A Comparison of Landforms and Processes Detection Using Multisource Remote Sensing Data: The Case Study of the Palinuro Pine Grove (Cilento, Vallo di Diano and Alburni National Park, Southern Italy)

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The authors presented a study aims at the recognition of landforms and processes related to flooding hazard using multi-source data (photogrammetric DTM and LiDAR-derived DTM). In detail, authors used 5m, 1m and 20cm resolution DTM for the recognition of landforms. They compared the outcomes of grid-based methods for landform extraction and 16 surficial process type assessment, leveraging various DEMs as input data. The study area is located in Southern Italy along the Cilento coastline, where recently a flooding events have been experienced.

Despite this research is related to a single case study, it represents an interesting work considering the peculiarity of the area (Cilento, Vallo di Diano and Alburni National Park, UNESCO Global Geopark and World Heritage), the landform recognition and process domains detection, and the multi-source dataset and methods used.

The present manuscript covers an area of interest to the journal's readership; the goals are evident, but some parts of the manuscript need a minor improvement.

The introduction appears comprehensive and well organised. The Study area is briefly presented while the Materials and Methods are readable, quite clear and "step by step" described. Results are evident and the successes are well highlighted however some "failures/limits" of used dataset and methods should be highlighted. Discussion and conclusion deal with implications of the work based on the results and should be improved by adding some further considerations. References seem to be appropriate.

In order to improve the manuscript, I recommend authors answer the attached questions/comments and minor points.

Comments for author File: Comments.pdf

Author Response

General comment 1: The authors presented a study aims at the recognition of landforms and processes related to flooding hazard using multi-source data (photogrammetric DTM and LiDAR-derived DTM). In detail, authors used 5m, 1m and 20cm resolution DTM for the recognition of landforms. They compared the outcomes of grid-based methods for landform extraction and 16 surficial process type assessment, leveraging various DEMs as input data. The study area is located in Southern Italy along the Cilento coastline, where recently a flooding events have been experienced.
Despite this research is related to a single case study, it represents an interesting work considering the peculiarity of the area (Cilento, Vallo di Diano and Alburni National Park, UNESCO Global Geopark and World Heritage), the landform recognition and process domains detection, and the multi-source dataset and methods used.
The present manuscript covers an area of interest to the journal's readership; the goals are evident, but some parts of the manuscript need a minor improvement.
The introduction appears comprehensive and well organised. The Study area is briefly presented while the Materials and Methods are readable, quite clear and "step by step" described. Results are evident and the successes are well highlighted however some "failures/limits" of used dataset and methods should be highlighted. Discussion and conclusion deal with implications of the work based on the results and should be improved by adding some further considerations. References seem to be appropriate.
General Response 1: Thanks for the thoughtful revision. In the first section of the discussions, limits related to grid-based techniques are highlighted. In the last sentence (lines 442-443 of the revised manuscript) the statement “... related to fuzzy boundaries and/or small groups of isolated cells …” has been added to further clarify that. Other considerations related to comment #9 have been also added in the conclusion (lines 532-544 of the revised manuscript).

In order to improve the manuscript, I recommend authors answer the attached questions/comments and minor points:

Comment 1: Page 2, Figure 1: The figure appears to be placed out of context. Please move it at the bottom of the chapter 1.
Response 1: The figure has been moved (p. 3 of the revised manuscript).

Comment 2: Page 2, line 74: Please change southern Italy in Southern Italy.
Response 2: Corrected (p. 2, line 87 of the revised manuscript).

Comment 3: Page 3, line 78: Please change the general in The general.
Response 3: Corrected (p. 2, line 90 of the revised manuscript).

Comment 4: Page 3, Figure 2: It would also be interesting to see the 20cm dataset in the same figure. If it has not been included for some reason (e.g., high resolution is not appreciated in the figure) please give reasons.
Response 4: The 20 cm dataset has been represented separately as it is one of the products of this work. Fig. 3 (p. 6 of the revised manuscript) shows the point cloud of the lidar survey, whereas Fig. 4 (p. 7 of the revised manuscript) shows the resulting DTM and contour lines. Captions of Figs. 3 and 4 have been updated highlighting that their contents are the results of our LiDAR survey.

Comment 5: Page 5, lines 142-151: Although the LiDAR survey was conducted with UAV equipped with GNSS in nRTK mode, which ensures good data quality, please clarify whether ground control targets (e.g., Quality Control Point - QCP) were used and, if not, explain how the error was checked.
Response 5: A small paragraph has been added in section 2.2.1 to clarify the concept and justify the choices (p. 5, lines 172-176 of the revised manuscript).

Comment 6: Page 6, lines 192-195: Please clarify even in this section whether analyses were conducted on all the available datasets (5m, 1m, 20cm).
Response 6: A sentence clarifying the matter has been added at the end of the paragraph: “The workflow outlined in the next sections has been implemented an all three datasets” (p. 7, lines 221-222 of the revised manuscript).

Comment 7: Pages 7-8, lines 239-264: For a better understanding, please add a summary table of 6 process domains detected, also relying on [10].
Response 7: Table 2 has been added (p. 9, line 266 of the revised manuscript).

Comment 8: Section 5 "Conclusions": It seems to lack considerations of the time offset of data acquisitions (i.e. different periods of data acquisition: 5m DTM acquired on 2011, 1m DTM acquired on 2013, UAV LiDAR survey carried out on?) and possible effects that this may have produced on the assessment.
Response 8: A paragraph has been added at the end of the conclusion section highlighting multitemporal implications (p. 21, lines 532-544 of the revised manuscript).

Comment 9: Section 5 "Conclusions": Given the specificity of the journal, it would be appropriate to add considerations about the different nature of the data (photogrammetric DTM and LiDAR-derived DTM) and possible effects that this may have produced on the assessment (e.g., issues on filtering and interpolation processes, etc.).
Response 9: Additions have been made in the conclusions section as recommended by the reviewer (p. 20, lines 512-525 of the revised manuscript).

Comment 10: Figures 11-12: Figures appears to be placed out of context. Please move them to the appropriate sections (e.g., 3.2 and 4 respectively).
Response 10: Figures have been moved (p. 17 and p. 19 of the revised manuscript.

Comment 11: Title: Considering the content of the paper and the datasets used, the title is thought to be slightly misleading. In fact, it should be specified that both photogrammetric and LiDAR data have been used. Given the above, please change the title as best you see fit.
Following some proposals:
•    Enhancing landforms and processes detection using multi-source data
•    Enhancing landforms and processes detection using photogrammetric and LiDAR data
Response 11: Considering also other reviewers comments, the title has been changed to: “Comparison of landforms and processes detection using multi-source data”.

Reviewer 2 Report

Comments and Suggestions for Authors

(1)   A key word in the title of this MS is “enhancing”, which I expect to see some improvement on the methodologies to enhance the landforms detection. However, as I read through the MS, the majority of this MS is with respect to the comparison among the DTMs with different spatial resolution. Could it be better to use “comparison of landforms and processes detection using different spatial resolution LiDAR data” to replace the current Title? I think it will be more precise to describe the major work done in this MS. However, I leave the decision to the authors.

(2)   In the abstract, the authors mentioned that AI-driven approaches should be used rather than the expert judgment-based approaches. However, I notice that the authors did not address this issue in their study, the methodologies used in the MS is basic, while some manipulated thresholds were used to identify the process domains.  

(3)   The current version of Introduction section is satisfactory, as it only provided quite basic information of the technologies, in particular the paragraph from Line 52-72. The authors should be focused on the current studies and their limitation to highlight the novelty and creativities in this study.

(4)   Page 4, Line 123-124. “Given that the DTMs from LiDAR provided by the Ministry do not meet the requirements of most experts and users of cartography…”. Please explain that why the DTMs and LiDAR data does not meet the requirements. In my opinion, DTMs and LiDAR data is also the 3D representation of land.

(5)   Page 8, Line 244-253. In this paragraph, the authors mentioned three thresholds, please specify the details of these thresholds.

(6)   Page 13. Figure 10 should be redrawn. The texts in the current figure are too small to recognize.

 

(7)   One major limitation of this MS is lacking novelties. The authors provide plenty of information on the source of the data they used, how they analyze the data, as well as the results demonstrated in detail, which I appreciate a lot. However, the methodologies the authors used are quite common and basic in geomorphology studies, no improvement has been made, which makes the whole of this MS seem like a technical note, rather than a research article. The authors should highlight the novelty and creativities in this MS. 

Author Response

Comment 1: A key word in the title of this MS is “enhancing”, which I expect to see some improvement on the methodologies to enhance the landforms detection. However, as I read through the MS, the majority of this MS is with respect to the comparison among the DTMs with different spatial resolution. Could it be better to use “comparison of landforms and processes detection using different spatial resolution LiDAR data” to replace the current Title? I think it will be more precise to describe the major work done in this MS. However, I leave the decision to the authors.
Response 1: Thanks for the comment. The “enhancing” statement in the title was initially thought to highlight the fact that higher resolution data produced generally better results. However, after considerations, it is indeed true that within this manuscript we apply the same methodology on the different datasets and then we compare results, so there is not really an “enhancement” in the methodology. For this reason, we agree with the reviewer and the title as been changed to “Comparison of landforms and processes detection using multi-source data”.

Comment 2: In the abstract, the authors mentioned that AI-driven approaches should be used rather than the expert judgment-based approaches. However, I notice that the authors did not address this issue in their study, the methodologies used in the MS is basic, while some manipulated thresholds were used to identify the process domains.
Response 2: AI-driven approaches were only mentioned among the various existing techniques both in the abstract and the introduction. The intention of the authors was not to state which method is better. To better clarify this issue, the sentence in the abstract has been changed to: “Over the last decade, various methods have been developed to achieve this goal, ranging from grid-based to object-based approaches, covering a range from supervised to completely unsupervised techniques. Attention has also been given to the level of expert judgment involvement compared to AI-driven strategies.” (p. 1, lines 10-13 of the revised manuscript).

Comment 3: The current version of Introduction section is satisfactory, as it only provided quite basic information of the technologies, in particular the paragraph from Line 52-72. The authors should be focused on the current studies and their limitation to highlight the novelty and creativities in this study.
Response 3: The introduction section has been expanded, highlighting limitations of filtering techniques and better specifying the main goal of this work (p. 2, lines 72-86 of the revised manuscript).

Comment 4: Page 4, Line 123-124. “Given that the DTMs from LiDAR provided by the Ministry do not meet the requirements of most experts and users of cartography…”. Please explain that why the DTMs and LiDAR data does not meet the requirements. In my opinion, DTMs and LiDAR data is also the 3D representation of land.
Response 4: The section has been integrated to better explain this concept (p. 4, lines 139-144 of the revised manuscript).

Comment 5: Page 8, Line 244-253. In this paragraph, the authors mentioned three thresholds, please specify the details of these thresholds.
Response 5: Thresholds and their identification are detailed in the lines following the specified interval. However, various integrations have been made in the paragraph for a better understanding of the threshold’s definition (p. 9, lines 277-289 of the revised manuscript).

Comment 6: Page 13. Figure 10 should be redrawn. The texts in the current figure are too small to recognize.
Response 6: Figure 10 has been modified (p. 15 of the revised manuscript).

Comment 7: One major limitation of this MS is lacking novelties. The authors provide plenty of information on the source of the data they used, how they analyze the data, as well as the results demonstrated in detail, which I appreciate a lot. However, the methodologies the authors used are quite common and basic in geomorphology studies, no improvement has been made, which makes the whole of this MS seem like a technical note, rather than a research article. The authors should highlight the novelty and creativities in this MS.
Response 7: Thanks for this consideration. Novelties within this manuscript are represented by the original grid-based classification, originally presented in [10], but further modified using DEV instead of DIFF in the framework of this research. The definition of process domains is quite an established technique, but in [10] more thresholds, thus domains, have been added, and they are further explored in this research. Also, to our knowledge, the application of such grid-based techniques and processes detection on a 20 cm dataset is not so common in literature, and, in our opinion, working with such high-resolution datasets, results are not to be taken for granted.
Because this was not so evident from the initial draft, to highlight these considerations, several additions to the manuscript have been made:

• the sentence “using a modified and updated version of the procedures presented in [10]” ha been added (p. 2, lines 85-86 of the revised manuscript);
• the sentence “an original” has been added (p. 7, line 224 of the revised manuscript);
• the sentence “hereby presented” has been added (p. 8, line 238 of the revised manuscript);
• the sentence "for DIFF values” has been added (p. 8, line 250 of the revised manuscript);
• the sentence " For the identification of elementary landforms, an original classification procedure based on DEV values, instead of the more common DIFF, has been applied.” has been added (p. 20, lines 504-505 of the revised manuscript);
• the sentence “Analyses carried out on the 20 cm dataset highlighted that…” has been added (p. 20, line 508 of the revised manuscript).

Reviewer 3 Report

Comments and Suggestions for Authors

DTMs with different resolutions have been used for landform recognition. The figures of the manuscript are with high quality. However, the writing of the manuscript should be improved. What is the scientific question to be solved? What are the new method for automated recognition the landforms? The whole manuscript should be carefully checked. For example, line 315"Therefor the next peak,"; line 381"scores (AUC < 0,7)".

Author Response

DTMs with different resolutions have been used for landform recognition. The figures of the manuscript are with high quality. However, the writing of the manuscript should be improved.

Comment 1: What is the scientific question to be solved?

Response 1: Introduction has been expanded better stating aims and goals.

 

Comment 2: What are the new method for automated recognition the landforms?

Response 2: Novelties have been highlighted throughout the manuscript.

 

Comment 3: The whole manuscript should be carefully checked.

For example, line 315 "Therefor the next peak,"; line 381"scores (AUC < 0,7)".

Response 3: Manuscripts has been carefully checked.

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

I appreciate the improvement in the revised manuscript. The authors have addressed most of my initial suggestions adequately. I believe the revisions in the current manuscript make this a stronger paper. Readers should be better able to access the findings of their study with greater understanding. I suggest a possible publication in Remote Sensing after a minor revision.

General comments

(1) The authors have modified the title of this MS according to the comment, which I believe will be much more suitable to describe the major work in the MS. However, the revised title seems to short, and I suggest “Comparison of landforms and processes detection using multisource data: a case study in Cilento coastline, Southern Italy” for their consideration. As the authors used a single case for comparison, it should be better to specify the detailed area of the case.

(2) Line 12-13, “Attention has also been given to the level of expert judgment involvement compared to AI-driven strategies.” As I commented in the first review round, the MS does not involve any AI-driven strategies, it is not appropriate to mention AI-driven strategies in this sentence. I suggest the authors delete this sentence in their revision.

 

(3) Figure 7 in Page 12. The panels in the figure are unclear, in particular the texts. Please redraw this figure accordingly.  

Author Response

Comments 1: The authors have modified the title of this MS according to the comment, which I believe will be much more suitable to describe the major work in the MS. However, the revised title seems to short, and I suggest “Comparison of landforms and processes detection using multisource data: a case study in Cilento coastline, Southern Italy” for their consideration. As the authors used a single case for comparison, it should be better to specify the detailed area of the case.
Response 1: Thanks for Your comments. Title has been changed to "Comparison of landforms and processes detection using multisource remote sensing data: the case study of the Palinuro pine grove (Cilento, Vallo di Diano and Alburni National Park, Southern Italy)".

Comments 2: Line 12-13, “Attention has also been given to the level of expert judgment involvement compared to AI-driven strategies.” As I commented in the first review round, the MS does not involve any AI-driven strategies, it is not appropriate to mention AI-driven strategies in this sentence. I suggest the authors delete this sentence in their revision.
Response 2: As suggested, the sentence has been removed.

Comments 3: Figure 7 in Page 12. The panels in the figure are unclear, in particular the texts. Please redraw this figure accordingly.
Response 3: Figure 7, along with Figures 9 and 10 have been redrawn.