Optimized Software Tools to Generate Large Spatio-Temporal Data Using the Datacubes Concept: Application to Crop Classification in Cap Bon, Tunisia

Round 1

Reviewer 1 Report

The article “Optimized Software Tools to Generate Large Spatio‐Temporal Analysis Ready Data (ARD) Using the Datacubes Concept Application to Crop Classification in Cap Bon, Tunisia” shows the classification approach for citrus groves among two other crop classes (olive groves and open field) located in the Cap Bon region, Tunisia. Abstract is not up to the par, not good but introduction is somehow ok, please try to modify as a typical structure of the introduction, please search for it.

 

1.       ABSTRACT: very hard to read and understand. It should be written in a language where scientific readers understand. Concepts are mixed and not clear. Please provide some background, theme of the study, significant results, and future outcomes from this study, they are missing here.

2.       INTRODUCTION: Authors cannot start a sentence by just writing that [15] has conducted research on…….. This is not good, write accordingly. Line 86: Complete the sentence!!!

3.       Figure 2. Acquisition dates of Sentinel‐2 and Sentinel‐1 data: Explanation??? Line 144-147 is does not mean anything.

4.       Sections 2.3.1 and 2.3.2 does not mean anything, they are already based in the QGIS, no need to define again.

5.       Results: Analysis of temporal signatures of crops: Try to connect figure 8 with the Calibration of the algorithms. In this way, the tables and figure would be more easy to read and understand. Now it’s a mixture.

6.       Section 4.2: Not good. Does not provide a good concept, hard to read.

7.       Conclusion: What’s new in it? Please mention your concluding remarks and innovatively in this section so that the reader can get a “TAKE AWAY” message.

8.       REFEREBNCES: Too old, update it.

 

 

Thanks

Author Response

Thank you for your comments, in the following parts, we have answered each comment.

1. ABSTRACT: very hard to read and understand. It should be written in a language where scientific readers understand. Concepts are mixed and not clear. Please provide some background, theme of the study, significant results, and future outcomes from this study, they are missing here.

The abstract has been entirely rewritten, taking into consideration all raised issues: clarification of concepts, background, significant results, and future outcomes from this study. Also includes comments from other reviewers.

2. INTRODUCTION: Authors cannot start a sentence by just writing that [15] has conducted research on…….. This is not good, write accordingly. Line 86: Complete the sentence!!!

You are right. These important details were considered in the revised version. Line 86 was corrected.

3. Figure 2. Acquisition dates of Sentinel‐2 and Sentinel‐1 data: Explanation??? Line 144-147 is does not mean anything.

Figure 2 was eliminated and was replaced by the number of acquisition dates of VV and VH of Sentine-1 and NDVI bands of Sentinel-2, which is actually more informative.

Lines 144-147 were eliminated.

4. Sections 2.3.1 and 2.3.2 does not mean anything, they are already based in the QGIS, no need to define again.

In section 2.3.1 we explained how we obtained and treated the ground truth data. We consider it a very important section that should not t be eliminated

In section 2.3.2 we explained the methodology of the data integration basing on the Tuplekeys and Datacubes concept. This explanation is the key to this work and we believe that should not be eliminated.

5. Results: Analysis of temporal signatures of crops: Try to connect figure 8 with the Calibration of the algorithms. In this way, the tables and figure would be more easy to read and understand. Now it’s a mixture.

A sentence was added that explain that the temporal signatures of crops was used for the calibration of the algorithms. Line 360.

6. Section 4.2: Not good. Does not provide a good concept, hard to read.

Some improvements were made to facilitate the reading the section 4.2, In this section we have tried to find explanation on why the data combination of optical and radar data did not improve the classification results unlike what was expected. Also, we have found other studies that the radar data did not improve their classification results. We think this type of discussion enrich the manuscript. We rewrote it to make it easier to understand.

6. Conclusion: What’s new in it? Please mention your concluding remarks and innovatively in this section so that the reader can get a “TAKE AWAY” message.

The conclusion was improved taking into consideration all your and other reviewers suggestions.

 

8. REFEREBNCES: Too old, update it.

Our bibliography contains 14 articles that were published in 2021 (6 studies) and 2022 (8 studies), and about 53 articles that were published between 2017 and 2022, 38 articles were published between 2000 and 2016 and only 8 articles were published before 2000. So, about more than 53% of the references were published in the last 5 years. Also, included references are necessary for readers to find all the information.

Reviewer 2 Report

This paper is well written and need only minor corrections as explained below:

1. The title of the paper is way too long. Can be shortened 

2. Line 14. In a context of constant climate change should be written in a context of changing climate.  Supervising should be replaced with monitoring.

3. In line 72, no proper flow from [5] and the primary objective. Fix the grammar.

4. In line 86, fix the typos before the parentheses.

5. In lines 126-127. Should be written tomato and pepper 

6. In lines 289-290. Do you mean statistics?

7.In line 344. Grammar issue (''explained by it is particular''). Please fix.

8. In line 350. Grammar issue. ''However, citrus trees have higher biomass this implies that it reflects higher.....''. 'r' in higher is bold.

Author Response

Thank you for your comments, in the following parts, we have answered each comment.

 

1. The title of the paper is way too long. Can be shortened

The title was shorted as following: Optimized Software Tools to Generate Large Spatio-Temporal Data Using the Datacubes Concept. Application to Crop Classification in Cap Bon, Tunisia.

2. Line 14. In a context of constant climate change should be written in a context of changing climate. Supervising should be replaced with monitoring.

In a context of constant climate change was changed to “in a context of changing climate”

Supervising was replaced to “monitoring”

 

3. In line 72, no proper flow from [5] and the primary objective. Fix the grammar.

The sentence was corrected as following: “In addition, another citrus classification study was carried out in Florida by [5], which main objective was to assess the effectiveness of using Sentinel-1A C-band Synthetic Aperture Radar SAR data to classify citrus fruits.”

4. In line 86, fix the typos before the parentheses.

The sentence was corrected as following: “involves dealing with many considerations such as different spectral, spatial resolution etc. (explained in detail in [21]).”

 

5. In lines 126-127. Should be written tomato and pepper

The sentence was corrected as following: “(essentially tomato and pepper).”

 

6. In lines 289-290. Do you mean statistics?

Yes, we mean statistics. Zonal Statistics is an analytical tool specifically for raster datasets.

7.In line 344. Grammar issue (''explained by it is particular''). Please fix.

The sentence was corrected as following: “This alternation can be explained by the sensitivity of VV to soil moisture.”

8. In line 350. Grammar issue. ''However, citrus trees have higher biomass this implies that it reflects higher.....''. 'r' in higher is bold.

The sentence was corrected as following: “However, citrus fruits have a higher biomass, which implies that they reflect more in the infrared than olives, thus presenting higher NDVI values.”

Reviewer 3 Report

This article develops a remote sensing data processing plug-in in QGIS software. The problems are as follows: 1) There is no problem with the method in the plug-in, but the access link to the plug-in should be published, and now we cannot judge the applicability of the plug-in. 2) The selection of feature variables is the key to classification accuracy. Figure 8 shows the time-series dynamics of NDVI/VH/VV, but only the mean value of the time-series data is used in the training model (line 293), which does not make good use of the time-series data and cannot reflect the advantages of your plug-in. 3) The key to this study is the convenient access to time-series remote sensing data, and the manuscript should be more detailed.

 

Author Response

Thank you for your comments, in the following parts, we have answered each comment.

This article develops a remote sensing data processing plug-in in QGIS software. The problems are as follows:

1-There is no problem with the method in the plug-in, but the access link to the plug-in should be published, and now we cannot judge the applicability of the plug-in.

As we stated in the line 275-276 that this plugin "began to be developed by Precisión Agroforestal y Cartográfica Universidad of Castilla-la Mancha (PAFyC-UCLM) in 2022 and still in constant development, that is why the plugin link is not yet available for the public: more formats of images collections and options are going to be added. Also, this actual plugin version still in internal evaluation version, and we will publish it in the next year when it is more tested and stable.

2-The selection of feature variables is the key to classification accuracy. Figure 8 shows the time-series dynamics of NDVI/VH/VV, but only the mean value of the time-series data is used in the training model (line 293), which does not make good use of the time-series data and cannot reflect the advantages of your plug-in.

Thank you for your comment, it was corrected in the line 322 as following” we computed the mean of the features at the plot level for all the available acquisition dates”, because we did not use only the mean value of the time-series, but we computed the mean of the feature (NDVI, VV and VH) at the level of each plot of our ground truth data and for all the available dates of the time-series. Also, as an observation for figure 8 (before) and 7 actually we specified in line 360-361 “These temporal signatures dynamics (obtained from all available dates) were used for the calibration of the algorithms.”

 

3-The key to this study is the convenient access to time-series remote sensing data, and the manuscript should be more detailed.

Thank you for your observation, we have tried to take it into consideration and many detailed was added to make the manuscript more understandable and easier to read. We have added more details for the section 2.3.2.1 concerning the creation parameters of LNG of the study area.

 

 

Round 2

Reviewer 1 Report

REV-2, PAPER: Optimized Software Tools to Generate Large Spatio-Temporal Data Using the Datacubes Concept. Application to Crop Classification in Cap Bon, Tunisia.

 

My previous comments were: Abstract is not up to the par, not good. ABSTRACT: very hard to read and understand. It should be written in a language where scientific readers understand. Concepts are mixed and not clear. Please provide some background, theme of the study, significant results, and future outcomes from this study, they are missing here.

So here, I noticed that the authors have tried to cover up things by editing the complete abstract but editing only few words. This is cheating, its non-negotiable.

 

 

INTRODUCTION: Authors cannot start a sentence by just writing that [15] has conducted research on.... This is not good, write accordingly. Line 86: Complete the sentence!!! Typical structure of the introduction, please search !!!

I am afraid, they could not work hard for this paper, writing the references as mentioned above and not properly modifying them. Typical stucture is missing.

 

Sections 2.3.1 and 2.3.2 does not mean anything, they are already based in the QGIS, no need to define again.

Here, still gaps present, did no worked hard in it.

 

Results: Analysis of temporal signatures of crops: Try to connect figure 8 with the Calibration of the algorithms. In this way, the tables and figure would be more easy to read and understand. Now it’s a mixture.

Here, the authors could not update the sections as requried, unfortunately.

 

Section 4.2: Not good. Does not provide a good concept, hard to read.

Still same as mentioned.

 

Thank you.

Reviewer 3 Report

Manuscript developed a new plugin. But we can't see the picture of the application plugin. The parameters of the study case are described in text, and I believe that the plugin's window for configuring the parameters is designed. The authors can provide some pictures of the windows involved in the study case in the Supplementary material.