Elevation Accuracy of Forest Road Maps Derived from Aerial Imaging, Airborne Laser Scanning and Mobile Laser Scanning Data

Round 1

Reviewer 1 Report

I want to thank you for the submitted paper "Elevation Accuracy of Forest Road Maps Derived from Aerial Imaging, Airborne Laser Scanning and Mobile Laser Scanning Data". Your work provides important insight into the precision of elevation data on forest road maps, which is crucial for sustainable forest management.

I was particularly impressed by your methodology for evaluating and comparing the precision of height data from different remote sensing technologies on three types of forest roads.

In addition, your observations on differences in precision depending on road type and scanning technology provide useful guidance for future researchers and practitioners in the field of forest management.

Given the importance of your work, I believe it will be of interest to Forests magazine readers. However, before we decide to accept the work for publication, I would ask you to consider and specify the answer to the question: Do you have data on the influence of weather conditions on the accuracy of the collected data and whether there was cloud cover at the time of aerial photography and whether it is a possible cause of worse data AI and time itself?

Other remarks and comments can be found in the pdf document of your text.

With respect

Comments for author File: Comments.pdf

Author Response

Dear reviewer,

We would like to thank you for the review of our manuscript. We studied and considered each of suggestion as well as comment very carefully.

In this letter, we report all revisions of the original manuscript related to the review. Each response contains a line number and description of related revision. In cases of disagreement or misunderstanding we tried to explain our point of view. However, we have been able to incorporate revisions to reflect most of the suggestions and comments provided. In the current version of manuscript are these revisions highlighted using the track changes function.

Reviewer comment: Do you have data on the influence of weather conditions on the accuracy of the collected data and whether there was cloud cover at the time of aerial photography and whether it is a possible cause of worse data AI and time itself?

Author response: Our study contains no special data to evaluate the influence of weather on elevation accuracy. However, the time and related conditions are chosen in order to acquire the RS data that can be used for the forestry mapping and forest inventory, thus the accuracy and good visibility of objects is a main point. This means cloudless, clear weather with an imaging date in spring or summer, with imaging between 10am – 14pm to avoid shadows of objects. We added/highlighted this information across the text of individual sections (i.e., 2.2-2.6) within the section Material and methods. Please see L135, L157, L167, L177, L194.

Reviewer comment: enter a total of 150 m, total of 100 m

Author response: Done. Please see L129-130.

Reviewer comment: Please write in a little more about the details of the GNSS instrument, its factory accuracy in the horizontal and vertical sense, and how much accuracy is guaranteed by the local SKPOS base system. And then take into account both possible measurement errors. Likewise, when it was surveyed, were the recordings made at the same time and under similar conditions when the position of the satellites was the same or not? State the model of the total station used in the measurement and the measurement system, was a prism used or some other method?

Author response: We did not write much information about this part of the manuscript, as we did not want to make it too long. However, it is true that specific information related to the devices was missing there. We added this information in section 2.2. Please see L137-138. Moreover, we improved the description of some parts of ground survey, such as corrections and accuracy. Please see L143-146.

Reviewer comment: Also, write the distance between the points within the profile, and whether this distance was always of the same value or the approximate distance between the points within one profile was determined.

Author response: Done. Please see L148-149.

Reviewer comment: Explain how the classification of ground points was derived and whether they were used in AI method ground control points, because without well-defined points, it is not possible to expect that the results of the model will be more precise.

Author response: Considering the length of manuscript and amount of technical details, we chose a compromise. We have provided key information and references. These references include all details. However, it’s true, this is extremely relevant point. Therefore, we improved description of all process, such as processing and post-processing of AI, ALS, and MLS data. Please see L161, L170, L180.

Reviewer comment: in the results, explain why the biggest error occurred on the asphalt road.

Author response: Extremely relevant point! This result was unexpected. However, it is confirmed by several independent assessments. We consulted the possible causes in a wider team of expert related to the RS. Based on this, we concluded that the main cause is the surface texture. Traditionally, the image correlation provides better results on surfaces with a heterogeneous texture, which, in our case, is a typical stone surface, compared to less accurate results obtained on a homogeneous surface (asphalt or panel). This practical experience with image matching for terrain generation in digital photogrammetry is also supported by the other authors e.g. Rahmayudi and Rizaldy (2016). We highlighted this result in discussion section. Please see L348-350.

Reviewer comment: quality of English language

Author response: All authors confirm that the manuscript was proof-read by MDPI Editing Services (ID: English-78592). The text has been checked for correct use of grammar and common technical terms, and edited to a level suitable for reporting research in a scholarly journal.

Yours sincerely,

Ivan Sačkov , Corresponding author

Author Response File: Author Response.pdf

Reviewer 2 Report

I have read with great interest the manuscript about the determination of the vertical profile of forests roads using diverse remote sensing techniques. The use of English is correct enough. The authors can find some comments aimed at helping them to improve the manuscript:

Certainly, the manuscript already contains many appropriate references on the topic addressed. While reading it, an article that dealt with similar issues came to my mind, in which the accuracy of the virtual recreation of a highway surrounded by a forest using UAV and SfM methods turned out to be critical, and particularly the elevation accuracy of the digital elevation model.

Iglesias, L.; De Santos-Berbel, C.; Pascual, V.; Castro, M. Using Small Unmanned Aerial Vehicle in 3D Modeling of Highways with Tree-Covered Roadsides to Estimate Sight Distance. Remote Sens. 2019, 11, 2625. https://doi.org/10.3390/rs11222625

The study area is fairly well described and graphically documented with the figures of the manuscript. However, I believe that the manuscript should include close-up images showing the macrotexture of the surfaces studied.

The road areas studied are small, in spite of the statistical analyses carried out on them. This is a severe limitation for the study and its contributions.

The equipment used for the data collection is described with sufficient detail.

Line 128. What do authors mean by pavement? Please, note that asphalt concrete and concrete are two types of pavements. Maybe granular material without binder? Other?

Figure 2 shows orthoimages with roads near forest with deciduous trees with overhanging branches. Were all the surveying data collected during the same period? Please, note that the tree foliage might affect the output data.

Line 175, section 2.6. The manuscript would benefit from a more detailed description of the tools and procedures used in ArcMap to produce the forest road maps. Particularly, while each data source may have different output formats the elevation models to be compared must be provided in the same format. Flowcharts of the procedures would help.

The statistical tests conducted seem appropriate for the purpose of the study. Nevertheless, I recommend the authors to illustrate the comparison of the elevation accuracy with vertical profiles, easily created in ArcMap with profile graphs on the TIN surfaces and superimpose the profiles with the three surveying techniques used:

https://desktop.arcgis.com/en/arcmap/latest/extensions/3d-analyst/creating-a-profile-graph-from-digitized-features-of-a-surface.htm

Previous studies have showed that the terrain relief may affect the elevation accuracy of elevation models produced from airborne laser scanning. The more rugged the relief, the smaller the elevation accuracy.

The authors should provide qualitative and quantitative data of how undulated or rugged the terrain of the case studies is. Therefrom, the authors could provide the range of validity of the elevation accuracy results and conclusions drawn.

Although the conclusions are supported by the results, the contribution of the study to the state of knowledge is scarce due to the limitation above highlighted.

Technical terms such as "pavement" should be refined.

Author Response

Dear reviewer,

We would like to thank you for the review of our manuscript. We studied and considered each of suggestion as well as comment very carefully.

In this letter, we report all revisions of the original manuscript. Each response contains a line number and description of related revision. In cases of disagreement or misunderstanding we tried to explain our point of view. However, we have been able to incorporate revisions to reflect most of the suggestions and comments provided. In the current version of manuscript are these revisions highlighted using the track changes function.

Reviewer comment: Certainly, the manuscript already contains many appropriate references on the topic addressed. While reading it, an article that dealt with similar issues came to my mind…

Author response: We added suggested reference. Please see L62.

Reviewer comment: The study area is fairly well described and graphically documented with the figures of the manuscript. However, I believe that the manuscript should include close-up images showing the macrotexture of the surfaces studied.

Author response: We added suggested visualization. Please see Fig 2 representing textures of all road surfaces (i.e., asphalt, concrete, and stone) created from the intensity value of the ALS data.

Reviewer comment: The road areas studied are small, in spite of the statistical analyses carried out on them. This is a severe limitation for the study and its contributions.

Author response: The study is a first step and consists of an analysis prepared on 1st class Slovakian forest roads. Based on the verified methodology, we would like to follow up in our further research on lower class road (2nd and 3rd categories) using available remote sensing techniques. Your comments are valuable to us and will be included in the next stage of the project, with further data samples to analyse. We hope, that for the first step, the provided data sample (7 test sections) meets the criteria for statistical inference, and thus, is enough to prepare the conclusions.

Reviewer comment: The equipment used for the data collection is described with sufficient detail

Author response: Considering the length of manuscript and amount of technical details, we have provided key information and references. Moreover, we added new information in the current version of manuscript.

Reviewer comment: Line 128. What do authors mean by pavement? Please, note that asphalt concrete and concrete are two types of pavements. Maybe granular material without binder? Other?

Author response: We consulted alternative terms for all road surfaces with experts related to the forest transportation. Based on this, we changed term “pavement” to term “stone” throughout the manuscript.

Reviewer comment: Figure 2 shows orthoimages with roads near forest with deciduous trees with overhanging branches. Were all the surveying data collected during the same period? Please, note that the tree foliage might affect the output data.

Author response: The time and related conditions are chosen in order to acquire the RS data that can be used for the forestry mapping and forest inventory, thus the accuracy and good visibility of objects is a main point. This means cloudless, clear weather with an imaging date in spring or summer, with imaging between 10am – 14pm to avoid shadows of objects. Moreover, AI and ALS data were obtained simultaneously. We added/highlighted this information throughout the text of individual sections (i.e., 2.2-2.6) within the section Material and methods. Please see L135, L157, L167, L177, L194.

Reviewer comment: Line 175, section 2.6. The manuscript would benefit from a more detailed description of the tools and procedures used in ArcMap to produce the forest road maps. Particularly, while each data source may have different output formats the elevation models to be compared must be provided in the same format. Flowcharts of the procedures would help.

Author response: Considering the length of manuscript and amount of technical details, we chose a compromise. We have provided key information and references. These references include all details. However, it’s true, this is extremely relevant point. Therefore, we improved description of all process, such as processing and post-processing of AI, ALS, and MLS data. Please see L161, L170, L180.

Reviewer comment: The statistical tests conducted seem appropriate for the purpose of the study. Nevertheless, I recommend the authors to illustrate the comparison of the elevation accuracy with vertical profiles, easily created in ArcMap with profile graphs on the TIN surfaces and superimpose the profiles with the three surveying techniques

Author response: We added suggested visualization. Please see Fig 4 representing longitudinal profiles of all road surfaces (i.e., asphalt, concrete, and stone).

Reviewer comment: Previous studies have showed that the terrain relief may affect the elevation accuracy of elevation models produced from airborne laser scanning. The more rugged the relief, the smaller the elevation accuracy.

Author response: Our study contains no special data to evaluate the influence of terrain relief on elevation accuracy. However, our results generally confirmed this hypothesis. The highest precision was found on asphalt roads derived from MLS and ALS. On the other hand, the lowest precision was found on all roads derived from AI data.

Reviewer comment: The authors should provide qualitative and quantitative data of how undulated or rugged the terrain of the case studies is. Therefrom, the authors could provide the range of validity of the elevation accuracy results and conclusions drawn.

Author response: Extremely relevant point! We added suggested information in terms of confidence interval for accuracy attributes and statistical tests. Please see L206, L209.

Reviewer comment: Although the conclusions are supported by the results, the contribution of the study to the state of knowledge is scarce due to the limitation above highlighted

Author response: We have tried to provide the most useful study possible. The motivation for realization resulted mainly from the following points: (1) relevant stakeholders need and therefore strictly require the forest road map containing precise information about related network of forest roads, (2) elevation accuracy of terrain primarily determines the overall precision and thus the usability of the forest road maps, and (3) aerial imaging, airborne laser scanning, and mobile laser scanning represent most innovative remote sensing technologies to acquire information about terrain on forest roads.

Reviewer comment: quality of English language

Author response: All authors confirm that the manuscript was proof-read by MDPI Editing Services (ID: English-78592). The text has been checked for correct use of grammar and common technical terms, and edited to a level suitable for reporting research in a scholarly journal.

Yours sincerely,

Ivan Sačkov, Corresponding author

Author Response File: Author Response.pdf

Reviewer 3 Report

The main purpose of the presented study is to evaluate and compare the accuracy of determining terrain heights on forest road maps obtained on the basis of data from three remote sensing technologies (AI, ALS and MLS) using five geospatial methods. The study was conducted on three different forest road surfaces (i.e. asphalt, concrete and pavement).

The study of the accuracy of determining heights by various modern methods of three types of logging road coatings is a scientific novelty.

The research corresponds to the subject of the journal. The presented conclusions are substantiated and confirmed by the results of the study. The article provides a good critical comparison of the results obtained with those previously obtained by other researchers.

The article is written in a good scientific understandable language, well structured. The materials and methods of research are sufficiently written. In the conclusions, the authors present good recommendations on the use of remote sensing methods under various conditions. This will attract readers not only from the scientific field, but also from industrial enterprises.

Author Response

Dear reviewer,

We would like to thank you for the review of our manuscript.

We have tried to provide the most useful study possible. The motivation for realization resulted mainly from the following points:

(1) relevant stakeholders need and therefore strictly require the forest road map containing precise information about related network of forest roads,

(2) elevation accuracy of terrain primarily determines the overall precision and thus the usability of the forest road maps,

(3) aerial imaging, airborne laser scanning, and mobile laser scanning represent most innovative remote sensing technologies to acquire information about terrain on forest roads.

The current version of the manuscript has been further improved. We were able to incorporate revisions of our manuscript to reflect most of the suggestions and comments from available reviews.

All authors confirm that the manuscript was proof-read by MDPI Editing Services (ID: English-78592). The text has been checked for correct use of grammar and common technical terms, and edited to a level suitable for reporting research in a scholarly journal.

Yours sincerely,

Ivan Sačkov, Corresponding author

Author Response File: Author Response.pdf

Reviewer 4 Report

The manuscript`s methodology appears to be weak and its findings seem to offer little in terms of originality to the scientific community.

Extensive editing of the English language is required.

Author Response

Dear reviewer,

We would like to thank you for the review of our manuscript. In this letter, we tried to explain our point of view. Please consider our arguments and current version of manuscript.

Reviewer comment: The manuscript`s methodology appears to be weak and its findings seem to offer little in terms of originality to the scientific community.

Author response (I. Sačkov): We have tried to provide the most useful study possible. The motivation for realization resulted mainly from the following points: (1) relevant stakeholders need and therefore strictly require the forest road map containing precise information about related network of forest roads, (2) elevation accuracy of terrain primarily determines the overall precision and thus the usability of the forest road maps, and (3) aerial imaging, airborne laser scanning, and mobile laser scanning represent most innovative remote sensing technologies to acquire information about terrain on forest roads.

The current version of the manuscript has been further improved. We were able to incorporate revisions to reflect most of the suggestions and comments from available four reviews.

Author response (L. Borowski): The manuscript is part of a broader analysis conducted for the Slovak National Forest Centre (NFC), which oversees a significant portion of Slovak forestry. The NFC's adoption of an appropriate model for gathering and processing information on forest road conditions is crucial for effective forest management, sustainable growth, and land protection initiatives. Our analysis aimed to provide insights into determining the most suitable model for forest road management, rather than developing the chosen methods (AI, ALS, and MLS). Therefore, we focused on analysing these three methods about the typical Slovakian forests land and field conditions, seeking concrete answers about their applicability to forest roads.

To achieve this, we conducted a rigorous experiment employing statistical inference, aligning with established benchmarks within the engineering discipline (geomatics). Thus, the results obtained underwent validation through the scientific method, providing a reliable basis for the NFC (or similar institutions in other countries) to make informed decisions regarding the selection of a model for forest road inventory. In this context, we hold a different perspective on the originality of our findings. While our research may not introduce novel concepts (we based on broadly used methods), it significantly contributes to existing knowledge by offering comprehensive insights into the accuracy of forest road inventories derived from diverse data sources.

We have underscored the original aspects of our study and highlighted its significance in enhancing understanding of forest mapping techniques and their practical implications, particularly in the realm of forest road inventory management. We hope that these findings will be duly considered by the National Forest Centre, and other centres or facilities, in their ongoing work to improve forest management practices.

Reviewer comment: Extensive editing of the English language is required.

Author response: All authors confirm that the manuscript was proof-read by MDPI Editing Services (ID: English-78592). The text has been checked for correct use of grammar and common technical terms, and edited to a level suitable for reporting research in a scholarly journal.

Yours sincerely,

Ivan Sačkov, Corresponding author

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

There are some comments that remain unresolved:

Figure 2 should include a graphical scale.

Figure 3 should include the depiction of the horizontal projection of the line whose profile has been retrieved on each elevation model plan.

The authors have not provided qualitative and quantitative data of how undulated or rugged the terrain of the case studies is.

Include in future lines of research a discussion on the need for further verifications, as they have not been addressed: The expansion of the study to broader areas, the fact that the terrain relief may affect the elevation accuracy of elevation models produced from airborne laser scanning.

Author Response

Dear reviewer,

We would like to thank you for the review of our manuscript. We studied and considered each of suggestion as well as comment very carefully.

In this letter, we report all revisions of the original manuscript. We have been able to incorporate revisions to reflect all of the suggestions and comments provided. In the current version of manuscript are these revisions highlighted using the track changes function.

Reviewer comment: Figure 2 should include a graphical scale.

Author response: We added required graphical scale. Please see Fig. 02a-c.

Reviewer comment: Figure 3 should include the depiction of the horizontal projection of the line whose profile has been retrieved on each elevation model plan.

Author response: We added required depictions. Please see Fig. 05a-c, 05e-g, 05j-k.

Reviewer comment: The authors have not provided qualitative and quantitative data of how undulated or rugged the terrain of the case studies is.

Author response: (1) We added new figure representing the ruggedness of road surfaces. Please see Fig. 02d showing a variability of elevation on forest roads (i.e., mean ±three times the standard deviation). (2) We added new sentence in conclusion section containing information that roads from this study were fully paved haul forest roads with a uniform surface and a longitudinal slope of up to 10%. Please see L424-425. However, we are not entirely sure if it is exactly what you requested. If you meant something else, please specify what means “qualitative and quantitative data”.

Reviewer comment: Include in future lines of research a discussion on the need for further verifications, as they have not been addressed: The expansion of the study to broader areas, the fact that the terrain relief may affect the elevation accuracy of elevation models produced from airborne laser scanning.

Author response: (1) We added new paragraph in conclusion section containing description of general contribution of this study and associated future lines. Please see L422-428. (2) The fact that the terrain may affect the elevation accuracy is mentioned in several places throughout the manuscript. In current version, we highlighted this fact even more. For example, please see L42, L103, L129, L265, L356, L427.

Reviewer comment: quality of English language

Author response: This study has been proof-read by MDPI Editing Services (ID: English-78592). Moreover, the current version has been additionally proof-read by professional EN editor. The text has been checked for correct use of grammar and common technical terms, and edited to a level suitable for reporting research in a scholarly journal.

Yours sincerely, Ivan Sačkov

Corresponding author

Author Response File: Author Response.pdf

Reviewer 4 Report

I am afraid that this paper has nothing to offer in improving any forest management practice, despite the use of multimodal data sources. 

There wasn`t any improvement in English language.

Author Response

Dear reviewer,

Even if your review is strictly negative, we would like to thank you for your opinion. Unfortunately, it seems that there is no way to change this. Our arguments, clarifications and corrections of manuscript are insufficient for you.

However, we still believe that this study can be beneficial for relevant stakeholders and valuable source of objective and scientific information for the special issue "Recent Advances in Forests Roads Research” of scholarly journal “Forests” (ISSN 1999-4907).

There are following key reasons:

(1) This study is realized based on needs of relevant stakeholders which strictly require the forest road map containing precise information about related network of forest roads.

(2) This study is the work of five authors from three institutions. The author's team consists of relevant academic and scientific workers with many years of knowledge and experience in R&D.

(3) This study fulfils the MDPI as well as general requirements for the scientific article, such as scope, structure, format, style, and language.

(4) This study has been thoroughly reviewed. Three reviewers provided clearly defined comments or suggestions. Based on this, this study was corrected and improved.

(5) This study has been proof-read by MDPI Editing Services (ID: English-78592). Moreover, the current version has been additionally proof-read by professional EN editor. The text has been checked for correct use of grammar and common technical terms, and edited to a level suitable for reporting research in a scholarly journal.

(6) This study provided clear and scientific information related to the elevation accuracy of forest road maps derived from aerial imaging, airborne laser scanning, and mobile laser scanning data, thus most innovative remote sensing technologies.

In any case, we are able to correct/improve this study. However, we need clear defined shortcomings, notes, requirements or suggestions. If you could provide these points, we might be able to improve the study according to your preferences.

Yours sincerely, Ivan Sačkov

Corresponding author

Author Response File: Author Response.pdf