Review Reports

Round 1

Reviewer 1 Report

The article describes the result of cluster analysis of satellite images and digitized geological maps. This allowed us to discover some details of the geological structure.
We believe that in the introduction it is necessary to focus in more detail on research using machine learning. The authors have given only one review article on this issue, it is necessary to indicate the main directions of research using machine learning.

 

Author Response

Dear Reviewer,

Please see the attachment.

Best regrads.

 

Author Response File: Author Response.pdf

Reviewer 2 Report

This manuscript presents an analysis of images from Sentinel 2A along Croatian region to perform more detailed lithological maps and to improve mineral exploration. Standard data processing approach has been then applied and a comparison with ground geological data has been presented.  I have appreciated the methodology for data interpretation, in spite of its simplicity. The results are clearly documented and illustrated, and the techniques for data interpretation are coherent and logical. The discussion of results and conclusions appear rather exhaustive. The references are adequate.

In conclusion I have found the obtained results are generally rather convincing to understand the lithology of a region and my suggestion is that the paper can be published on Remote Sensing.

Author Response

Dear Reviewer,

Please see the attachment.

Best regrads.

Author Response File: Author Response.pdf

Reviewer 3 Report

Some technical issues can be addressed or clarified to improve the quality of the manuscript:

(1)  The introduction section did not point out the necessity underlying this research. In other words, it is suggested to highlight the key issues to be solved in this study.

(2)  Lines 73-75: ‘Remote sensing data processing using ML algorithms at the medium-scale map level....A new approach for…’. It is advisable to review the previous studies on this topic for supporting the ‘new approach’ proposed in this study.

(3)  Lines 143-145 introduced the K-means and Random forest algorithms used in this study. Please provide more details on the ML model construction and validation.

(4)  The spectral signatures in Figure 6 (i.e., the last column) were difficult to read. Please improve their resolution.

(5)  Please fine-tune Figure 7 for improving the readability.

(6)  For Table 1, what’s the physical meaning under the ‘p < 0.000’.

(7)  Please reorganize section 4, it is difficult to get the point.

(8)  For the conclusions, it is suggested to focus more on the implications of this study for geologists in the discrimination of rock units in Karst Terrains.

Author Response

Dear Reviewer,

Please see the attachment.

Best regrads.

Author Response File: Author Response.pdf

Round 2

Reviewer 3 Report

Thanks to the contribution of all the authors, the manuscript has been carefully revised according to the reviewer comments.