Combination of UAV and terrestrial photogrammetry to  assess rapid glacier evolution and map glacier hazards

Fugazza, Davide; Scaioni, Marco; Corti, Manuel; D'Agata, Carlo; Azzoni, Roberto Sergio; Cernuschi, Massimo; Smiraglia, Claudio; Diolaiuti, Guglielmina Adele

doi:https://doi.org/10.5194/nhess-18-1055-2018

Articles | Volume 18, issue 4

https://doi.org/10.5194/nhess-18-1055-2018

© Author(s) 2018. This work is distributed under
the Creative Commons Attribution 3.0 License.

Special issue:

The use of remotely piloted aircraft systems (RPAS) in monitoring...

https://doi.org/10.5194/nhess-18-1055-2018

© Author(s) 2018. This work is distributed under
the Creative Commons Attribution 3.0 License.

Articles | Volume 18, issue 4

Research article

|

05 Apr 2018

Research article |

| 05 Apr 2018

Combination of UAV and terrestrial photogrammetry to assess rapid glacier evolution and map glacier hazards

Davide Fugazza, Marco Scaioni, Manuel Corti, Carlo D'Agata, Roberto Sergio Azzoni, Massimo Cernuschi, Claudio Smiraglia, and Guglielmina Adele Diolaiuti

Download

Final revised paper (published on 05 Apr 2018)
Preprint (discussion started on 06 Jun 2017)

Interactive discussion

Status: closed

AC: Author comment | RC: Referee comment | SC: Short comment | EC: Editor comment

- Printer-friendly version

- Supplement

RC1: 'Referee comment of the paper by Fugazza et al.', Anonymous Referee #1, 16 Jul 2017
- AC1: 'Point by point response to reviewer 1', Davide Fugazza, 24 Sep 2017
RC2: 'Review of the manuscript: Combination of UAV and terrestrial photogrammetry to assess rapid glacier evolution and conditions of glacier hazards. Fugazza et al.', Saskia Gindraux, 18 Jul 2017
- AC2: 'Point by point response to reviewer 2', Davide Fugazza, 24 Sep 2017

Peer-review completion

AR: Author's response | RR: Referee report | ED: Editor decision

ED: Reconsider after major revisions (further review by Editor and Referees) (27 Sep 2017) by Paolo Tarolli

AR by Davide Fugazza on behalf of the Authors (18 Oct 2017) Author's response Manuscript

ED: Referee Nomination & Report Request started (23 Oct 2017) by Paolo Tarolli

RR by Saskia Gindraux (03 Nov 2017)

Suggestions for revision or reasons for rejection

Review of the manuscript:
Combination of UAV and terrestrial photogrammetry to assess rapid glacier evolution and conditions of glacier hazards. Fugazza et al.

General comments:
In this manuscript, the authors describe and analyse geomorphological features on the tongue of a hazard-prone glacier in the Italian alps with the help of three different (close-range) remote sensing methods. They found that UAV- and terrestrial photogrammetry are the best surveying techniques to assess ice thickness change and map glacier hazards.
Compared to the first manuscript, the authors reduced significantly the length of the manuscript which was preconized by both reviewers, set the aims more clearly and re-organized the different sections in a much better way. However, there are still several points of the paper that require substantial improvements, such as:

1) This version of the paper in my opinion, is poorly written. I had large difficulties understanding several sections of the manuscripts, due to the use of unconventional words and wrong sentence construction. As this is not my mother-tongue either, I could not give suggestions for every case and did not highlighted all of them. A English proof reading is surely needed before considering this paper for publication.

2) The individual methods and comparison of the methods used as suggested in this version of the manuscript are not new. The only aspect that could have been interesting and relatively new, is the merging of different datasets. However, the authors only merged the terrestrial and UAV photogrammetry point clouds, without giving a quantitative explanation of how this merged point cloud is much better than the terrestrial one alone. Looking at Figure 6, and the very sparse point cloud they obtained from UAV photogrammetry compared to terrestrial photogrammetry, I doubt that the merging of both point clouds did make a big difference. It would however be good if the authors could give more information on this point. Moreover in this study, this merged point cloud is only used to map the hazards (along with the orthophoto), which I guess, could have also been done with the point cloud from one method only (the terrestrial photogrammetry). So I am not sure if this merged point cloud was really necessary. I think the authors should emphasis more on the scientific value of this study.

3) I think that the authors wrote a lot about the differences between the methods they used, but very little refer to other literature. There are a lot of papers (below only a few for example) comparing point clouds and DEMs generated from different surveying methods.
a. Baltsavias, 1999: A comparison between photogrammetry and laser scanning
b. Rayburg et al. 2009: A comparison of digital elevation models generated from different data sources
c. Naumann et al. 2013: Accuracy comparison of digital surface models created by unmanned aerial systems imagery and terrestrial laser scanner
In most sections of the results and discussion, the authors report results that have already been found in other studies. For instance in the discussion section, where the advantages and drawbacks from all methods are explained, the authors cite publications that found the same results (but several years ago). As several papers are already stating these, I think this information should not be discussed anymore, but taken as granted. I suggest the authors to integer older and newer publications related to the comparison of point clouds and DEM specifically.

4) To my point of view, the authors are often making statements such as: “The UAV-based DEMs hold the potential to become a standard tool to investigate geodetic mass balance”, but the authors did not try to do this and the publications that succeeded to do it are very rare. Another one: “The final accuracy of our UAV photogrammetric products was nevertheless adequate to investigate ice thickness change over two years,…”. The reader don’t know for what it is “adequate”, as the authors found DEM errors over two meters (for these two years). They do not give any percentage error that would mean on the total melt. Moreover, the authors do not state why they need ice thickness data. We only know that ice thickness change is related to an increase of natural hazards.
I think in general, the manuscript needs to be more carefully written, with results presented in a more quantitative robust way.

I think the manuscript requires again major revision before being considered for publication in NHESS. More specific and short comments are reported in the supplementary material as .pdf.

Comments on Figures and Tables:

Figure 1:
- I would zoom in more as there is a lot of space under the reference area, and eventually show the positions of the terrestrial pictures.
- It’s hard to make the difference between the green triangle and the green points. Can you change the colour? If you do this, I think you can also remove the “ (in 2016 two different…)” in the caption, because we can see it on the map.
- Is the number after +/- based on 1 standard deviation? Or 2?

Figure 2:
- It is very hard for the reader to have an idea about the size of the feature. Could you insert a scale?
- This Figure shows the hazardous features on the glacier but these are not the ones that you survey and studied. So I am wondering what kind of information the reader gets out of this Figure. I would maybe recommend to put them on the right side of Fig.1 where there is some space left before the caption’s end, and delete the Figure 2.

Figure3:
- Very clean Figure 
- I would only centre the b in the white case

Figure 4:
- I think this figure does not give much information to the reader. My suggestion is that you either put the location of the images on Fig 1., or that you merge a and b.
- Caption: Small question in (a). In the text you say that you took 134 pictures. Are they all displayed here? I have the feeling that they don’t, so did they all align in the software?

Figure 5:
- The background image is very dark. It would be nice to see more lighter colours
- Also here on the image it is hard to see how big these features are. As you don’t have much space, I suggest you make a similar scale everywhere and that you add it on top of the 100 scale bar and you show the number here.

Figure 6:
- Caption: Maybe you could mention that the scale bars don’t have the same scale?

Figure 7:
- Caption: “L” in Location
- In the caption and in the main text you use orthomosaic and orthophotos. Please stick to one term.
- I would add after (a) and (b), situation in 2014 (situation in 2016) or something similar, so that you don’t start with a year.

Figure 8:
- Maybe consider to change the total ice thickness change! In the manuscript you are stating values to -30 and -50m that we don’t see on the map! The reader needs to calculate if he/she has the yearly values.

Table 1:
- In the caption, it would be nice if you could state something like:” DEM 2007 from aerial multispectral survey, DEM2014 and DEM 2016 from UAV photogrammetry.” So that the reader do not need to go back to the text to remember which DEM is which.

Table 2:
- This table definitely need some adjustments, because it’s not easy to read.
1. The “sample window” text could be rotated 90°, and use all space above the numbers (merge cells).
2. The meaning of k should be explained in the caption
3. You could choose 3 abbreviations (UAV, TP and TLS) in the table and explain them in the caption, so that the text is less squeezed.
4. Watch that you use capital letter at the beginning or you text everywhere.
5. The numbers with 1645+/- 54 need to be in one line! Otherwise the reader asks: what is the number below it? There seems to be a bit of space left on the right side of your table to enlarge it (till the level of your caption right).

Table 3:
1. The “sample window” text could be rotated 90°, and use all space above the numbers.
2. Same comment as above with the abbreviations
3. Explain what is “–“ and Ref. in the caption?
4. The caption could be a bit more elaborated!

Table 4:
- The text could be all set on left side of the cell
- Caption: Could you explain what is sigma?

Referee Report: PDF

Hide

RR by Anonymous Referee #3 (05 Dec 2017)

Suggestions for revision or reasons for rejection

The authors have benefited from two sets of very thorough reviews on their original manuscript. I have read these reviews and the authors’ responses, which are mostly appropriate and well-justified. Importantly, the authors have shortened many sections and undertaken some restructuring which has improved the flow and clarity of the manuscript.

I have provided a few additional points for consideration prior to publication. These are minor and should not take the authors very long to address.

Line 50 – ‘sensing’, not ‘Sensing’

Line 63 and throughout – ‘photogrammetry’, not ‘Photogrammetry’

Line 69 – remove ‘slowly’

Line 74 – would be useful to quantify ‘rapidly downwasting’ here – approximately how much surface lowering/terminus retreat is there per year?

Line 102 – is it appropriate in NHESS for authors to list articles that have been ‘submitted’? ‘In press’ or ‘Accepted’ – yes; ‘submitted’ – I’d perhaps not be comfortable with including this. This requires an Editorial decision.

L160 – I assume that no independent check data were acquired to truly test the ‘accuracy’ of the UAV-SfM data during the model generation stage? If so, the wording needs changing here – the RMSE for the PhotoScan ‘markers’ in this situation is simply a reflection of the internal project consistency (i.e. how well the software can shift/rotate/transform the data to fit the user-placed markers). It provides no true measure of accuracy, for which an independent set of check points is required. I would strongly suggest authors clarify this briefly, or change their terminology. I note that there has already been some discussion about the appropriateness of the term ‘accuracy’ elsewhere in the manuscript (see p24 of the author response), and request that the authors consider this a little further in this paragraph.

L261 – these sentences are a vast oversimplification of the M3C2 algorithm, which is complex and requires some more explanation – i.e. what was the radius for surface normal estimation, what was the value of the registration that was specified prior to analysis? A sentence explaining why you chose this method over others (e.g. 2.5D raster subtraction) would be useful – e.g. was the topography complex enough to require its use?

L371 – ‘thickness’, not ‘Thickness’

Table 1 – column 1 - please add some annotation to help the reader understand the source of each DEM – e.g. (TLS) / (UAV-SfM) etc

Figure 4 – requires a distance scale on both or either of the panels

Figure 6 – I take issue with the different scales on each of these panels – it is impossible to compare like-for-like. Strongly suggest authors modify this so that the point densities between panels are directly comparable.

Hide

ED: Reconsider after major revisions (further review by editor and referees) (07 Dec 2017) by Paolo Tarolli

AR by Davide Fugazza on behalf of the Authors (18 Jan 2018) Author's response Manuscript

ED: Referee Nomination & Report Request started (11 Feb 2018) by Paolo Tarolli

RR by Saskia Gindraux (14 Feb 2018)

ED: Publish subject to minor revisions (review by editor) (18 Feb 2018) by Paolo Tarolli

Dear author,
Your paper has been further revised by one reviewer, the one who raised major review, and by myself. The reviewer proposed at this stage minor, however highlighting the fact that the paper looks like a case study. I deeply read the paper again, and according to the goals of the Special Issue (where you submitted the work) on "The use of remotely piloted aircraft systems (RPAS) in monitoring applications and management of natural hazards", i think that this research, even if case study, is welcome. It is absolutely in line with the special issue purpose.
Summarizing: you provided a suitable reply to all the comments (major and minor) raised at the previous review stage. From my side, I think that only minor issues remain in work, mainly related to the section 5.2, and the figure quality and clearness. Here the details:
- section 5.2. Please improve (adding just 3-4 sentences at the end of the session), the discussion about hazards, and implications of this study. A reader should understand why it is so important to survey such environment and which kind of hazards can be mapped and monitored.
- "Meters" in the maps should be "m".
- Fig. 5. It should be improved in the clearness and font size of the numbers/letters. I suggest to reset the figure and set up it in a new layout (you can use any GIS), where the legend and scale should be one, background white, and number/letter font sizes larger.
- Fig. 6. I printed it; it is very dark, very difficult to read. I think you can play with the colour options and light as you wish in GIS. The scale should be one. Please avoid to insert letter in box, just simply add it as (a)(b)(c)
- Fig. 7. Here please add just one legend and one scale, with glaciers outlines in two colours.
- Fig. 8. This can be improved. Legend, scale and north arrow should be outside the figure. Otherwise, it is difficult to feel about the background. Make clearer the glaciers outlines: make the width size larger, try to use different colours (blue with blue on the ground is difficult to read)
Best regards
Paolo Tarolli

Hide

AR by Davide Fugazza on behalf of the Authors (06 Mar 2018) Author's response Manuscript

ED: Publish as is (10 Mar 2018) by Paolo Tarolli

AR by Davide Fugazza on behalf of the Authors (12 Mar 2018)

Short summary

This paper describes the surveys we performed in 2014 and 2016 by means of UAVs and terrestrial photogrammetry to monitor the Forni Glacier, one of the largest glaciers in the Italian Alps. We investigated the hazards related to the glacier collapse, which have been increasing recently due to the high ice melting rate. Our approach is feasible and low cost and we will repeatedly monitor the glacier to provide rapid hazard detection services to help the tourism sector.