The Capabilities of Optical and C-Band Radar Satellite Data to Detect and Understand Faba Bean Phenology over a 6-Year Period
Round 1
Reviewer 1 Report
Comments and Suggestions for AuthorsInteresting work, good results for the sensors integration and the SAR index. DL models seem very adequate for phenology issues. Few questions remained:
- Please correct mention to Figure 2 on page 4
- Not sure if Figure 2 is needed
- What FVL means?
- Did you considered using EVI instead of NDVI?
- Please correct legend of Table 2 for "Summary"
- Did you considered resampling Landsat OLI data to 10 m (such as Sentinel 2 data)?
Author Response
Please find our responses in the attached file
Author Response File: Author Response.docx
Reviewer 2 Report
Comments and Suggestions for AuthorsReview Comments for remotesensing-3592452
Capabilities of optical (NDVI) and C-band radar satellites data to detect and understand faba beans phenology during a 6-year period
1. Overall Evaluation
This study represents the first attempt to utilize multi-sensor satellite data for long-term monitoring of faba bean phenology, filling a critical research gap in remote sensing for this crop (existing studies have predominantly focused on soybeans). By integrating multi-source satellite data (Sentinel-1/2, Landsat-8), the authors rigorously analyze the potential of optical and radar data for phenological monitoring of faba beans. The methodology is robust, with an extended temporal coverage (2016–2021), and the conclusions hold valuable implications for agricultural remote sensing applications. However, certain aspects require further refinement to enhance scientific rigor and readability.
2. Possible Problems and Recommendations
(1) Regarding faba beans as an intercrop:
How did the authors accurately identify and extract phenological information for faba beans used as an intercrop? How were faba beans distinguished from other crops?
Section 5.4:
Figure 14a displays the phenological cycle observed via satellite data for a faba bean intercrop during 2019–2020 in the Auradé field.
(2) Table 1:
The correlation coefficients lack annotations for significance levels (p-values), which should be supplemented.
(3) Cloud masking algorithm:
The specific cloud detection methods for Sentinel-2 and Landsat-8 need clarification to enhance reproducibility.
(4) Sowing strategy impacts:
While Section 5.3 mentions "autumn vs. spring sowing," it does not validate these claims with field management data (e.g., tillage records). The authors previously noted that "The French RPG does not distinguish between sowing periods," which is regrettable.
(5) Orbit difference analysis:
Figure 12 shows sensitivity variations between orbits 30 and 110, but the analysis lacks deeper explanations integrating incidence angles or polarization mechanisms. Theoretical support from radar scattering models may be added.
Author Response
Please find our responses in the attached file
Author Response File: Author Response.docx
Reviewer 3 Report
Comments and Suggestions for AuthorsThis study presents a comprehensive analysis of faba bean phenology monitoring using multi-sensor satellite data (Sentinel-1/2, Landsat-8) over six years in southwestern France. The integration of optical and radar data to address cloud limitations and improve temporal resolution is commendable. However, several issues need to be addressed. Below are the major concerns and suggestions for improvement:
1 The research involves 78 fields spanning six years, but the count of fields differs greatly each year (e.g., 4 fields in 2017 compared to 30 in 2021). This discrepancy creates concerns regarding the statistical reliability of comparisons across years.
2 The study identifies discrepancies in NDVI values between Sentinel-2 and Landsat-8 caused by residual cloud and shadow impacts (Figure 11). Nevertheless, the suggested 40% cloud cover threshold seems arbitrary and may not effectively address radiometric distortions.
3 The connection between radar backscatter (γâ‚€VH/VV) and biophysical parameters like LAI and biomass lacks a mechanistic explanation.
Author Response
Please find our responses in the attached file
Author Response File: Author Response.docx
Reviewer 4 Report
Comments and Suggestions for AuthorsIntroduction:
I would like if trends in green and dry faba bean production were compared over the same time period.
I would like if there were more explanation on radiometric theory between different radar signals
1. Study site:
Figure 1: The country map in the top right should have a box that spatially matches the large panel. The green circle represents an ambiguous area.
I do not understand how the crop can be planted in autumn and harvested in June July, unless it winters over like wheat. Please explain the differences between faba beans which are grown over the winter and those which are not if this is the case. *I now see in Figure 6 that this is the case.
2. Dataset:
The seasonal pattern of global radiation in 2017 is quite different than other years. Qualitatively explain or try to deduce why. Also, I do not believe global radiation is the correct term for the variable plotted and described. Perhaps it is insolation or shortwave radiation.
List the range or spectrum of wavelengths utilized by Sentinel and Landsat
3. Methodology
Equation 1 should be formatted for uniformly aligned height of terms
What are the goals and research questions of this study? Describing these (possibly in earlier sections as well) will give data sets and equations more context.
It strikes me as a little odd that this section prefaces / summarizes other sections. This is sometimes done in the introduction, but this is fine.
Which radar signals do we expect to best represent vegetative health / cover? State hypotheses
4. Results
The reader knows very little about how or why you modeled radar and NDVI
Sometimes the radar polarization font carries onto subsequent words by mistake
I do not understand what you are trying to say about air temperature here "The year 2018 is also remarkable because it presents the highest cumulative rainfall, combined to standard cumulative air temperatures."
5. Discussion:
Figure 11: Use the same point size scale for a) and b)
Describe how destruction of cover crop by farmers may appear different than harvesting through the lens of remote sensing
7. Conclusion (Which should be 6)
Tell me more about faba bean as a summer crop
Overall:
I would be helpful for all of us if there were line numbers listed. Many grammar mistakes. Stronger motivation for this research is needed throughout the paper.
"provides" to "provide"
wheat, and rapeseed
There is bad syntax when referring to Figure 2
"fields delineation" to "fields' delineation"
"such cloud" to "such as cloud"
", phase" to " and phase"
"specifics behaviors" to "specific behaviors"
The quantity of water in soil is referred to as 'soil moisture' or 'water content', not humidity
"value of the angle of incidence at which radar images are acquired" to "incidence angle"
"fava" to "faba"
Author Response
Please find our responses in the attached file
Author Response File: Author Response.docx
Round 2
Reviewer 4 Report
Comments and Suggestions for Authors***Second Round of Revisions***
Comments by page number, line number:
2, 51: It is more accurate to say "The VH/VV ratio is more sensitive to vegetation structure and moisture signals than changes in soil moisture and incidence angle".
Figure 1: Very nice! Too large in the paper, but I believe the editors can fix this.
5, 26: I am also curious why 2017 has the least incident solar radiation of all years from July to September. And did something happen in June this year to flip the anomaly? It is worth mentioning if known.
Radar polarization font carries over to more text on page 13, line 5 (2 instances of this). Elsewhere this error looks to be corrected. Line numbers per page are helpful, but line number for the whole document is even better.
Overall: grammar and clarity is improved. Theoretical explanations are helpful and insightful. I approve of changes in this round of revisions. There are a few minor things to clean up.
Author Response
please find our responses in the attached file
Author Response File: Author Response.docx