6 Years of SAR (Sentinel-1) and Optical (Sentinel 2, Landsat-8) Acquisitions over Agricultural Surfaces in Southwestern France

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This manuscript aimed to assess the temporal stability of radar and optical time series over 6 years for eight major world crops. The authors have clearly invested substantial effort in this work. The current manuscript contains various comparisons concerning optical and SAR data in monitoring culture as well as their potential influence sources. It also gives some primary conclusions. However, prior to its publication, there are several concerns that needs to be well addressed. My primary critiques revolve around the interpretation of results. As the authors mentioned in the Highlights, the first main finding and implication seem a little weak as highlights, and these can not be regard as your scientific contributions. For now, it is more like an experimentation report rather than a high-level sciencific paper with novelty, there are too much unnecessary information (both text and figures). Thus, the current manuscript is overly verbose overall and should be streamlined. More importantly, the impact of climate on temporal satellite signal profiles seems a little unconvincing since there is no direct influence or process. For example, you mentioned that the earliest NDVI curves correspond to the warmest years at the start of winter in the case of durum winter wheat. Isn’t it obvious? This can not be concluded as the influence of climate on satellite signals. I strongly recommend that the authors provide more direct evidence or a more detailed theoretical justification to substantiate this proposed pathway.

 

Major comments:

In the Introduction, it is better to begin with a brief illustration the importance of  agricultural resources monitoring.

The section number and the reference style are totally confusing for me to follow.

Description of the satellite data: No need to give a such detailed description. Not so relavant with your focus, also too many references. Please modify this part as a more concise one. Actually, Figure 6 is enough to reflect the key information of all the avialable data.

Is it necessary to keep Figure 5 in the context? It looks like that text in Line 235-255 and 268-296 also can be shortened. Too much for now.

There is a total of 6 figs before method part, in my opinion, it is quite redundant.

 

Minor comments:

Line 30 GAI, please give the full description at its first apperance.

Line 86 Remove the second comma.

Line 88 Why the ‘and/or’? What do the authors exactly imply?

Line 89 There is no need to simply list this many references without necessary disscussion.

Line 97 Currently, the Sentinel-1 constellation has two satellites, namley 1A and 1C. 1B has stopped operation since powersupply unit failure. It is better to clarify this for readers.

Line 98-100 References and demonstration needed. What are the shortcomings of these published studies?

Line 101-102 Again, I think the point is to conduct comprehensive literature summary (e.g. problems and limitations) instead of simply listing several references in the part especially when this is also your reasearch focus.

Line 110 Too many references.

Line 119-120 Explain why the authors choose these crops.

Line 120-123 The description about study area can be moved its corresponding section.

Figure 1. Nice figure. Better to plot the latitude/longitude.

Line 156-160 There seems to be some typesetting errors in the caption.

Figure 2. What is the difference between Relative coverage (%) (a) and number of parcels (b)? Is it necessary to use both indices?

Line 189-190 What is the data source? from field sensors, nearby AWS or reanalysis data? What kind of equipments?

Line 190 Give specific numbers for the minimum and maximum.

Line 214 Keep a consistent reference style.

Figure 7. The description in the figure is ambiguous, please modify.

Where are the sections 3.1-3.5 in the Methodology?

Line 379 Keep the reference style uniform.

Figure 9. For Sentinel-2, 385 or 313?

Line 418-419 You mean nearly 80% Sentinel-2 images with cloud cover equals 0? Seems impossible for me. By the way, how do you calculate the cloud cover?

Figure 11. and 12. Please remove them and shorten the corresponding text. Line 483-499 is much more vital than the previous two paragraphs. It is better to reorganize this section.

Line 521-590 Similar as optical data, there is too much detail which needs largely corrections and deletions. Or you can move thses three sections (Optical data, SAR data, Impact of SAR orbit on phenological monitoring) into the supplementry materials.

Line 831-832 I don’t see there is any strong connection between temperature/precipitation accumulation and growth rates, what is the meaning?

There are also some typos in the references, such as ‘Remote sensing of environment’ in Line 940, should it be Remote Sensing of Environment? All first letters should be captialized. Some with doi while some not.

Comments for author File: Comments.pdf

Comments on the Quality of English Language

The manuscript requires extensive language polishing to meet the standards of a scientific publication. There are some unclear sentences, repeat and unnecessary description throughout the text, which hinder the reader's understanding. I strongly recommend professional English editing to improve the flow, coherence, and overall readability.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

Thank you for submitting this manuscript. The subject is engaging, and the effort invested in the research is notable. However, further effort is required to provide a clearer vision and presentation of the work.

Specific Comments by Section

Abstract

The abstract is excessively long—approximately double the size of typical abstracts. Please consider summarizing it significantly.

Materials

The Materials section is lengthy and requires better organization.

• Please maintain 1.1 Study Area for the geographical description only. The other factors should be grouped under distinct subheadings such as Land Use or Crop Classification, Meteorological Data, Soil Analysis, etc.
• Under each subtitle, avoid re-listing all values from tables and figures. It is sufficient to include the nature or description of the data, the source, and a few representative ranges.
• Correct the heading to 1.2 Description of the Satellite Data (instead of repeating 1.1). Structure this section further using sub-subtitles (e.g., 1.2.1, 1.2.2, 1.2.3) for the different satellite orbits used (Sentinel-1, Sentinel-2, and Landsat).
• Please remove the numbering within the descriptive text below the subtitles; it will be better presented as normal paragraphs.

Methods

• The mentioned sections (3.1, 3.2, and 3.3) could not be found in the rest of the manuscript. Please remove these references from the methodology text and from Figure 7.
• Please unify the citation style throughout the manuscript. The style at line 379 is inconsistent with the rest of the paper.
• Consider presenting the NDVI equation within the Methodology section.

Results and Discussion

Impact of the cloud cover on optical images repetitiveness:

• Line 418 marks the first mention of the vegetation index GAI. If this index was used in the study, please include it in the Methodology section; otherwise, it should be removed from the Results and Discussion.
• The symbols H and S are not defined in the graph or its accompanying caption.
• There is a lack of citations to other research papers to enrich the discussion (e.g., citing authors who have previously addressed low NDVI accuracy under high cloud cover).

SAR data:

• The first three paragraphs along with Figure 14, which describe the used downloaded SAR images, should be moved to the Materials section.
• Equation 2 and its description belong in the Methodology section.
• Line 585 uses a different citation style.

Impact of SAR orbit on phenological monitoring:

• The subtitle has incorrect numbering; please remove it.
• Line 625 uses a different citation style.
• Line 699 uses a different citation style.
• Is there an explanation, supported by references, for the different attitude of the results between orbit 30 and orbit 110?

SAR-NDVI comparisons per crop type:

• The subtitle has incorrect numbering; please remove it.

Interannual comparison of temporal satellite signal profiles – double logistic regression:

• The subtitle has incorrect numbering; please remove it.
• Line 794 uses a different citation style.
• Tables 5 and 6: The tables are not clear. Please consider decreasing the font size to improve the visual structure and readability of the tables.
• Table 6 at line 874 is a repeat number. Please change it to Table 7 and ensure this correction is reflected in the main text as well.
• In lines 863–872, please clearly state which figure you are using to compare its attitude with the values in Table 7.

Conclusion and Perspectives

• The opening phrase "This publication aimed to..." should be converted to "This study aims to..." for clarity and consistency.
• The conclusion is very long. Please consider summarizing it more concisely.
• Avoid mentioning the subtitles' names from the Results section before concluding their findings. The conclusion should summarize the entire paper in a structured but continuous narrative, flowing like a story, rather than a point-by-point recap.

General Formatting and Consistency

• Please ensure that sentences and figure captions are not broken by new lines at the beginning or end.
• The symbols used in the figures must clearly fit on the intended plots, and adjacent plots within the same figure should not be broken by a new line.
• It is better to avoid starting a sentence or paragraph with a figure number.
• Table 1 lists symbols for the used crops, but the text sometimes uses the symbols and sometimes the complete names. Please unify the use; decide whether to use symbols only in tables/figures or consistently throughout the text.
• Please unify the numbering structure of the entire paper. The Materials section starts with subsection 1.1, while the main Materials section lacks the number '1', and numbering is then absent in subsequent main sections (Methodology, Results and Discussion). If numbering is used, it should start from the Introduction and be applied consistently to main sections and all subsections. Alternatively, remove all numbering and rely solely on titles and subtitles.
• Unify the citation style across the entire manuscript, as differences are present.

Comments for author File: Comments.pdf

Comments on the Quality of English Language

Please avoid long sentences as a 5 -7 lines sentence is hard to follow.

Please avoid starting the sentence or paragraph with the figure or tables number.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

Dear authors, 

I just think you did an outstanding work joining radar and optical images to study crops behavior, taking 8 different cultures from south France. Your study will serve to other researchers to conduct similar approach to know deeply about other species from different latitudes and heights about ground level.

I highlight mainly how you take a large group of images and synthetize them on graphs, and relating them with NDVI index, these help to understand better their behavior and growth limits.

Also, I got a better understanding about polarizations and their impact to record vegetation growth.

I just applaud your big effort doing this work, and share it around.

Best regards, 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 4 Report

Comments and Suggestions for Authors

Please refer to the attached file.

Comments for author File: Comments.pdf

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Reviewer 5 Report

Comments and Suggestions for Authors

This manuscript provides a substantial and well-structured contribution to the field of crop monitoring through the integration of long-term SAR and optical remote sensing data. The research leverages six years of high-resolution Sentinel-1 (SAR) and Sentinel-2/Landsat-8 (optical) data over agricultural areas in southwestern France to model crop dynamics for eight globally significant crop types. The study is methodologically rigorous, comprehensive in scope, and the findings are highly relevant to agricultural remote sensing and crop phenology modeling. The use of double logistic regression models to describe seasonal crop dynamics is well-founded and statistically robust (R² > 0.92 in most cases), enabling reliable tracking of interannual variation in crop development. The complementary strengths of SAR and optical data are clearly demonstrated, with careful attention given to challenges such as cloud cover, angular geometry, orbit effects, and polarization configurations. The consideration of both co-polarization and cross-polarization backscatter, as well as their ratio, provides a nuanced view of the radar signal’s sensitivity to phenological stages. The study would be further strengthened by tighter integration of in-situ validation data, more explicit discussion of model transferability, and clarification of reproducibility elements. Some restructuring of dense results sections could also help improve readability. Nonetheless, the methodological framework and empirical findings offer significant value to researchers and practitioners working in satellite-based agricultural monitoring.

Despite its strengths, the manuscript would benefit from addressing several weaknesses and clarifying key aspects.

1. Missing information on in-situ validation (Lines ~55–58, 767–789):
While the manuscript states that the sites belong to instrumented networks (JECAM, ICOS), it is unclear how, or if, in-situ data (e.g., biomass, LAI, yield) were quantitatively used to validate the modeled time series. While there is some reference to departmental yield data, no direct field-level validation of the satellite signal modeling is performed. The lack of ground-truth validation limits the interpretation of model robustness across the full range of crop conditions and interannual variability.

2. Limited discussion on transferability (Lines ~783–789):
Although the manuscript claims the modeling framework is transferable to other agro-climatic regions, this is not demonstrated or supported by specific evidence. A short discussion on potential challenges for extrapolation—such as soil moisture variability, crop phenological plasticity, or differences in tillage and irrigation practices—would improve the utility of the study for broader audiences.

3. Clarity on methodological reproducibility (Lines 278–341):
The processing pipeline is well-described but could benefit from clearer separation of pre-processing versus modeling steps. Providing access to code (e.g., for the logistic regression or NDVI interpolation) or processed datasets as supplementary material would enhance transparency and reusability. This is especially important for the double logistic fitting procedures.

4. Excessive length and density in some sections (e.g., Lines 343–749):
Several results sections (notably 4.3 and 4.5) are lengthy and verbose, potentially limiting reader engagement. These could be more digestibly structured, perhaps through summarizing paragraphs at the end of each crop section, or by moving some secondary figures and detailed year-by-year discussions to supplementary material. The current format includes extensive appendices, but the manuscript body itself could also be streamlined.

5. Incomplete referencing in some claims (e.g., Lines 575–579):
Some of the observed radar signal behaviors (such as double-bounce effects or diurnal signal shifts) are stated as known phenomena but could be more consistently backed by references to support readers less familiar with SAR theory.

6. Under-discussed spectral differences between sensors (Lines 432–436):
The authors state that wavelength differences between Sentinel-2 and Landsat-8 do not significantly affect NDVI when vegetation is developed. However, this conclusion would benefit from a citation or quantitative test of spectral band matching or cross-calibration, especially at low NDVI values.

7. Satellite data access and processing software (Lines 243–249):
More clarity is needed on whether the VisioTerra POF-ML pipeline is proprietary or publicly available. If the software is not open-source, this could present reproducibility challenges. Explicitly stating its accessibility (or providing an alternative method) is advised.

8. Implications for operational monitoring (Lines 785–790):
The conclusion hints at applications for anomaly detection and operational crop monitoring. However, it would be useful to elaborate on the conditions or indicators that would define a “phenological anomaly,” and how the double logistic model might integrate into an early warning system or decision support tool.

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

Thank you for submitting your manuscript. The update has notably improved the clarity and quality of the work. However, a few modifications are still required to provide a clearer vision and presentation of the research.

Materials:

• Replace the subtitle ‘2.2.1 Optical data’ with ‘2.2.2 Optical data’.

Methodology:

• Please add the primary source/reference for the NDVI equation.

Results & Discussion:

• Line 347: Remove the repeated word ‘cover’.
• Figure 7: Correct the gray symbol value in the legend to ‘cc > 40%’ instead of ‘cc < 40%’.
• Tables 5 and 6: These look much better. I recommend including vertical column lines or increasing the spacing between the ‘Year’ and ‘R²’ columns to better distinguish their values.

Comments for author File: Comments.pdf

Author Response

Please see the attachment.

Author Response File: Author Response.pdf