Quantifying sediment mass redistribution from joint time-lapse gravimetry and photogrammetry surveys

Mouyen, Maxime; Steer, Philippe; Chang, Kuo-Jen; Le Moigne, Nicolas; Hwang, Cheinway; Hsieh, Wen-Chi; Jeandet, Louise; Longuevergne, Laurent; Cheng, Ching-Chung; Boy, Jean-Paul; Masson, Frédéric

doi:https://doi.org/10.5194/esurf-8-555-2020

Articles | Volume 8, issue 2

https://doi.org/10.5194/esurf-8-555-2020

© Author(s) 2020. This work is distributed under
the Creative Commons Attribution 4.0 License.

https://doi.org/10.5194/esurf-8-555-2020

© Author(s) 2020. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 8, issue 2

Research article

|

22 Jun 2020

Research article |

| 22 Jun 2020

Quantifying sediment mass redistribution from joint time-lapse gravimetry and photogrammetry surveys

Maxime Mouyen, Philippe Steer, Kuo-Jen Chang, Nicolas Le Moigne, Cheinway Hwang, Wen-Chi Hsieh, Louise Jeandet, Laurent Longuevergne, Ching-Chung Cheng, Jean-Paul Boy, and Frédéric Masson

Download

Final revised paper (published on 22 Jun 2020)
Preprint (discussion started on 18 Jul 2019)

Interactive discussion

Status: closed

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

- Printer-friendly version

- Supplement

RC1: 'See comments in attached PDF.', Anonymous Referee #1, 07 Sep 2019
RC2: 'Gravimetry part and its interpretation', Anonymous Referee #2, 26 Sep 2019
AC1: 'Reply to reviewer and updated manuscript', Maxime Mouyen, 21 Nov 2019

Peer-review completion

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

AR by Maxime Mouyen on behalf of the Authors (02 Dec 2019) Author's response Manuscript

ED: Referee Nomination & Report Request started (18 Dec 2019) by Giulia Sofia

RR by Anonymous Referee #3 (09 Mar 2020)

Suggestions for revision or reasons for rejection

I have read this paper with some interest, as it presents a novel method to combine gravity and photogrammetry to constrain sediment mass redistribution in a very active fluvial environment. While interesting, I admit that I am not really convinced that this is currently a particularly useful method and that it’s worth doing the gravimetry work when just doing the photogrammetry will get you most of the way to the same answer, and doing both photogrammetry and the density measurements will get you all the way there without having to buy extra equipment. However, I recognize that this is a new application of gravimetry, and I do believe that explorations of new techniques are valuable and should be published.
That being said, I do have some issues with the manuscript in its present form. I will note that I am not an expert in gravimetry, so I am approaching this manuscript from a different perspective than the previous reviewers did. Some of my comments do repeat issues raised by the previous reviewers that still remain. This does make me wonder if some of the more technical gravity-related issues were also not sufficiently addressed, but again, I don’t have the expertise to evaluate this.

The framing in terms of erosion and sediment transport is not that convincing to me. You are comparing this sort of study to measures of suspended sediment transport, but these are very different things, so I don’t see how this is a useful contrast. Sediment fluxes in a river are an integrated signal of erosion upstream, while your method is measuring local changes, but not fluxes. I am not at all convinced that a denser network of gravimeters could measure sediment fluxes (see below), so your sentence in line 475-476 that contrasts this method with rating curves and sediment concentration data does not seem like a valid comparison. In general, I find the second half of the conclusions overly optimistic. You have a proof of concept for one type of gravimetry study. You propose a second type of application, but this should be treated as quite speculative, given that it’s completely untested. For the application that you do have the proof of concept for, it seems like it should be contrasted against existing methods to do similar local-scale change analyses: so photogrammetry only studies, terrestrial or airborne lidar, analysis of changes using satellite imagery or air photos, etc.

The photogrammetry is a major component of this study, but you are missing some critical information in your presentation of the methods and analysis. It may be good to look at James et al., 2019, ESPL (Guidelines on the use of structure-from-motion photogrammetry in geomorphic research) for an overview of what should be included in a paper that uses this technique. As it is, this part of the study could not be reproduced with the information given. How were the control and check points measured? What did they consist of (existing features, targets that you laid out, etc)? What was the error for these point locations? Where are the check points (and why are these not shown in figure 6)? How was the processing done? What exactly was done in Pix4Dmapper and what was done in ContextCapture? What parameters were used during processing? What is the uncertainty of the resulting model? How was it assessed? You give accuracy in the response to the reviews that seems extremely optimistic. Accuracy is related to ground sampling distance, and vertical accuracy will be a multiple of the GSD, so to have 3-5 cm accuracy for a survey with a GSD of 10-15 cm is unlikely. You say that shifting the prisms by +/- 5 cm has a small effect, but what about +/- 10-15 cm or more?

Section 6.2: seems like the correction for the river water should be in the methods, rather than in the discussion.

Section 6.4: I have major issues with this section. Your calculations show that you are unlikely to detect changes in suspended sediment concentration, and could detect changes in the alluvial thickness locally, but it does not follow that you get any information about sediment flux. In a river with an alluvial bed, steady bedload transport could occur with no local change in bedload thickness. If there are changes, you’d be measuring only divergences in bedload flux, not the flux itself.
However, even if you can’t get fluxes, the ability to continuously measure changes in the channel bed geometry during a large flood would be very interesting. Taiwanese rivers often have substantial erosion/deposition/channel shifts during large floods, but currently we can only measure before and after changes and not the dynamics during floods. These changes can be a real problem for discharge gauging (you can see this in the water level time series in fig. 12). This seems like a more feasible application of a network of in-stream gravimeters. But again, this is not bedload flux – I don’t see how the described method could measure fluxes.
I would also caution that installing anything in the bed of a Taiwanese river is asking for that thing to be destroyed, so that is a practicality to keep in mind.

Hide

ED: Reconsider after major revisions (10 Mar 2020) by Giulia Sofia

AR by Maxime Mouyen on behalf of the Authors (20 Apr 2020) Manuscript

ED: Referee Nomination & Report Request started (21 Apr 2020) by Giulia Sofia

RR by Anonymous Referee #3 (08 May 2020)

ED: Publish as is (08 May 2020) by Giulia Sofia

ED: Publish subject to technical corrections (18 May 2020) by A. Joshua West (Editor)

AR by Maxime Mouyen on behalf of the Authors (19 May 2020) Author's response Manuscript

Short summary

Land erosion creates sediment particles that are redistributed from mountains to oceans through climatic, tectonic and human activities, but measuring the mass of redistributed sediment is difficult. Here we describe a new method combining gravity and photogrammetry measurements, which make it possible to weigh the mass of sediment redistributed by a landslide and a river in Taiwan from 2015 to 2017. Trying this method in other regions will help us to better understand the erosion process.