Multi-Indicator Drought Variability in Europe (1766–2018)

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Ionita et al. present a comprehensive multi-indicator reconstruction of European droughts (PDSI, SPEI, and SMI) from 1766 to 2018. The topic is timely and scientifically relevant. However, the manuscript would benefit from improved clarity in objectives, enhanced figure quality, a more quantitative comparison among indicators, and a stronger discussion connecting the results to climate dynamics and previous studies. Therefore, I recommend a Major revision. Below are point-wise detailed comments.

Title

The title is informative but too long; consider shortening it to “Reconstructing Multi-Indicator Drought Variability in Europe (1766-2018)” to emphasise the multi-indicator approach.

Abstract

• The abstract is well-structured but somewhat descriptive. Add quantitative metrics, such as specifying correlation coefficients between indicators or percentages of drought area changes.
• Clarify the novelty: highlight that this is the first simultaneous reconstruction comparison of PDSI, SPEI, and SMI across Europe.
• The abstract should mention key findings about atmospheric drivers (CCA results).

Introduction

• The first two paragraphs provide an excellent context for drought impacts, but they are repetitive and offer brief statements about agriculture and ecosystems; instead, they should focus more on research gaps.
• Lines 66-70: Emphasise the data limitation gap, lack of long-term, spatially consistent datasets, as the main reason for this study.
• Line 80. The debate section is strong, but it should include a concise paragraph summarising why different drought indicators yield divergent conclusions.
• The objectives are clearly listed; however, merge similar objectives (iii-v) to avoid redundancy. Mention explicitly that temporal consistency and cross-validation among indicators were analysed.
• Lines 124-128. No need.
• At the end, add one line connecting this study with climate teleconnections (NAO, AMO) and how they may influence drought patterns; this will enhance the scientific depth.

Data and Methods

• Line 143: Too much background on PDSI; actually, it’s the intro part. Here, summarise the advantages/disadvantages in bullet form or a concise paragraph.
• Lines 181–182, For SPEI, mention explicitly how stable oxygen isotopes from tree rings were calibrated with observed data.
• Lines 182-183 provide the spatial resolution of the SMI model outputs and validation R² values.
• Line 208: The trend analysis briefly introduces the threshold for significance (e.g., p < 0.05) and explains why the modified Mann-Kendall test was chosen over the original.
• Line 219 Canonical correlation analysis-include the percentage of variance explained by each mode and justify using the first 10 EOFs; also state the software/tool used (e.g., MATLAB, Python).

Results

• Line 327. Explain why 10-year windows were chosen instead of 5- or 15-year windows, and provide a literature reference.
• After line 391, add a short interpretation on how SPEI’s temperature sensitivity affects recent decades’ dryness detection.
• Figures 5-7: Improve readability by making the legend fonts larger. Add a north arrow and scale bar.
• Lines 428-429, The CCA section is strong; however, mention which atmospheric pattern corresponds to each mode (e.g., NAO, East Atlantic pattern). Add correlation coefficient tables.

Discussion

• Generally, the discussion is well summarised but highly descriptive. Include more quantitative comparisons between indicators (e.g., “PDSI and SMI correlated at r = 0.78, p < 0.01”).
• Lines 545-553, Discuss potential uncertainties due to different calibration periods and proxy availability.
• Lines 572-573, Excellent statement about “cold droughts”; add supporting references from 2022-2025 (e.g., Nature Climate Change, Climate Dynamics).
• Line 578 presents a final paragraph that briefly connects drought variability to large-scale climate modes (NAO, AMO, ENSO).

Conclusions

• The conclusions are clear but lengthy. Condense to 4–5 strong, outcome-driven statements focusing on: Major findings (indicator agreement/divergence), Policy implications, Future research directions.

 

• Lines 619-624: The phrase “not categorically unprecedented” is excellent; emphasise this finding as the key contribution.
• Add a final line connecting this work to climate-resilient water management strategies.

Author Response

Ionita et al. present a comprehensive multi-indicator reconstruction of European droughts (PDSI, SPEI, and SMI) from 1766 to 2018. The topic is timely and scientifically relevant. However, the manuscript would benefit from improved clarity in objectives, enhanced figure quality, a more quantitative comparison among indicators, and a stronger discussion connecting the results to climate dynamics and previous studies. Therefore, I recommend a Major revision. Below are point-wise detailed comments.

Response: We want to thank the reviewer for the appreciation/suggestions/comments/feedback that will help us improve our manuscript, and for taking the time to read and review our paper. Please find the detailed responses below and the corresponding revisions/corrections highlighted in red in the re-submitted files.

 

Title

The title is informative but too long; consider shortening it to “Reconstructing Multi-Indicator Drought Variability in Europe (1766-2018)” to emphasise the multi-indicator approach.

Response: Thank you for the suggestion. We modified the title according to the reviewer suggestion, be we had to remove the word “Reconstruction”, as we used already published data, and we did not produce by ourselves a reconstruction.

Abstract

• The abstract is well-structured but somewhat descriptive. Add quantitative metrics, such as specifying correlation coefficients between indicators or percentages of drought area changes.

Response: Thank you for the suggestion. We improved the abstract in the revised version of the manuscript, and we included interannual magnitudes/areas (PDSI 1874; SPEI 2003; SMI 1868), decadal statistics (PDSI 1941–1950; SPEI 2011–2018), and area-trend estimates at European and regional scales.

• Clarify the novelty: highlight that this is the first simultaneous reconstruction :

Response: Thank you for the suggestion. We improved the abstract in the revised version of the manuscript, and explicitly stated that the simultaneous analysis over 1766–2018 across three independent reconstructions.

• The abstract should mention key findings about atmospheric drivers (CCA results).

Response: Thank you for the suggestion. We improved the abstract in the revised version of the manuscript, and we included the key findings about atmospheric drivers (CCA results).

 

Introduction

• The first two paragraphs provide an excellent context for drought impacts, but they are repetitive and offer brief statements about agriculture and ecosystems; instead, they should focus more on research gaps.

Response: Thank you for the suggestion. We improved the introduction in the revised version of the manuscript, and we modified and improved the first two paragraph in order to avoind the repetition and to focus more on the research gaps. Lines 40-52

• Lines 66-70: Emphasise the data limitation gap, lack of long-term, spatially consistent datasets, as the main reason for this study.

Response: Thank you for the suggestion. We modified the lines according to the reviewer's suggestion and emphasized the data limitation gap. Lines 63-70.

• Line 80. The debate section is strong, but it should include a concise paragraph summarising why different drought indicators yield divergent conclusions.

Response: Thank you for the suggestion. We improved this section and moved it to the discussion section as suggested by one reviewer.

• The objectives are clearly listed; however, merge similar objectives (iii-v) to avoid redundancy. Mention explicitly that temporal consistency and cross-validation among indicators were analysed.

Response: Thank you for the suggestion. The objectives were improved according to the reviewer's suggestion.

• Lines 124-128. No need.

Response: We removed the mention lines.

• At the end, add one line connecting this study with climate teleconnections (NAO, AMO) and how they may influence drought patterns; this will enhance the scientific depth.

Response: Thank you for the suggestion. Our CCA was designed to extract summer large-scale circulation patterns linked to drought variability, and the resulting Z500 modes do not project cleanly onto canonical teleconnections (e.g., NAO, EA); to avoid over-attribution, we therefore did not invoke named patterns.

Data and Methods

• Line 143: Too much background on PDSI; actually, it’s the intro part. Here, summarise the advantages/disadvantages in bullet form or a concise paragraph.

Response: Thank you for the suggestion. We modified the text accordingly and added a sentence to summarise the advantages/disadvantages of the PDSI index. Lines 113-119.

• Lines 181–182, For SPEI, mention explicitly how stable oxygen isotopes from tree rings were calibrated with observed data.

Response: We appreciate your suggestion. However, the methodology how stable oxygen isotopes from tree rings were calibrated with observed data has been fully documented in the original publication where the proxy record was first presented.

 

• Lines 182-183 provide the spatial resolution of the SMI model outputs and validation R² values.

Response: The resolution of the SMI model output is 0.5° x o.5°. The statistics of the SMI model output are already discussed in Rakovec et al., (2022) and is beyond the scope of our paper.

https://agupubs.onlinelibrary.wiley.com/doi/full/10.1029/2021EF002394

 

• Line 208: The trend analysis briefly introduces the threshold for significance (e.g., p < 0.05) and explains why the modified Mann-Kendall test was chosen over the original.

Response: We appreciate your suggestion. We will state explicitly that two-sided tests at α = 0.05 are used to assess significance (reporting p < 0.05 p-values). We chose the modified Mann–Kendall (Hamed & Rao; Hamed) because hydroclimatic series exhibit serial correlation and ties that violate the original MK independence assumption; the modification adjusts the MK variance without prewhitening, thereby controlling Type-I error while preserving the trend signal.

 

• Line 219 Canonical correlation analysis-include the percentage of variance explained by each mode and justify using the first 10 EOFs; also state the software/tool used (e.g., MATLAB, Python).

Response: Thank you for the suggestion. In the revised version of the manuscript, we improved this section according to the reviewer's suggestions. All the analyses were made using MATLAB, but we are not allowed to specify this in the manuscript text.

Results

• Line 327. Explain why 10-year windows were chosen instead of 5- or 15-year windows, and provide a literature reference.

Response: Explanation for chosen 10-years window: The choice for a 10-year average was motivated by previous studies, which have shown that multi-year droughts most of the time occur in clusters of 4 up to 10 consecutive years [28,33,60,61] and it also allows us to make a proper comparison between the different drought indicators at decadal time scale. Lines 304-307.

[https://doi.org/10.1038/s43247-022-00648-7

https://doi.org/10.1029/2021EF002394.

https://doi.org/10.1038/s41612-020-00153-8.

https://doi.org/10.2166/nh.2012.024 ]

• After line 391, add a short interpretation on how SPEI’s temperature sensitivity affects recent decades’ dryness detection.

Response: Thank you for the suggestion. We modified the text accordingly and added a sentence with explanations. Lines 352-358.

• Figures 5-7: Improve readability by making the legend fonts larger. Add a north arrow and scale bar.

Response: Modified as suggested.

• Lines 428-429, The CCA section is strong; however, mention which atmospheric pattern corresponds to each mode (e.g., NAO, East Atlantic pattern). Add correlation coefficient tables.

Response: We do not mention any well-known teleconnection pattern. The role of the CCA analysis is to identify the large-scale atmospheric circulation patterns associated with drought variability, which can be independent of NAO and EA. None of our CCA spatial patterns, in terms of Z500, doesn’t resemble any well-known teleconnection pattern. Moreover, we are performing the CCA analysis on the summer season, when NAO, EA and other pre-defined teleconnections are much weaker, which can be one reason for the fact that the CCA doesn’t capture any of them.

 

Discussion

• Generally, the discussion is well summarised but highly descriptive. Include more quantitative comparisons between indicators (e.g., “PDSI and SMI correlated at r = 0.78, p < 0.01”).

Response: We want to thank the reviewer for the appreciation of our manuscript and for your suggestion. In the revised version, we improved the discussion section of our manuscript according to the reviewer's suggestion.

• Lines 545-553, Discuss potential uncertainties due to different calibration periods and proxy availability.

Response: Thank you for the suggestion. We modified the text accordingly and added new explanations as suggested. Lines 547-564.

• Lines 572-573, Excellent statement about “cold droughts”; add supporting references from 2022-2025 (e.g., Nature Climate Change, Climate Dynamics).

Response: Thank you for your appreciation and the suggestion. We already cited some papers regarding the cold droughts. If the reviewer knows other relevant papers on this topic, please let us know their DOI number, and we will check them out and cite them accordingly.  

• Line 578 presents a final paragraph that briefly connects drought variability to large-scale climate modes (NAO, AMO, ENSO).

Response: Thank you for the suggestion. Our CCA is explicitly summer-focused and yields Z500 patterns that do not project cleanly onto canonical modes (e.g., NAO/EA), consistent with the weaker summertime expression of these teleconnections; we therefore avoided over-attribution.

Conclusions

• The conclusions are clear but lengthy. Condense to 4–5 strong, outcome-driven statements focusing on: Major findings (indicator agreement/divergence), Policy implications, Future research directions.

Response: Thank you for the suggestion. We modified the conclusion in the revised version of the manuscript according to the reviewers' suggestions. Lines 596-634.

 

• Lines 619-624: The phrase “not categorically unprecedented” is excellent; emphasise this finding as the key contribution.

Response: Thank you for the suggestion. We modified the conclusion in the revised version of the manuscript according to the reviewers' suggestions and emphasised this finding as the key contribution. Lines 596-634..

 

• Add a final line connecting this work to climate-resilient water management strategies.

Response: Thank you for the suggestion. We modified the conclusion in the revised version of the manuscript according to the reviewers' suggestions and added a line to connect this work to climate-resilient water management strategies. Lines 596-634.

 

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

Summary:

This study evaluated three drought indices (PDSI, SPEI, and SMI) in the context of drought events in Europe extending back to the 1700s. Each index was found to have varying effectiveness in quantifying the drought events. Atmospheric circulation drivers were also examined. Overall, I think this is a solid study that will be important for research and climate assessments around drought. There were a few areas of the manuscript that I felt required some additional clarification concerning the methods. Because some of my questions involve the methodology, I think the paper should be reviewed again after revision. I outline my specific comments and questions below.

Comments/questions:

Concerning the methods, there seem to be elements in this section that are either missing or could use some additional explanation. These include: 1) How were the regional averages computed? How was the drought area detection analysis conducted? How was the percentile analysis conducted? Over what periods were the drought indices computed and examined (i.e. monthly, seasonally, annually)? If drought indices were computed annually (?) then why compare them with 500hPa heights only for JJA and not the full corresponding period?

Figure 4: I suggest flagging the grid points in each panel that met the “drought” criteria as it will be easier to compare with Figure 3. Please describe the units of the data shown in the panels in the caption.

Figure 5: How was drought area defined here? Was the average index values just for the drought area or for all of Europe?

What is the “drought” data referred to panel (e) of Figure 8-10.

Some limitations to consider for the discussion and elsewhere: 1) The time-scale of the drought event vs. the time-scale of the indicator variable. Would a drought over just two seasons be well captured in an annual index? Some might be better at that than others. It might be worth also describing and citing the specific drought events here or earlier in the Introduction to better provide context as well. 2) The atmospheric driver analysis is challenged somewhat using the Valler et al. (2022) data set that features a long time series but much higher uncertainty than a standard reanalysis like ERA5. It might be worthwhile comparing events since 1940 with ERA5 (or similar) reanalysis to assess consistency.

 

Author Response

This study evaluated three drought indices (PDSI, SPEI, and SMI) in the context of drought events in Europe extending back to the 1700s. Each index was found to have varying effectiveness in quantifying the drought events. Atmospheric circulation drivers were also examined. Overall, I think this is a solid study that will be important for research and climate assessments around drought. There were a few areas of the manuscript that I felt required some additional clarification concerning the methods. Because some of my questions involve the methodology, I think the paper should be reviewed again after revision. I outline my specific comments and questions below.

Response: We want to thank the reviewer for the appreciation/suggestions/comments/feedback that will help us improve our manuscript, and for taking the time to read and review our paper. Please find the detailed responses below and the corresponding revisions/corrections highlighted in red in the re-submitted files.

 

Comments/questions:

Concerning the methods, there seem to be elements in this section that are either missing or could use some additional explanation. These include: 1) How were the regional averages computed? How was the drought area detection analysis conducted? How was the percentile analysis conducted? Over what periods were the drought indices computed and examined (i.e. monthly, seasonally, annually)? If drought indices were computed annually (?) then why compare them with 500hPa heights only for JJA and not the full corresponding period?

Response: Thank you for pointing this out—we will clarify these items in Methods. Regional means were computed as area-weighted (cos φ) averages of all land grid cells within MED, CEU, and NEU masks. Drought area was quantified per year and indicator as the fraction of grid cells below a region-specific Q10 threshold, where Q10 is the empirical 10th percentile of that indicator’s distribution over the common baseline (1766–2008) for each region and for Europe; exact Q10 values and series are reported in Table 1 and corresponding figures. The percentile analysis therefore uses region-wise empirical distributions, and extremes are summarized by event magnitude (M) and areal coverage (A), with decadal statistics computed as arithmetic means over calendar decades. Regarding temporal aggregation, all reconstructions used here are summer-targeted: OWDA scPDSI (JJA), SPEI2 for July–August (2-month), and SMI aggregated to JJA, analyzed at interannual and decadal scales. Consequently, the circulation analysis uses JJA 500-hPa heights to match the seasonality of the drought fields;

 

Figure 4: I suggest flagging the grid points in each panel that met the “drought” criteria as it will be easier to compare with Figure 3. Please describe the units of the data shown in the panels in the caption.

Response: We agree with this comment, but due to the fact that we are dealing with different indicators, which define drought based on different thresholds, especially SMI, is rather difficult to have a new figure with the suggested modifications. Nevertheless, to make it clearer for the reader, we added in the figure caption what the observed colors mean.

 

Figure 5: How was drought area defined here? Was the average index values just for the drought area or for all of Europe?

Response: We added the definition of drought area in the Figure caption of Figure 5.

 

What is the “drought” data referred to panel (e) of Figure 8-10.

Response: Thank you for the suggestion. In the revised text of our manuscript, we improved the figure captions of figures 8-10 to make them clearer about what is represented in the figures.

 

Some limitations to consider for the discussion and elsewhere: 1) The time-scale of the drought event vs. the time-scale of the indicator variable. Would a drought over just two seasons be well captured in an annual index? Some might be better at that than others. It might be worth also describing and citing the specific drought events here or earlier in the Introduction to better provide context as well. 2) The atmospheric driver analysis is challenged somewhat using the Valler et al. (2022) data set that features a long time series but much higher uncertainty than a standard reanalysis like ERA5. It might be worthwhile comparing events since 1940 with ERA5 (or similar) reanalysis to assess consistency.

Response: Thank you for the suggestion. We revised our manuscript according to the reviewers’ suggestions, and this is an improved version of our manuscript. Thank you for raising the timescale issue. To clarify, all drought indicators in our study are defined for the summer season, not as annual indices. We have revised the Methods to state this explicitly and added citations.  We agree that ERA5 is the preferred benchmark after 1940; however, our aim is to diagnose multi-century variability and circulation drivers, which necessitates a product spanning the pre-instrumental era. We therefore used the EKF400v2 paleo-reanalysis (Valler et al., 2022) despite its larger uncertainties, because it enables a consistent, long-period analysis that ERA5 cannot provide alone.

 

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

In this study, historical and recent drought patterns across Europe were investigated using three complementary drought indicators of PDSI, SPEI, and SMI for a long period were investigated. The analysis covers the years 1766-2018 and includes short-term and multi-year drought events. Although the topic is interesting, there are several aspects that should be revised to improve the technical quality of the manuscript.  Some of my comments related to the manuscript are listed as follows;

Three different indices were used to examine the most severe drought in Europe in both temporal and spatial terms. However, these indices show different periods as the driest for Europe. It is also mentioned that in smaller areas, the indices provide more consistent results compared to the results for the entire continent. Thus, the title of the study does not fully reflect its content.

The objective of the research is not presented clearly or in sufficient detail, which weakens the overall novelty of the study.

Each index provides information about drought changes from a different point of view. Although the authors state that  these perspectives are combined, in reality, the indices are only presented together rather than being integrated. There is no more information related to the mathematical combination and use of their individual strengths.

The drought indicators used in the study are PDSI, SPEI, and SMI. In the introduction, only the abbreviations are provided without their full form. Additionally, Figure 1 is also presented without explaining how these indices are calculated, their value ranges, or the variables they depend on.

Before presenting Figure 1, the authors attempt to show that there are different interpretations in the literature about past drought events. To support this, they compare three indices to show that they produce different results when identifying whether recent droughts are really unusual and which regions were mostly affected. However, since the scientific background of the indices is not explained, these visuals are somewhat confusing. To better support the discussions, these explanations could be moved to the Discussion section with appropriate references.

In line 191, it is stated that: “Both indices have a 0.5 × 0.5 spatial resolution and cover the period 1901–2021.” In this explanation, the metric unit of spatial resolution should be clearly specified.

Between lines 182 and 188, the SMI (Soil Moisture Index) used in the study is defined. However, as there are several definitions of SMI in the literature, this paragraph should be revised for clarity.

Between lines 190 and 230, many references and methods are mentioned, but it is unclear which method was used for what purpose. The authors should clarify the role of each method mentioned in this chapter. Additionally, this chapter is much shorter than the other sections. Thus, expanding this section is recommended.

The information about the data used to calculate the indices over a long period is also insufficient. The data sources, values, and details of how they were used are only briefly mentioned. This part should be expanded with more detailed explanations.

The PDSI, SPEI, and SMI indices identify different periods as the driest in Europe, and the spatial patterns presented in the maps also show differences. The authors consider this a positive aspect, suggesting that multiple indicators complement each other. However, if different indices are used for the same purpose, one should be able to reflect the “true” drought conditions better or worse. Large discrepancies among them raise questions about their reliability. The authors might consider incorporating satellite-based indices, at least for the recent period, to better support the results.

The results section should also be strengthened. There are too many explanations in parentheses and too many numerical details, which make the manuscript confusing. If so many numbers are presented in the text, the added value of the tables and figures becomes unclear.

The study mainly concentrates on historical drought events but provides only a limited discussion of future projections under climate change. In addition, some of the figures and explanations are presented in a highly technical manner, which may make them difficult for readers without specialized knowledge to follow the content.

Author Response

In this study, historical and recent drought patterns across Europe were investigated using three complementary drought indicators of PDSI, SPEI, and SMI for a long period were investigated. The analysis covers the years 1766-2018 and includes short-term and multi-year drought events. Although the topic is interesting, there are several aspects that should be revised to improve the technical quality of the manuscript.  Some of my comments related to the manuscript are listed as follows;

Response: We want to thank the reviewer for the appreciation/suggestions/comments/feedback that will help us improve our manuscript, and for taking the time to read and review our paper. Please find the detailed responses below and the corresponding revisions/corrections highlighted in red in the re-submitted files.

 

Three different indices were used to examine the most severe drought in Europe in both temporal and spatial terms. However, these indices show different periods as the driest for Europe. It is also mentioned that in smaller areas, the indices provide more consistent results compared to the results for the entire continent. Thus, the title of the study does not fully reflect its content.

Response: Thank you for the suggestion. We modified the title of the manuscript. We decided to use the title suggested by Reviewer No. 1.

 

The objective of the research is not presented clearly or in sufficient detail, which weakens the overall novelty of the study.

Response: Thank you for the suggestion. The objectives were improved according to the reviewer's suggestion. Lines 90-95.

 

Each index provides information about drought changes from a different point of view. Although the authors state that  these perspectives are combined, in reality, the indices are only presented together rather than being integrated. There is no more information related to the mathematical combination and use of their individual strengths.

Response: Thank you for this constructive point. Our intention with “integration” was a comparative, cross-validated synthesis rather than a mathematical fusion of indices; a formal merger is non-trivial because PDSI, SPEI2, and SMI target different processes and time scales, use different calibrations, and therefore are not commensurate in a single metric without imposing ad-hoc weights that risk false precision. To address your concern and make the synthesis more explicit, we revise wording (“simultaneous, cross-validated analysis” rather than “integrated”) and highlighted that our CCA isolates a common circulation-linked component across indicators, which functions as an independent consistency check.

The drought indicators used in the study are PDSI, SPEI, and SMI. In the introduction, only the abbreviations are provided without their full form. Additionally, Figure 1 is also presented without explaining how these indices are calculated, their value ranges, or the variables they depend on.

Response: Thank you for the helpful suggestion. We defined the indices in the introduction at their first mention. Our manuscript is a secondary analysis of three previously published drought reconstructions—Palmer Drought Severity Index (PDSI) [31], summer Standardized Precipitation–Evapotranspiration Index (SPEI2) [28], and Soil Moisture Index (SMI) [32]. Because we analyze these published reconstructions rather than recompute the indices, it would be inappropriate to integrate detailed calculation methods from the three source papers here; those methodological descriptions (algorithms, parameterizations, calibration choices, and uncertainties) properly belong in—and are fully documented by—the original publications [31,28,32].

 

Before presenting Figure 1, the authors attempt to show that there are different interpretations in the literature about past drought events. To support this, they compare three indices to show that they produce different results when identifying whether recent droughts are really unusual and which regions were mostly affected. However, since the scientific background of the indices is not explained, these visuals are somewhat confusing. To better support the discussions, these explanations could be moved to the Discussion section with appropriate references.

Response: Thank you for the suggestion. We moved these explanations to the Discussion section. Lines 454-490.

 

In line 191, it is stated that: “Both indices have a 0.5 × 0.5 spatial resolution and cover the period 1901–2021.” In this explanation, the metric unit of spatial resolution should be clearly specified.

Response: Thank you for the suggestion. In the revised version of the manuscript, we corrected the misspellings.

 

Between lines 182 and 188, the SMI (Soil Moisture Index) used in the study is defined. However, as there are several definitions of SMI in the literature, this paragraph should be revised for clarity.

Response: Thank you for the suggestion. In the revised version of the manuscript, we added a sentence to explain this index. Lines 158-160.

 

Between lines 190 and 230, many references and methods are mentioned, but it is unclear which method was used for what purpose. The authors should clarify the role of each method mentioned in this chapter. Additionally, this chapter is much shorter than the other sections. Thus, expanding this section is recommended.

Response: Thank you for the suggestion. In the revised version of the manuscript, we divided this section into 2 subsections in order to make a better difference between the data used and the methods. Also, the methods subsection was improved according to the reviewers' suggestions. Lines 188-216.

 

The information about the data used to calculate the indices over a long period is also insufficient. The data sources, values, and details of how they were used are only briefly mentioned. This part should be expanded with more detailed explanations.

Response: Thank you for the suggestion. Our manuscript does not recompute drought indices; it analyzes three reconstructions already published in separate papers (PDSI/OWDA, summer SPEI2, and SMI). Reproducing the full derivation and calibration procedures from those source studies would duplicate their methods and is outside the scope of this secondary analysis. That said, we agree the data provenance and usage need to be clearer for readers, and we improved the Data and Methods section in the revised version of the manuscript.

 

The PDSI, SPEI, and SMI indices identify different periods as the driest in Europe, and the spatial patterns presented in the maps also show differences. The authors consider this a positive aspect, suggesting that multiple indicators complement each other. However, if different indices are used for the same purpose, one should be able to reflect the “true” drought conditions better or worse. Large discrepancies among them raise questions about their reliability. The authors might consider incorporating satellite-based indices, at least for the recent period, to better support the results.

Response: Thank you for your insightful comment highlighting the differences among the PDSI, SPEI, and SMI drought indices. We agree that the discrepancies in identifying drought severity and spatial patterns raise important questions regarding the varying sensitivities and methodological differences of each indicator. Our intention in presenting multiple indices was to capture the multidimensional nature of drought, acknowledging that no single index perfectly represents all aspects of drought conditions. While we appreciate the suggestion to incorporate satellite-based indices to support the analysis, currently, the scope of our study focuses on long-term reconstructions derived from proxy and meteorological data to provide a century-scale context. 

 

The results section should also be strengthened. There are too many explanations in parentheses and too many numerical details, which make the manuscript confusing. If so many numbers are presented in the text, the added value of the tables and figures becomes unclear.

Response: Thank you for the suggestion. In the revised version of the manuscript, we improved the results section and made it clearer and easier to read. We also made sure that the data presented in the tables is presented in the main text too.

 

The study mainly concentrates on historical drought events but provides only a limited discussion of future projections under climate change. In addition, some of the figures and explanations are presented in a highly technical manner, which may make them difficult for readers without specialized knowledge to follow the content.

Response: Thank you for the suggestion. In the revised version of the manuscript, we improved the discussion section according to the reviewers' suggestions. Our scope is a multi-century analysis of published reconstructions, so full projection work is beyond this study; however, we  added a concise paragraph linking our findings to recent projection evidence and policy-relevant risks.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

The authors have satisfactorily addressed all my concerns; therefore, I recommend the paper for publication.

Author Response

We want to thank the reviewer for the appreciation/suggestions/comments/feedback that will help us improve our manuscript. 

Reviewer 2 Report

Comments and Suggestions for Authors

I think that authors did a generally good job address my review questions. However, I did not find revisions to the methods section in accordance with the points in your response to my first comment. I suggest making sure to fully address this so that readers clearly understand how those tasks/quantities were computed for your study.

Author Response

I think that authors did a generally good job address my review questions. However, I did not find revisions to the methods section in accordance with the points in your response to my first comment. I suggest making sure to fully address this so that readers clearly understand how those tasks/quantities were computed for your study.

 

Response: We want to thank the reviewer for the appreciation/suggestions/comments/feedback that will help us improve our manuscript. Please find the detailed responses below and the corresponding revisions/corrections highlighted in red in the re-submitted files.

Response: Thank you for the suggestion. In the revised version of the manuscript, we have now added how each indicator has been computed, but also to new sub-sections, namely: Regional and European mean indices and Percent area under drought (A). We hope by adding this information the text is now more clear and easier to follow.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

Thanks to the the authors for their efforts. Textual edits were performed by taking some of the criticisms into consideration. However, a few points were overlooked due to the purpose and scope of the study. The first of these is that the mathematical backgrounds of the indexes used are not presented.

*Although it is stated that indexes used in the study have been previously proposed by other researchers and relevant references are provided, if the numerical values ​​that enable the creation of the graphs in the study come from these indexes, and if the maps are thematic presentations of the index values, how these indexes were calculated and what each parameter represents should also be stated. If there are very long and complex step-by-step formulations at this point, the final formulation of each index can be presented with explanations. There is no strict rule that the mathematical equation of an index can only be included in the reference publication. On the contrary, presentation by the researcher using it is a requirement rather than a preference.

*While the article has been edited throughout, the presentation of the results section, in particular, remains quite confusing. Too many jumbled sentences, too many parenthetical explanations, etc.

*Why is the SMI graph different in appearance from the PDSI and SPEI graphs in Figure 6?

*In the absence of any innovative suggestions or approaches, the authors should reconsider the necessity of so many figures/graphs to explain why three different indexes give significantly different or partially similar results depending on the size of the region in terms of the most severe drought.

Author Response

Thanks to the authors for their efforts. Textual edits were performed by taking some of the criticisms into consideration. However, a few points were overlooked due to the purpose and scope of the study. The first of these is that the mathematical backgrounds of the indexes used are not presented.

Response: We want to thank the reviewer for the appreciation/suggestions/comments/feedback that will help us improve our manuscript. Please find the detailed responses below and the corresponding revisions/corrections highlighted in red in the re-submitted files.

Response: Thank you for the suggestion. In the revised version of the manuscript, we have now added how each indicator has been computed, but also to new sub-sections, namely: Regional and European mean indices and Percent area under drought (A). We hope by adding this information the text is now more clear and easier to follow.

*Although it is stated that indexes used in the study have been previously proposed by other researchers and relevant references are provided, if the numerical values ​​that enable the creation of the graphs in the study come from these indexes, and if the maps are thematic presentations of the index values, how these indexes were calculated and what each parameter represents should also be stated. If there are very long and complex step-by-step formulations at this point, the final formulation of each index can be presented with explanations. There is no strict rule that the mathematical equation of an index can only be included in the reference publication. On the contrary, presentation by the researcher using it is a requirement rather than a preference.

Response: In the revised version of the manuscript we added, for each indicator, how the index is computed and the data set used to compute it. See the new sub-sections: Regional and European mean indices and Percent area under drought (A).

*While the article has been edited throughout, the presentation of the results section, in particular, remains quite confusing. Too many jumbled sentences, too many parenthetical explanations, etc.

Response: We have strengthened the Results section in the revised manuscript. Nonetheless, because we incorporated extensive requests from more than six reviewers over a prolonged review process, we adopted balanced compromises where comments diverged.

*Why is the SMI graph different in appearance from the PDSI and SPEI graphs in Figure 6?

Response: Because SMI has only positive values and if we use bars like for PDSI and SPEI it is very difficult to compare them, since both PDSI and SPEI have both positive and negative values. We have tested different approached to compare them, and this combination was the one that was giving the proper visual results for what we wanted to show.

*In the absence of any innovative suggestions or approaches, the authors should reconsider the necessity of so many figures/graphs to explain why three different indexes give significantly different or partially similar results depending on the size of the region in terms of the most severe drought.

Response: We do not agree with this comment. In the other papers, the authors focused their analysis mainly on the reconstruction of the three indices and on some special events like 2018. In our study we go beyond this analysis and try to identify long-lasting drought events over the past *250 years, but also emphasizing the need of a multi-proxy approach for future studies, especially when one wants to emphasize the “extremeness” of recent drought events. We wanted to emphasize that this approach is rather subjective, depending on the advantages/disadvantages of different drought indicators. None of the data/studies used in our study hasn’t properly discussed these issues and we strongly believe that there is a need for such discussions. Thus, we believe that each figure is necessary.

Author Response File: Author Response.pdf