-ring reconstructions, so you can app. even know about the density of available tree-ring evidence in the areas you have discussed), and what “well-verified” information means in that context.
Moreover, it is worth to check some tree-ring based hydroclimate reconstruction papers regarding the studied regions, because these papers also individually refer to the particular season(s) on which the reconstruction primarily reflects. Thus, in this way you can avoid generalisations, and can actually tell in which of your studied regions the tree-rings react more on winter and in which region e.g. on summer precipitation.
Detailed individual comments of the reviewer with page numbers:
I think the reviewer’s suggestions are highly relevant concerning pages 1 and 3, and I find particularly useful and important all the comments related to the Results chapter, particularly the ones reflecting on the case studies.
Regarding the Sections the reviewer suggested to remove: I fully agree that 3.2.3, 3.2.4 should be removed as in the documentary evidence there is no any indication of a drought, and so no comparison can be presented. However, I think you could perhaps utilise some of the information from 3.2.3 in 3.2.2 (as a comparison or additional information?); similarly, some of the information in 3.2.4 could be utilised perhaps in 3.2.5. I agree with the comments regarding 3.2.6: please, provide more details on droughts or if there is no more evidence, why and what extent you connect this “sterility and death” to drought. Please, do consider the suggestions of the reviewer concerning Sections 3.2.7.10.
In general, many of the problems described by the reviewer could be avoided if the author provides more information on the source, not only with referring, but also on how many sources reflect on drought. My general impression is that in most cases probably only one or two short references are available, while in a few cases (esp. towards the end of the study period) “abundant” information is available. Please, give more exact information. What does “abundant” mean? What does each source suggest? If they are too many to describe one by one, you can group them. It would be generally good to get exactly the information included in the sources. Same goes for dating: we only get information on the year when the report comes from – is it possible that no more information is provided in these sources? Because, of course, the season of report (date) would be often highly relevant. Please, when it is possible, do give the exact date of report, and not only the year.
I agree with the reviewer that quotation would also help in the more accurate presentation of the cases. It would also help the reader if you provide a short conclusion on what exactly the sources tell us about the drought, and how certain, based on the documentary evidence, is that really a drought occurred. You provide excellent conclusion in each case concerning the results derived from tree-ring evidence. It would be much easier to compare the two different set of data if interpretation and conclusion would be available on both sides and not only on the tree-ring (or speleothem) side.
Concerning the final conclusions, the reviewer raises a rather interesting question – worth for, for example, another topic to elaborate in the Discussion chapter. I agree with the reviewer that in the Discussion/Conclusion the conclusions are too general and too “overall” compared to the fact that here we are talking about less than a dozen case studies, sometimes with rather sporadic documentary evidence. It does not necessarily means that the conclusions are wrong, but they should be phrased differently, more concentrating on the fact that we are not talking about the conclusions of a 5-6 hundred-year systematic comparison, but only some cases from the 16th and the early 17th centuries.
I find the suggestions of the reviewer regarding a separate section on societal impacts highly relevant: if one of your main goals are to present societal impacts, then this topic must be much more emphasised and systematically discussed within the Results chapter. Additionally, you also found it an important aim to “gain further insights into the seasonality and severity of droughts” – it is true, there are some short reflections on these questions, but especially on seasonality you could provide more information. If society, seasonality and severity detection are major goals, then they should systematically appear (or the statement that no information is available)) in the discussion conclusions of each case studies. Moreover, the major results related to your major goals should as well appear in the final conclusion.

Comments regarding Reviewer No. 2.
Most of the major points, mentioned by the second reviewer, are already addressed in connection with the first review, so I only reflect on questions that differ from the comments of Reviewer No. 1, and/or the second reviewer discussed it in more detail than the first.
Minor points: I agree with both reviewers that, since no any drought was described in documentary evidence in Canada during this period, Canada should be taken out of the referred regions. However, I do not see any problem to keep North America in the title, and I think there is no reason to take Canada off from Table 1. Additionally, in the Introduction I find it important to keep the information that the author indeed studied the sources referring to Canada, but no drought-related evidence was found in contemporary documentary evidence (i.e. “lack of observed drought”) – with the additions suggested by the second reviewer.

Dear Author,

please, take into consideration the minor suggestions of Referee No. 1, and finalise the manuscript.

Thank you and best wishes