Sentinel-1 Data for Winter Wheat Phenology Monitoring and Mapping

Round 1

Reviewer 1 Report

Nasrallah et al. investigated the use of Sentinel-2 SAR data for mapping wheat phenology stages. The method is to fit three Gaussian kernels to detect the maximum values that are unique to phenology stages. The method is also validated with a satisfying result. This study provides a good example for use of S1 data to agricultural applications.

 

Major comment:

this is kind of a “hindcasting” study, showing a method assuming the full time series is available. For operational use, how the normalization is conducted if a full time series is not available? For a nowcasting application, you may have a part of time series (now, the current moment), the length of time series may be not enough to fit a gaussian kernel and show the max value/location/date. I would expect the authors to clarify this issue. Clarify if the method can be applied to a different species of wheat? The authors already have a short discussion, more in-deep discussion is expected for the potential challenge. The impact of background soil water content to desired signal (phenology) could be clearer.  It seems that a different English punctuation system is used. The current manuscript is quite long although it is acceptable to me.  I did found that terrain correction is applied. What's the resolution of DEM data you used for this 10-m application?

 

Specific comments:

 

L26-27:  it is unclear what is “Gaussians” only from the abstract.

L122: “had had started”

L153: make sure the ranges are consistent to the figure 1.

L208: S1 also should be defined in the main text.

L210: sympathized → synthesized?

Figure 4: usually a figure should be fit in a single page.

Fig. 7 and Fig. 8: it seems that Gaussian curves cannot fit some part of curve in the time series.

Fig 6-8: although julian days are fine, it is better to show DOY, or month-year format for the x axis.

Figure .11: this provides a good example that the gaussian fitting may be not enough to capture the curve shapes.

L365: soil moisture. I would suggest to use IEM and water-cloud model to analyze the sensitivity of soil moisture to your result although it is not mandatory. Suggested literature:

Bai et al., First assessment of Sentinel-1A data for surface soil moisture estimations using a coupled water cloud model and advanced integral equation model over the Tibetan Plateau, Remote Sensing Volume 9, Issue 7;

He et al., Simulation and SMAP Observation of Sun-Glint Over the Land Surface at the L-Band, IEEE Transactions on Geoscience and Remote Sensing, vol. 55, no. 5

L665: this conclusion is derived based on a full time series. For operational application, i. e., with a “half” time-series, the fitting of gaussian curve may be impossible, could the same rmse be achieved?

 

Author Response

Dear reviewer, we highly thank you for your comments. Please see the attachment.

Author Response File: Author Response.docx

Reviewer 2 Report

Sentinel-1 Data for Winter Wheat Phenology Monitoring and Mapping

The study refers to the use of Sentinel-1 data for winter wheat phenology and mapping in the Bekaa plain of Lebanon. For achieving this, the temporal variation of Sentinel-1 signal is analysed based on in-situ phenology measurements. In addition, the authors compared the results with NDVI measurements from Sentinel-2 data. The authors found that each crop stage was identified better using different polarization (VV/VH at low incidence angle (32°-34°) for germination and harvesting; VV at low incidence angle (32°-34°) for heading; and  VH polarization at high incidence angle (43°-45°) for soft dough). While, the Root Mean Square Error (RMSE) of the estimation of the phenological phases’ was varying from 0.6 days up to 5.5 days according each crop stage. Finally, the authors stated that Sentinel-1 are powerful as a tool for crop monitoring, to serve enhanced crop management and production handling

Comments

Abstract

L15-36. Past tense must be used since the study refer to a completed work and not on a work in progress.

L37. “Sentinel-1” instead of “Sentienl-1”.

Introduction

L40. The authors need to use present tense or to indicate from which time period sufficient and stable crop productivity is considered a global desire.

L40-47. References should be provided to support the statements of this paragraph.

Winter Wheat Phenology: Economic and Agronomic Importance

L99-150. I recommend this section to be incorporated with the previous (Introduction). The information provided for wheat needs to be reduced since the focus of this journal is remote sensing and not agriculture.

Study Site

L151-172. This section needs to be incorporated in the material and methods section.

Material and Methods

L176-188. Past tense must be used in this paragraph.

L192-194. This needs to be incorporated in the previous paragraph in order to help in the justification of using the NDVI. In addition, better justification for using NDVI is needed.

L196-209. Past tense must be used in this paragraph.

L219-237. Past tense must be used in these paragraphs.

L238-256. Past tense must be used in these paragraphs.

A subsection regarding the software used and the statistical analysis that was conducted must be introduced in this section.

Results

L287-300. These paragraphs must be incorporated in the previous section. Past tense must be used.

NDVI Temporal Profiles

The authors need to use past tense. A table with descriptive statistics should be introduced for helping the readers to better interpret the figures.

Sentinel-1 Temporal Profiles

The authors need to use past tense. A table with descriptive statistics should be introduced for helping the readers to better interpret the figures.

L324-325. “Similar results were pinpointed in a recent study [32]” instead of “A similar result..”.

Optimal S1 Configuration for Mapping the Three Phenological phases (Germination, Heading and Soft dough)

The authors need to use past tense.

L373-375. This sentence must be rephrased.

Optimal S1 Configuration for Mapping Harvesting (West and North Bekaa)

The authors need to use past tense.

Figure 5 and L396-405. These must be moved in the beginning of the Results section.

Smoothing and Gaussian Fitting

The authors need to use past tense.

Germination, Heading, Soft Dough and Harvesting Mapping

The authors need to use past tense.

Figure 9. The authors need to provide figures with the same resolution. The figures should have outlines.

Discussion

S1 Versus NDVI Temporal Behavior

L519-558. The authors need to use past tense.

The authors could compare the results between NDVI and S1 data using appropriate statistical analysis (e.g. using Pearson correlation) in order to assess better the performance of S1 data on phenology estimation.

Influence of S1 Incidence Angle

The authors need to use past tense.

L570-576. This paragraph should be removed and incorporated to the Materials and Methods section.

Wheat Phenology Mapping

The authors need to use past tense.

L629-632. These sentences must be removed and incorporated in the previous sections.

Quality Indicator, Strengths and Perspectives

The authors need to use past tense.

L687-692. This paragraph must be incorporated to the following section (Conclusions).

Concluding remarks

The section must be renamed to Conclusions.

L694-695. The authors must remove the “First experience with”.

L699-700. The authors need to rephrase this sentence for providing a better meaning with the previous sentence.

L700-705. I recommend the authors to use “observed” instead of “seen”.

L711-717. It is better to use passive voice for presenting your future work.

References

L772-776. The authors need to properly write the title of the manuscript using small letters also.

General Comment

I recommend this manuscript to be accepted after minor revision. The study provides an innovative method to assess the phenological stages of winter wheat by incorporating Sentinel-1 data. Sentinel-1 data are considered valuable in agriculture because they can provide information even during cloudy weather, which this is not the case when using Sentinel-2 data (or data from other optical satellites). Moreover, the results of their study are very promising for adoption of their approach to other crops of high importance (e.g. potato, maize, other cereals and vineyards).

Author Response

Dear reviewer, we highly thank you for your comments. Please see the attachment.

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

Thanks to the authors for carefully addressing my comments.