Soil Salinity Assessing and Mapping Using Several Statistical and Distribution Techniques in Arid and Semi-Arid Ecosystems, Egypt

Round 1

Reviewer 1 Report

The manuscript entitled “Soil Salinity Assessing and Mapping Using Several Statistical and Distribution Techniques in Arid and Semi-Arid Ecosystems, Egypt” presents interesting study on evaluation of soil salinity based on remote sensing data.

The manuscript is quite well prepared, however contains some drawbacks.

 

Detailed comments:

1. Please use the same font size, e.g. In line 24 font size is smaller than in subsequent lines.

2. Table 1 is exactly the same which is in the Wikipedia. I suggest to present this climate data in the figure instead of table and present only selected variables, e.g. max, min. and average temperature.

3. Please add source of data for Fig. 1. Is it Landsat 8?

4. The satellite imagery was captured during vegetation period or for bare soil. Please notice that plant cover change the reflectance and values of the indices will be dependent on plant cover. Please explain how you avoided effect of various land cover and seasonal changes which cause different value of the spectral indices.

5. Please provide more details about field electrical conductivity measurements. Please add map which present points where the measurements were performed.

6. Fig. 4 present linear regression equations and coefficients of determination. Please change the caption of the Fig. 4 to be more informative. Why in the charts in Fig. 4 there are much less points than 100? It is based on all sampling points? Please explain what data was used for these regression analyses. Moreover, some relationships, for example Fig.4C and 4D seems to be not linear. Please try to use other type of regression than linear.

7. In all the analyses only univariate regression was applied. I recommend to apply multivariate regression. Multivariate regression should be applied based on stepwise selection of independent variables because most of the spectral indices are strongly correlated. Moreover, please consider to apply PCA for evaluation of multivariate relationships.

8. Line 442: Do you mean Table 9 in this study (there are no such table) or in the study presented in the reference [38}. Please be more specific. It would be better to describe the result not only cite another study in the Discussion.

9. The chapter “3.5. Tukey's range, significant and difference analysis (model validation)” demands more detailed description of the methods and input data. Why only 30 observations were used not all 100 observations? Did you applied Tukey HSD test? It is post-hoc test used together with ANOVA. Please provide more details about application of this test in Material and Methods.

10. The last sentence in Conclusions suggest that the results are universal for any conditions. There are no limitations? Probably it can be applied in similar conditions not in any conditions. Please specify the criteria, e.g. land cover during acquisition of satellite imagery.

 

Author Response

Dear Reviewer,

Thank you very much for handling our manuscript. We revised our manuscript in light of the comments from reviewers. Following we interpret how we revised the manuscript and addressed the concerns of reviewers. We confirm that editor / reviewer comments have been considered and addresses in this revised version using track changes.

We hope that you will find this manuscript suitable for publication, and we are glad to address any remaining concerns.

On behalf of my co-authors; I would like to submit the revised version of the enclosed manuscript.

Author Response File: Author Response.docx

Reviewer 2 Report

1. The spatial position points of ground measured points, including distribution and distance of sample points, are not seen in the whole paper, which is directly related to the accuracy of covariance interpolation. Please add the description and distribution map. 2. In the inversion of soil salinity by Landsat-8 satellite, pixel values and measured values at the same location are mentioned in the paper, but this involves the problem of point plane scale. How does the author consider the correspondence between satellite pixel scale and ground sample point scale? If it's in pixels, should I think about multi-point average, or single point? 3. It can be seen from the paper that there is a high correlation between different salinity indices. If so, should principal component analysis be carried out first to build a statistical model?

Author Response

Dear Reviewer,

Thank you very much for handling our manuscript. We revised our manuscript in light of the comments from reviewers. Following we interpret how we revised the manuscript and addressed the concerns of reviewers. We confirm that editor / reviewer comments have been considered and addresses in this revised version using track changes.

We hope that you will find this manuscript suitable for publication, and we are glad to address any remaining concerns.

On behalf of my co-authors; I would like to submit the revised version of the enclosed manuscript.

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

I still don't understand what Tukey's test was used for. The comparisons were performed between paired data. Why for t-test for paired data was not used?

Why my recommendation to use multiple regression was not taken into consideration?

The changes in the manuscript are not tracked and because of that it is difficult to recognize what was changed.

Author Response

Dear Editor and reviewers,
Thank you very much for handling our manuscript. We revised our 
manuscript in light of the comments from reviewers. Following we interpret how we 
revised the manuscript and addressed the concerns of reviewers. We confirm that 
editor / reviewer comments have been considered and addresses in this revised 
version using track changes.
We hope that you will find this manuscript suitable for publication, and we 
are glad to address any remaining concerns.
On behalf of my co-authors; I would like to submit the revised version of the 
enclosed manuscript.

Best regards

Prof. A. Scopa

Author Response File: Author Response.docx