Review Reports

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The manuscript builds monthly SSI (standardized soil moisture index) for the Ebro Basin (2010–2023) by downscaling SMOS surface soil moisture with DISPATCh to 1 km using MODIS LST/NDVI, then compares SSI against station-based SPI and a 1.1 km gridded SPI. It finds broadly positive SSI–SPI correlations (best at mid/long integration windows), frequent “k vs. k+1” scale alignment (SSI-k matches SPI-(k+1)), and spatial divergences linked to land cover/management (dense forests, urban areas, rice-paddy irrigation, snow). The study argues SSI is operationally useful where in-situ data are scarce.

Major comments

1. Sensing depth / representativeness not specified
   The manuscript repeatedly refers to surface soil moisture from SMOS and its DISPATCh downscaling to 1 km, but it never states the effective physical sensing depth (e.g., ~top few cm) or how the 1 km SSI should be interpreted versus root-zone conditions. Please (a) state the assumed depth of the microwave retrieval; (b) discuss how DISPATCh affects vertical representativeness; and (c) caveat interpretation in snow/forest/irrigated areas where surface–root-zone coupling can decouple. See Section 3.1 description of SMOS “surface soil moisture” and DISPATCh derivation.
2. Land-cover impacts treated only indirectly (NDVI classes)
   The paper acknowledges vegetation influences and stratifies results by NDVI into three bins, finding decreasing SSI–SPI correlation as vegetation density increases (R≈0.68 → 0.58). However, there is no explicit comparison by land-cover class (e.g., grassland vs. forest). Please add an analysis stratified by ECOCLIMAP classes (cropland/grassland/shrub/forest/urban) and report per-class correlation, slope, bias, and any lag differences. This will convert the qualitative interpretation into quantitative evidence.
3. No in-situ soil-moisture validation
   Validation uses (i) station-based SPI (239 stations) and (ii) 1.1 km gridded SPI from LCSC; there is no ground soil-moisture or other hydrologic state validation. The abstract and discussion emphasize robustness “in areas with scarce in-situ measurements,” but for a methods paper this is a gap. Please add at least a limited in-situ comparison (soil probes at 5–10 cm where available) and/or cross-validation with a hydrologic response variable (streamflow, groundwater, or reservoir storage indices) to demonstrate skill beyond precipitation proxies.
4. Alternative satellite products only cited, not compared
   SMAP and Sentinel-3 are mentioned in the context of DISPATCh extensions and literature, but the study does not provide a cross-sensor consistency check. A brief sensitivity test (e.g., SMAP-based DISPATCh over a subset period) would strengthen generality claims and help clarify whether the core findings (e.g., SSI–SPI at k vs k+1) are sensor-agnostic.
5. Uncertainty quantification
   Downscaling relies on LST/NDVI-based SEE and trapezoid assumptions. Please provide an uncertainty analysis (e.g., perturb LST extremes, NDVI, and vegetation fraction; or compare linear vs original DISPATCh parameterizations) to show how retrieval/ancillary errors propagate to SSI and to the reported statistics. The DISPATCh equations and inputs are clearly presented—this section would benefit from error propagation results.

 

Author Response

Dear reviewer, thank you for your time and valuable comments. We have taken all your suggestions into consideration and striven to rewrite the paper accordingly.

In the attached document you will find a point-by-point response to your comments.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

 

Comments to the Author

This study constructs a high-resolution (~1 km, monthly) Standardized Soil Index (SSI) from satellite soil moisture (SMOS + DISPATCH) and compares it with station-based SPI and gridded SPI. It systematically examines the correspondence across integration windows and temporal lags (δ) and the spatial heterogeneity of these relationships, offering practical value for regional drought monitoring. Overall, the manuscript is well structured and potentially publishable after substantial revisions. Below are detailed, actionable comments.

1. Insufficient justification for the reduction in station samples and its impact.
   The manuscript reports a reduction from 328 stations to 239 stations for SPI (2010/06–2023/05, no missing data), yet only 201 stations are used in the subsequent timescale correspondence analysis. The reasons for the loss of 38 stations (e.g., QC thresholds, missing-data ratios, spatial co-registration failures, continuity criteria) and the implications for statistical robustness are not articulated. It is recommended to provide a “sample flow” figure/table in the Methods or Data Availability section, listing exclusion rules and remaining sample counts at each stage; additionally, conduct a robustness analysis using the 239-station baseline (compare correlations, significance, and effect sizes to the 201-station results) to evaluate potential selection bias.
2. Insufficient treatment of regional heterogeneity and robustness in distributional assumptions.
   Based on KS tests, >60% of pixels are best fit by a Beta distribution, and SSI is parameterized accordingly; however, roughly 40% of the domain is not best described by Beta, and the associated uncertainty and potential bias are not quantified. It is recommended to adopt zone-specific best-fit distributions or, in parallel, provide a non-parametric quantile approach, and compare sensitivities against the current parameterized SSI. Please also report spatial maps and histograms of KS statistics and p-values, including thresholds, to verify that distributional choices do not systematically bias particular land-cover or climate zones.
3. The core claim that “the optimal SPI window tends to be one month longer than the SSI window” requires rigorous statistical validation.
   This conclusion currently relies mainly on frequency histograms and mean correlations, which provide limited evidential strength. It is recommended to perform station-level paired comparison tests (e.g., Steiger’s test) or bootstrap confidence intervals and present the distribution of effect sizes (e.g., Δr). Please also apply multiple-comparison control (FDR, Benjamini–Hochberg) to demonstrate that the observed pattern is statistically robust rather than incidental.
4. Transparency in resampling and temporal aggregation for gridded SPI.
   LCSC SPI is aggregated from 1.1 km weekly to monthly and co-registered to 1 km, but the spatial resampling kernel/interpolation scheme, cross-month weekly weighting, and missing-data imputation rules are not specified. Please detail the spatial interpolation and temporal aggregation strategies, and quantify, in representative subregions, how different choices affect correlation, bias, and RMSD to assess smoothing and aggregation bias.
5. Mechanistic support is needed for the performance decline at the 24-month window.
   The decline is currently attributed to limited sample length and over-smoothing, but the relative contributions are not disentangled. It is recommended to conduct rolling leave-one-year-out validation and subperiod comparisons (drought vs. wet regimes) to differentiate the effects of sample length from smoothing; additionally, perform low-frequency filtering sensitivity tests (varying windows/filters) and report impacts on r, KGE, and slope.
6. Traceability of quality control and missing-data handling is insufficient.
   QC masks, thresholds, and treatments for SMOS, LST, and NDVI (including RFI/cloud contamination and temporal interpolation) are not presented in a consolidated way, hindering assessment of their influence on SSI. Please provide a quality-control workflow table listing native QC flags, thresholds/rules adopted in this study, and pixel rejection rates (by season/land cover).
7. Extrapolative language should be tempered.
   Phrases such as “surpasses existing indices,” “applicable to any region,” and “globally robust” are not fully supported by the current evidence. Please adjust to formulations such as “comparable to existing approaches” or “competitive,” unless supported by open benchmarks or cross-regional validation; alternatively, include a systematic comparison with public benchmarks in the Supplement.
8. Cross-sensor validation with additional soil-moisture products is lacking.
   Relying solely on SMOS + DISPATCH may conflate findings with sensor/processing-chain sensitivities, limiting generalizability. It is recommended to incorporate SMAP (or multi-source fusion products) in representative subregions for cross-validation, compare the consistency of SSI construction across sensors, and discuss sources of divergence.

 

Author Response

Dear reviewer, thank you for your time and valuable comments. We have taken all your suggestions into consideration and striven to rewrite the paper accordingly.

In the attached document you will find a point-by-point response to your comments.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

This manuscript presents a comprehensive evaluation of the Standardized Soil Moisture Index (SSI) derived from remote sensing data as a tool for monitoring multiple types of droughts. I think there are several issues need to be addressed to strengthen the manuscript's scientific rigor and clarity.

1 The use of SSI for hydrological drought monitoring is not sufficiently explained within the existing literature. The authors should clarify how their method improves upon previous studies that used model-based soil moisture for SSI.

2 The claim of "global coverage and enhanced spatial detail" is not fully supported, as the study only focuses on the Ebro Basin. A wider discussion on how applicable the findings are elsewhere is needed.

3 The vegetation impact analysis based on NDVI is superficial. The authors should provide more detailed explanations for the observed correlation differences across NDVI ranges, such as the role of evapotranspiration, root depth, and land management practices.

4 The spatial representativeness of meteorological stations is not discussed. How do station density and placement influence the comparison between SPI and SSI, especially in topographically complex areas like the Pyrenees?

5 The correlation between SSI and SPI, approximately R ≈ 0.6–0.67, is moderate. The authors should consider whether this level of agreement is adequate for practical drought monitoring and discuss how uncertainties in soil moisture measurements could impact the findings.

Author Response

Dear reviewer, thank you for your time and valuable comments. We have taken all your suggestions into consideration and striven to rewrite the paper accordingly.

In the attached document you will find a point-by-point response to your comments.

Author Response File: Author Response.pdf

Reviewer 4 Report

Comments and Suggestions for Authors

The work contains very interesting conclusions regarding the integration times and lags of the two water-related indexes. What really bothers me is that the authors do not clearly state what is new and what is an already designed index, used to specific Ebro basin area. The reader may have impression that SSI (from satellite SM) examined in the paper is something new, but Standardized Soil Moisture Index (SSI) - from fusing the space-borne SMAP soil moisture data with the NLDAS climate index was already in paper: Standardized Soil Moisture Index for Drought Monitoring Based on SMAP Observations and 36 Years of NLDAS Data: A Case Study in the Southeast United States, Yaping Xu 1,*, Lei Wang 1, Kenton W Ross 2, Cuiling Liu 1, Kimberly Berry 3 (the the publication is given as an example, the authors do not have to cite it).

 

 

General: more literature in Introduction is needed. For example in sentences that starts in L32, L38, L80, L81

L39 and further: from the beginning of the paper one should know what the numbers near SSI and SPI are (the integration times in months)

L32: …i.e., streamflow… - comma not necessary here

L56: Please explain what authors mean by “climatology”

L59: please explain what “SPI-6-SPI-9” means. “SPI-6 minus SPI-9” or “from SPI-6 to SPI-9” or maybe something else. In further parts of manuscript (Table1, Table2) the “-“ means correlation or “and”, which is even more confusing

L301: literature reference please for empirical Gringorten formula

Figure3: Captions of x-axes are “SPIs with Maximum Correlation”. I believe it should be “integration times of the SPI (in months)” or something similar. Also y-axes are not “Frequency” but should be “number of occurrences”

Table1 and Table2: minus or dash (“-“) between two SSI-a and SPI-b is confusing. Please use a different symbol, as at first glance it seems like a minus sign, when in reality it's a correlation. Perhaps "&" would be better?

Author Response

Dear reviewer, thank you for your time and valuable comments. We have taken all your suggestions into consideration and striven to rewrite the paper accordingly.

In the attached document you will find a point-by-point response to your comments.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

Because the analysis relies on SMOS surface soil moisture (~0–5 cm) downscaled with DISPATCh, the product primarily reflects near-surface wetness rather than root-zone conditions. This vertical constraint limits—though does not entirely preclude—the ability to distinguish land-cover-specific drought behavior (e.g., cropland vs. grassland vs. forest) when dynamics are governed by deeper soil layers, canopy interception, or management (irrigation). Moreover, processes related to groundwater recharge occur on longer timescales and at depth and are not directly captured by a surface index; consequently, the current approach should not be used to infer recharge anomalies without independent hydrologic corroboration.

Actionable requests

1. State prominently (Abstract/Methods/Limitations) that the SSI here represents 0–5 cm surface moisture, not root-zone status, and discuss expected surface–root-zone coupling/decoupling by land cover and season.

2. Expand the land-cover analysis (ECOCLIMAP classes) to report correlation, slope, bias, 95% CIs, and preferred lag (k vs. k+1), clarifying where surface SSI best/least tracks drought behavior.

3. Provide at least limited hydrologic cross-validation relevant to recharge (e.g., piezometer heads or a simple baseflow index) in two representative sub-regions (forest and irrigated).

4. Add a brief use-case table distinguishing applications suited to surface SSI (agrometeorological monitoring, snow/irrigation signals) from those requiring root-zone or groundwater metrics.

These additions would right-size the claims, clarify scope across land uses, and prevent over-interpretation regarding groundwater recharge while preserving the manuscript’s central contribution.

Author Response

Dear reviewer, thank you again for your time and valuable comments. We have taken all your suggestions into consideration and striven to rewrite the paper accordingly.

In the attached document you will find a point-by-point response to your comments.

Author Response File: Author Response.docx

Reviewer 2 Report

Comments and Suggestions for Authors

No.

Author Response

We sincerely thank Reviewer 2 for their positive assessment of the revised manuscript and for confirming that all their previous comments and suggestions have been adequately addressed.

Reviewer 3 Report

Comments and Suggestions for Authors

I have no more questions.

Author Response

We sincerely thank Reviewer 3 for their positive assessment of the revised manuscript and for confirming that all their previous comments and suggestions have been adequately addressed.