Next Article in Journal
Short-Term River Flow Forecasting Framework and Its Application in Cold Climatic Regions
Previous Article in Journal
Deformation and Stability Characteristics of Layered Rock Slope Affected by Rainfall Based on Anisotropy of Strength and Hydraulic Conductivity
Article
Peer-Review Record

Application of Multi-Source Data Fusion Method in Updating Topography and Estimating Sedimentation of the Reservoir

Water 2020, 12(11), 3057; https://doi.org/10.3390/w12113057
Reviewer 1: Anonymous
Reviewer 2: Anonymous
Reviewer 3: Anonymous
Water 2020, 12(11), 3057; https://doi.org/10.3390/w12113057
Received: 5 September 2020 / Revised: 21 October 2020 / Accepted: 26 October 2020 / Published: 30 October 2020

Round 1

Reviewer 1 Report

The article presents a fusion method usually applied for oceans - adapted with good results for an underwater terrain of a reservoir. There are several issues that need more explicit discussions.

  • What is the rate of change of the reservoir sedimentation? In respect with his one can decide whether fusing data collected in different time intervals leads to relevant results or not. The question is due to the information that sonar sounding data were collected in 2013 – 2015 (line 86), manual measurements were performed in 2014 – 2016 (line 90) and other data are from 2017 (line 92).
  • The information about the topography of the reservoir coming from different parts of the article needs more clarification: “… may experience a marked change in several decades” (line 240) and “may experience rapid changes after decades of operation” (line 346). This does not indicate the change rate actually. Is it known? Can it be predicted? Are the driving factors known and studied?
  • Although the article repeatedly tackles about accuracy, it is no evident how is the accuracy calculated for this method and what are the comparison results? The comparison with what reference? Which are the trusted measurements to be considered as reference?

 

Author Response

Dear Reviewer,

Thank you very much for your advice. We have substantially revised our manuscript after reading your comments. Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

The paper is well written. The topic is of interest for a broader public.

It gives a very practical application of data fusion form different sources.

From a methodological perspective, the paper has one weak point:

The main methods are not properly presented and explained.

 

  1. L114: Use different correction methods: please name them.
  2. L116: Use the remove-restore-method: there is not explanation nor reference. What does this method do?
  3. L118: The Kriging interpolation method. THE Kriging method does not exsist. Kriging interpolation methods is a broad family of of geostatistical interpolation methods, please specify. Kriging interpolation is based on a semi-variogram which need to be adapted (and interpreted in a scientific paper). CoKriging also uses a Co-Variable for interpolation which is also suitable for terrain interpolation. In addition you can also introduce anisotropy to account for structure like ridges etc... 

Figure 3:

"Generation low resolution grid by interpolation": What techniques is used. If you use Kriging, you also have to show the semi-variogram fit and the parameterisation. 

"Generating high resolution" by idensifiying: What idensifiying means...

"Generating the final DEM": How this is done, since you already have a high resolution grid in the step before.

You compare a methodology which uses Kriging interpolation. Kriging interpolation is very sensitive against its parameterisation. The semi-variogram function needs to be carefully choosen and adjusted. This includes anisotropy/direction. Kriging tends to produce errors if not properly parameterised.

Form you paper, the reader is not able to reproduce the methodology.

The semi variogram function show range (the max. distand in which correlation occurs) and sill (points below show positive correlation) and nugget (short range variability in the data). Therefore, from my perspective, it is needed to interpret your results.

Author Response

Dear Reviewer,

Thank you very much for your advice. We have substantially revised our manuscript after reading your comments. Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

The manuscript title: Application of Multi-Source Data Fusion Method in Updating Topography and Estimating Sedimentation of the Reservoir clearly explain the fusion methods to carry out a detailed topography of the bed of a river, in order to verify this method and in the future to carry out and improve the comparison of the evolution of this river. Which makes it a good tool for analysis, planning and management (risks, flods, etc). The quality of the images and the information is good and there presentation clear, for all these reasons i recommend the publication of this manuscript after minor changes (attached in the pdf)

Comments for author File: Comments.pdf

Author Response

Dear Reviewer,

Thank you very much for your advice. We have substantially revised our manuscript after reading your comments.

Round 2

Reviewer 1 Report

The authors introduced the required information in regard with the rate of change of the reservoire sedimentation and some new references about the topography changes. The answer about the accuracy is also satisfactory. 

Author Response

Dear Reviewer,

Thank you very much for your advice and comments. It's a great help for us to rethink and improve the paper.

 

Reviewer 2 Report

The paper has improved. But I still miss an adequate discussion of the Kriging method. It is not enough to simply use the ArcGIS geospatial analyst.

You write "The method was based on the Spherical semi-variogram model, and the model parameter, nugget, and partial sill are optimized using cross validation with a focus on the estimation of the range parameter by using the Geostatistical Analyst Tool of ArcGIS."

What's really interesting to the reader is how this semi-variogram looks like since the shape of this curve and (if possible to present) the underlying data points allow to understand the degree of spatial correlation: Kriging assumes that the distance (or direction) between sample points reflects a clear spatial correlation. That can be used to explain the variation in the surface.

In addition, if your structure implies a strong anisotropy (each valley has it) or you have other co-variables than topography, Ordinary Kriging will not work. In this case eg. IDW (Inverse Distance Weighted) is a better choice. Or you have to introduce Co-Kriging and/or Anisotropy.

For me, the selection of Ordinary Kriging as an interpolations algorithm is still not proven by your text. And in addition, the only advantage of this methodology is to directly display the spatial correlation in your data. And this is not done.

Ordinary Kriging, if not properly used, may produce artefacts in your interpolated surface, especially in the case spatial correlation does not exist or large anisotropy produce good results along the valley direction but bad results for the cross-section.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Back to TopTop