A Virtual Reality Simulation of a Real Landslide for Education and Training: Case of Chiradzulu, Malawi, 2023 Landslide

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

“A virtual reality simulation of a real landslide for education and training: case of Chiradzulu, Malawi 2023 landslide” is a great idea for an educational activity and even a performance! The title reflects what is proposed, even if the development of the manuscript does not always remain coherent. To demonstrate the correctness of the simulation it is necessary to intervene on some parts. The introduction highlights above all the advantage of modeling and visualizing phenomena in order to intervene promptly and effectively. However, in the case of landslides, a large amount of data is needed to reach this “level”. This is confirmed by numerous published articles that have gone a long way in modeling and visualizing (simply on the monitor) landslides, which could also be considered and added to the list of references. The reported materials only allow you to simulate a landslide in space through a comparison of images, for the rest of the data other sources are needed and should be reported. Meanwhile, in figure 3 the landslide should be indicated with a graphic symbol. Regarding the methods, the description does not highlight what information the various GIS applications need for simulation. It is not equally clear how the alternative scenarios were created, in which portion of the area affected by the landslide would the density of trees be increased? It would be interesting to visualize by adding a new figure what happens in one of these cases. Beyond the simplification of the phenomenon, in discussing the results you could highlight the parts that in your opinion do not fully respond to reality. It seems presumptuous to me to think that we can intervene with works without knowing the characteristics of the landslide in detail. As for the figures, I ask you to improve 1 and 3, and as already said I would add another to show the consequences of a landslide with vegetation "different" from the real one. Finally, the bibliography, although updated, lacks some recent articles (also published in MDPI journals) relating to space-time simulations of landslides that actually occurred, useful for understanding the difficulty of space-time modelling. For this reason, I propose to publish it after having made the requested changes (major revision). See the attached file

Comments for author File: Comments.pdf

Author Response

Dear Reviewer

Thank you for taking the time to peer-review our manuscript. We appreciated the constructive feedback. We have made some changes in the manuscript using red color. Our specific response to you comments can be found below. In addition, we have responded to your additional great feedback in the attached file again using red color to represent changes made relevant to each suggestions or comments.

Comment 1

A virtual reality simulation of a real landslide for education and training: case of Chiradzulu, Malawi 2023 landslide” is a great idea for an educational activity and even a performance! The title reflects what is proposed, even if the development of the manuscript does not always remain coherent. To demonstrate the correctness of the simulation it is necessary to intervene on some parts. The introduction highlights above all the advantage of modeling and visualizing phenomena in order to intervene promptly and effectively. However, in the case of landslides, a large amount of data is needed to reach this “level”. This is confirmed by numerous published articles that have gone a long way in modeling and visualizing (simply on the monitor) landslides, which could also be considered and added to the list of references.

We fully agree with you that some clarifications needed to be made. Unlike mechanistic/deterministic or probabilistic landslide models, our simulation was a visualization effort. We aimed to realistically visualize what happened (in our real landslide event) and connect it with the role that reforestation/deforestation can play in minimizing potential impacts.  To make this aspect more clear we have pointed out this in many relevant sections.  We particularly added to the sentence on lines 80-82 to include modelling and added a few additional references for readers.

The reported materials only allow you to simulate a landslide in space through a comparison of images, for the rest of the data other sources are needed and should be reported. Meanwhile, in figure 3 the landslide should be indicated with a graphic symbol. Regarding the methods, the description does not highlight what information the various GIS applications need for simulation. It is not equally clear how the alternative scenarios were created, in which portion of the area affected by the landslide would the density of trees be increased? It would be interesting to visualize by adding a new figure what happens in one of these cases. Beyond the simplification of the phenomenon, in discussing the results you could highlight the parts that in your opinion do not fully respond to reality. It seems presumptuous to me to think that we can intervene with works without knowing the characteristics of the landslide in detail. As for the figures, I ask you to improve 1 and 3, and as already said I would add another to show the consequences of a landslide with vegetation "different" from the real one.

We added a few lines in the data collection part to clarify the type of data collected for our purpose. Again it should be emphasized that we were not modeling the mechanics of the landslide. We made some changes in Figure 1, Figure 2 (caption) and Figure 3 (caption). It should be noted that in Figure 3, the whole image is the area impacted by the landslide flow, thus it is not possible to use a specific symbol to represent it. Among the scenarios we had, the one with fewer trees correspond to what indeed happened. We have clarified this better in our revisions. As recommended we also added another Figure (7) that represent the visualization of the actual landslide that is our scenario with fewer trees. 

Finally, the bibliography, although updated, lacks some recent articles (also published in MDPI journals) relating to space-time simulations of landslides that actually occurred, useful for understanding the difficulty of space-time modelling.

We have updated the manuscript to include an additional sentence on this body of literature and some recent examples (lines 79-82).

For this reason, I propose to publish it after having made the requested changes (major revision). See the attached file

Thank you for your comments in the attached file. You will see our responses addressed in the revised version for every point. We have used red color to specify where changes have been made.

Thank you very much.

Reviewer 2 Report

Comments and Suggestions for Authors

Dear authors,

in the introduction you list a few obstacles to the wider use of VR in geohazard research, but your own manuscript itself demonstrates that one of of the major points is : missing valuable 3D data and complexity of physical similations (in 3D), changing scale and resolution. These elements are barely discussed in your manuscript - or related information is even simply missing.

You took drone imagery ...but don't tell us at which resolution they were taken, scale information is missing on most of your views/maps. Those drone images were they used to create a high-resolution surface model?

And, according to your views you only worked with a game-like environment, probably using game physics. This certainly also influences the perception of the user - seeing a virtual real world .. or one that looks like a common game environment. In a game-like environment risk perception is highly reduced (especially for 'gamers').

If you used your drone imagery for 3D model creation and visualisation in VR, please, state this clearly - and provide views with scale and resolution information. If it was just used as a basis for setting up a game environment, please, provide arguments why you did so - as this environment is known of being of far lower value than the 'digital twin' of a real environment.

 

yours

reviewer H

Author Response

Dear Reviewer

Thank you for taking the time to peer-review our manuscript. We appreciated the constructive feedback. We have made some changes in the manuscript using red color. Our specific response to you comments can be found below. 

 

1. in the introduction you list a few obstacles to the wider use of VR in geohazard research, but your own manuscript itself demonstrates that one of the major points is : missing valuable 3D data and complexity of physical simulations (in 3D), changing scale and resolution. These elements are barely discussed in your manuscript - or related information is even simply missing.

Thank you very much for the note. Agree with the point. We believe that we have been able to address some of these obstacles in our visualization by using a realistic 3D model of the study area combined with field observations and visualizing the actual landslide using this 3D model. Of course the higher the resolution of the terrain the more computation power is needed to render and run the simulation in the current VR devices without lag and interruptions. We added a line in our discussion section to highlight this point (lines 310-312).   

2. You took drone imagery ...but don't tell us at which resolution they were taken, scale information is missing on most of your views/maps. Those drone images were they used to create a high-resolution surface model?

Drone imagery was not used to create the surface or 3D model in this case (although it could be used). However, we used drone images and mapping primarily as background information and creating the actual landslide geographic boundaries and extent, impacted areas and features, flow directions, type of materials, vegetation coverage and type, etc. for the VR application. Having said that we agree with the reviewer that this should be stated more clearly in the paper. To address this we particularly added a couple of lines in our methodology section (lines 17-169). 

3. And, according to your views you only worked with a game-like environment, probably using game physics. This certainly also influences the perception of the user - seeing a virtual real world .. or one that looks like a common game environment. In a game-like environment risk perception is highly reduced (especially for 'gamers').

Although we used Unity Game Engine to create the environment in the VR, we built the environment very much close to the real environment in terms of terrain, landcover, location of buildings, etc.  As such it gives a more realistic impression of the environment compared to typical game environments. We added a few lines clarify this further (Lines 178-180).

4. If you used your drone imagery for 3D model creation and visualisation in VR, please, state this clearly - and provide views with scale and resolution information. If it was just used as a basis for setting up a game environment, please, provide arguments why you did so - as this environment is known of being of far lower value than the 'digital twin' of a real environment.

Thank you very much for the comment. As mentioned earlier we did not use the drone image for the 3D model. We used drone imagery mostly for creating the actual landslide boundary areas, impacts, materials, land cover, and background analysis. We made this clear in the text. Also changes figure captions where we references the drone images. We also mentioned the type of drone used to capture the images and create the map based on the drone images.

Thank you very much.

 

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

I confirm my positive opinion as an educational activity of the manuscript ‘A virtual reality simulation of a real landslide for education and training: case of Chiradzulu, Malawi 2023 landslide’ that was returned to me. You have succeeded in intervening on the text and figures following most of my requests. I would like to emphasise that some aspects that are more important for scientific publication than for the success of the simulation needed to be deepened (e.g. characters of the landslide in space, accuracy of the method, restitution of information that is not only visual of the landslide's mode of incidence). I propose to publish it in this version without further modifications.

Author Response

Comments and Suggestions for Authors

I confirm my positive opinion as an educational activity of the manuscript ‘A virtual reality simulation of a real landslide for education and training: case of Chiradzulu, Malawi 2023 landslide’ that was returned to me. You have succeeded in intervening on the text and figures following most of my requests. I would like to emphasise that some aspects that are more important for scientific publication than for the success of the simulation needed to be deepened (e.g. characters of the landslide in space, accuracy of the method, restitution of information that is not only visual of the landslide's mode of incidence). I propose to publish it in this version without further modifications.

Response:

Thank you very much for the feedback and accepting the paper for publication. However, we hope that some of the additional changes that we made in response to reviewer 2 has enhanced the paper.

Reviewer 2 Report

Comments and Suggestions for Authors

Dear authors,

as you did not provide any details about the surface modelling and the landslide simulations (besides the fact that you did not use drone imagery for the modelling), the manuscript as practically not been improved. I still have the impression that you used game physics for the simulations .. and then .. you manuscript does not fit in a 'geohazards' journal.

yours

reviewer H

Comments on the Quality of English Language

no special comment

Author Response

Comments and Suggestions for Authors

Dear authors,

as you did not provide any details about the surface modelling and the landslide simulations (besides the fact that you did not use drone imagery for the modelling), the manuscript as practically not been improved. I still have the impression that you used game physics for the simulations .. and then .. you manuscript does not fit in a 'geohazards' journal.

yours

reviewer H

Response:

Dear Reviewer. Thank you very much for your feedback. We apologize if the revisions were not satisfactory to you. Respectfully we think that there is a misunderstanding between us. The misunderstanding in our view is probably rooted in:

1. “numeric modeling” vs “nonnumeric modeling” of different phenomenon. In this VR application we neither intend nor used a numeric modeling approach to model the landslide because our application’s primary goals were to provide a conceptual understanding of the landslide and educate users about their impacts under different conditions. For such purposes we do not need to have numeric, or data driven models. Having said that we agree that we must be clearer on this. As such the following was added to the Methods section along with a new figure to explain our approach:

While there are different methods to model natural hazards such as landslides for virtual reality applications including numeric and non-numeric models, in this study we use non numeric approach because our goal is to provide a conceptual understanding of the landslide, educate users about potential impacts of landslides and how human interventions in the environment (e,g. deforestation) increases the level of vulnerability and damages.     

We also added the following at the end of the Discussion section to emphasize the importance of using numeric models in future versions of the application:

Finally, it is important to mention that the application developed here was not based on a numerical model of the landslide. A future enhancement of this application could benefit from a numeric model based on geological, topographical, hydrological, meteorological, seismic, land cover and human factors. Such numerical model can provide a more data driven basis for risk mitigation and reduction measures that can be also visualized in before their implementation. While incorporating such models in virtual reality is still challenging, it is expected that with the rapid improvements in VR hardware and software, most of these challenges will be resolved.        

 

2. Scope. Respectfully we disagree with the notion that “you used game physics for the simulations .. and then .. you manuscript does not fit in a 'geohazards' journal”. The journal Aims and Scope among other items has:

“Applications: detecting, monitoring, mapping, and modeling hazards (frequencies and magnitudes; spatial and temporal variability); risk assessment and damage scenarios; multi-hazard analyses and cascading effects; risk and vulnerability reduction; physical, social and economic resilience; disaster response and reconstruction; social exposure; and risk perception, communication and education.”

We believe that our paper is very much within the scope of the journal by providing an communicational and educational tool for the geohazards that can help long term risk and vulnerability reduction, and enhance risk perception. These are the areas that are often neglected.

Thank you very much.