Dear Referee,

We would like to thank you for the time and effort put into reviewing the manuscript. This response carefully addresses all your comments (our response is marked with an R and written in italics).

Best regards,

Heidi Kreibich on behalf of all authors

Referee #1

The presented database is the result of a large effort in collecting impact and risk management data of both flood and drought events, more specific: two floods or two droughts that occurred in the same area (Kreibich et al. 2017, 2019). The 45 paired events in the database occurred in 42 different study areas (there are three study areas for which there are two paired events). These basins have different socioeconomic and hydroclimatic contexts, and are spread across all continents. The dataset is unique in covering floods and droughts, in the number of cases assessed and in the amount of qualitative and quantitative socio-hydrological data contained. The dataset underlies the publication on multi flood-drought assessment by Kreibrich et al. (2022)

In order to collect the data, templates were provided to structure the event characteristics using review style of reporting. In addition, semi-quantitative data was collected on: management, hazard, exposure, vulnerability and impacts. The main sources of the data vary from peer-reviewed papers to reports (by governments, NGOs, etc) to newspapers. The campaign started at the EGU General Assembly in April 2019 in Vienna.

The study and the database are unique and I am supporting the publication. The efforts to collect these data must have been a lot of work and the rigorous verification process is convincing. The database sets the scene for new scientific work on multi risk & hazards, for which  empirical data is now lacking.

R: We thank the Referee for the positive feedback. Below, we have responded to all the comments point by point.

I only have some minor edits and suggestions:

The data base is guided by a document which describes the sampling methods and the data processing. This latter part is quite important, because it describes the quality assurance process, using peer review by a core group and cross checking. What I miss here a bit, is that the guiding document does not address the differences in the quality of the sources of the data. One may assume that a peer reviewed paper may provide data which has already undergone a review process, whether a newspaper article does not. I have checked the excel files, and the sources are mentioned. But there is no judgement on the resulting quality ?

R: We have tried to compile as detailed and complete data as possible on the events. As peer reviewed data sources were not available for all indicators of all events, we have also relied on reports, newspaper articles or expert knowledge. We agree that it is useful to provide some judgement about the quality of the various data sources. In the revised version of the manuscript as well as of the data description, we added a personal judgement on the quality of the sources of the data. For instance, in the manuscript in section “2.3.1 Compilation of key data” (lines 287-296): “Our aim was to compile as complete data as possible on the events, but not for all indicators of impacts, hazard, exposure, vulnerability and management shortcomings of all events peer-reviewed data sources were available. Thus, we also resorted to e.g. newspaper articles or expert knowledge. For transparency reasons, and to give data users the opportunity to judge the quality of the data themselves, data source information (citations, references) is also compiled in the key data table. According to our personal assessment, the sources of the data are categorised in descending quality as follows: scientific study (peer reviewed paper and PhD thesis), report (by governments, administrations, NGOs, research organisations, projects), own analysis by authors, based on database (e.g. official statistics, monitoring data such as weather, discharge data, etc.), newspaper article, and expert judgement.”

The role of the references are in the “reference tab” of the key data is not clear: these do not correspond with the references in the PairedEvents document. I assume it is for the key data? However, it misses a link between the key data tab. Using only the current notation of source (“report” or “scientific study”) is not enough to locate which of the references belong to which piece of information. It is also not notated if the source contains information on the first or second paired event. Therefore my recommendation for the authors is to make this link more clear, e.g., in the key data source columns refer to which source contains the data and to include in the references tab whether the reference contains information for the 1st / 2nd / 3rd or all events.

R: Indeed, the ‘reference tab’ of the Key data table contains all the references cited in the ‘key data tab’. Starting from the reference, you can find its citation (and as such the data for which it is the source) in the key data tab by using the ‘search’ within the excel document. This works very well, although it might appear not straightforward. However, our idea was the other way around: we expect that data users/readers start from the key data in the ‘key data tab’, which contain the citations of the data sources. For instance, Paired event with ID1, data column “Management: Problems with water management infrastructure”, event 2 (this is cell F3 in the key data tab) contains the citation: “(Zhang et al. 2017)”. To find the full reference to this citation, you go to the reference tab, look under ‘ID of Paired event: 1’ and find it in line 5 (last line of the ID1 block as the references are sorted alphabetically per paired event). This is described in section 3.2.2 Key data table of the manuscript; however, we realise that the header of the columns in the key data tab were misleading. We have now changed the header of every second column from “source of data” to “category of data source”. Also, the description of the key data table in the data description file has been improved accordingly. We refrain from adding the event year to the references tab because the listed references refer to particular event characteristics and thus are context specific. We are convinced that the access to references should start from the context information which is given in the key data tab. We will include a README tab into the Key data table, to even better inform the data user how to use the key data table.  

The templates / attachments are missing in the index, and are not referred to in the document. Section 3.1 mentions that there are templates, but it would be useful to refer here to the templates in the document itself.

R: In the revised version of the data description, we have added the Appendix with the template to the “Table of contents”. Additionally, we refer to the Appendix in section 3.1 where the templates are mentioned. The provision of the templates in the Appendix of the data description is now also mentioned more clearly in the manuscript (section 2.2 Data acquisition, line 244 and in the conclusions lines 520-521)

Dear Referee,

we would like to thank you for the time and effort put into reviewing the manuscript. This response addresses your comment and includes the revised wording of our manuscript.

Best regards,

Heidi Kreibich on behalf of all authors

Referee #2

This work provides a large database concerning a number of paired flood and drought events occurred worldwide across several decades. It is the result of significant efforts and commitment of a large community of experts working together within the framework of the Panta Rhei Research Initiative of the International Association of Hydrological Sciences (IAHS) (Working Groups “Changes in flood risk” and “Droughts in the Anthropocene”).

The result is a unique database that may sustain further investigations of the community and I certainly support its publication. The value of the work should also be recognized considering that it attempts to compare, across time and space, events that are, for their nature and for the different contexts/condition on which they occur, unavoidablement different. This appears evident looking at the descriptions of the events (for example, events diverge in terms of the magnitude of the precipitation, initial basin conditions, extent of levee failure, tidal elevation, etc). In my opinion, this aspect represents both the strength and the limit of the database.

The attempt to provide a rational way to compare such events represents the most challenging part of the work. Although the applied approach (i.e., quality assurance procedures, etc.) is convincing and represents the best way of acting, I still be doubtful regarding the possibility to drive conclusions on risk and impact patterns, or on the effectiveness of management solutions, when boundary conditions and characteristics of the two compared events are different. Thus, my only request is to made an additional effort (perhaps on section 4?) in trying to emphasize this aspect, stressing the limit of the current database and thus promoting caution on the interpretations of the outcomes.

R: The following explanation is now included in Section 4 lines 453-460: “While the variables describing the first and second event of a pair are comparable, variables and data quality differ strongly between the paired events. The great heterogeneity of data and events represents both the strength and the weakness of the Panta Rhei dataset with regard to comparative analyses. As quantitative comparative analyses across all paired events are impossible, such analyses can only be undertaken on the basis of the indicators-of-change. Although these indicators were created with great care according to the quality assurance protocol, they are subject to uncertainties and caution is required when interpreting the results.”