An Improved Method Constructing 3D River Channel for Flood Modeling
Round 1
Reviewer 1 Report
The manuscript is interesting and of good level. It can be of interest for the Computational Hydraulic community. I recommend that it is accepted for the publication on Water after a minor revision to be performed on the points highlighted in the following.
Introduction
The Author have to describe better the consequences of an erroneous description of the river topography. The Introduction is too concise.
River terrain construction in the planar coordinate system
The description of the reconstruction method of the river topography is not clear. Roughly speaking each point (interpolated or surveyed) should be described by means of its three spatial coordinates. Formula (1) e.g. gives the coordinate y of the jth interpolated point.What about x and z? Moreover, equations (2) are given without any explanation. The Authors have to re-write this part of the manuscript Accounting for these points.
Constructing error treatment
The observation done in point 3 holds for this point. Again only the x coordinate of the interpolated point is given (formula 4), nothing is said about the coordinates y,z and the text is too concise.
Benchmark introduction
I assume that formula 7 describes the thalweg as a curve in the plane x,y: i.e. it gives the y coordinate in terms of the x coordinate of the thalweg. Then formula (8) gives the z coordinate of each point belonging to a tranverse section on the thalweg. The adopted fomulas are not clear anyway, because each point belonging to a given section (interpolated or surveyed) is defined by its three spatial coordinates. From formulas 7,8 it is not clear as these coordinates are obtained.
Figure 9 does not give a clear graphical representation. It would be better to represent the sections on a plane rather than in the 3D space.
In the text and in table 4 the acronym of the Root Mean Square Error (RMSE) is reported as EMSE.
In conclusion: a minor revision is recommended before the manuscript is accepted. The revision has to be performed on the abovementioned points.
Author Response
Responses to Reviewers
Dear editor and reviewers:
We would like to thank the editor very much for giving us a chance to revise the paper, and also thank the reviewers so much for the constructive suggestions which help us significantly improve the quality of the paper. According to the comments and suggestions, we have made a thorough revision and resubmit the new version of the manuscript entitled “An Improved Method Constructing 3D River Channel for Flood Modelling” (ID: WATER_426279). In the revised manuscript, we highlighted all the changes in red. Should you have any query for the revision, the authors are pleased to make the further revision.
With best regards
The Authors
The manuscript is interesting and of good level. It can be of interest for the Computational Hydraulic community. I recommend that it is accepted for the publication on Water after a minor revision to be performed on the points highlighted in the following.
Introduction
The Author have to describe better the consequences of an erroneous description of the river topography. The Introduction is too concise.
Response: Thank you for the suggestion. The detailed description of the erroneous river topography is added in the Introduction (line 57) as:
In the actual situation, the river trend is complicated and it may be meandering. That means independent variable, x, is not monotonically increasing, and there may be a regression of x value (a fixed x may correspond to two y values) as shown in Figure 1 (a). In this case, it is prone to cause constructing errors by using CHS approach as plotted in Figure 1 (b) in section A
(a)
(b)
Figure 2. Meandering river channel (a) and erroneous river trend constructed by using the Cubic Hermite Spline (CHS) method (b).
And the Introduction is more detailed by describing the importance of topography for rain flood process numerical simulation. The revised text is as follows (line 33),
River channel terrain, a key spatial layer estimating channel networks in a catchment, slope gradients, flow direction and other controlling factors of the water movement, is one of the most important spatial input data sets in hydrological modelling (Bourdin et al., 2012). Without high-resolution topography, there will be a deviation in computed hydraulic factors, such as water depth and flow velocity, and in turn the rain-runoff process, thus affecting flood assessment and management.
Bourdin, D.R., Fleming, S.W., Stull, R.B., 2012. Stream flow modelling: A primer on applications, approaches and challenges. Atmosphere-Ocean 50, 507-536.
River terrain construction in the planar coordinate system
The description of the reconstruction method of the river topography is not clear. Roughly speaking each point (interpolated or surveyed) should be described by means of its three spatial coordinates. Formula (1) e.g. gives the coordinate y of the jth interpolated point. What about x and z? Moreover, equations (2) are given without any explanation. The Authors have to re-write this part of the manuscript accounting for these points.
Response: The formula 1 is used to determine the y valued of the considered point Pj whose x value of xj is linear interpolated between xi of section Si and xi+1 of section Si+1. The parameters applied in the formula 1 are computed by formula 2 and 3. When computing the z value of the point Pj, the formula 6 (Section 2.4) is utilized to computed the elevation zj.
This part of the manuscript has been rewritten in the end of the section 2 line 154 as:
After this step, both the planar and the vertical positions of the new point Pj are known. In a sum, Equation (1) is used to determine the y valued of the considered point Pj whose x value of xj is a given one linear interpolated between sections Si and Si+1. I. The parameters applied in Equation (1) are computed by Equations (2) and (3), while When Equation (6) is utilized to compute the z value of the point Pj. Other inserted points can be interpolated in the same way and terrain point clouds between the given sections can be generated.
Constructing error treatment
The observation done in point 3 holds for this point. Again only the x coordinate of the interpolated point is given (formula 4), nothing is said about the coordinates y, z and the text is too concise.
Response: To be honest, the problem is very similar to last comment. The make it more understandable, in addition to the revision in the end of Section 2, the second paragraph in Section 2.3 line 140 is revised as:
In Equation (4), the parameter a and b are obtained as that in Equation (2) and Accordingly, the parameter is modified as….
Benchmark introduction
I assume that formula 7 describes the thalweg as a curve in the plane x, y: i.e. it gives the y coordinate in terms of the x coordinate of the thalweg. Then formula (8) gives the z coordinate of each point belonging to a traverse section on the thalweg. The adopted formulas are not clear anyway, because each point belonging to a given section (interpolated or surveyed) is defined by its three spatial coordinates. From formulas 7,8 it is not clear as these coordinates are obtained.
Response: The formulas 7 and 8 are cited from Morales-Hernandez et al. [22]. For arbitrary x, its y coordinate can be computed by formula 7, while the z value can be obtained from formula 8. The formula 8 has two independent variables x and n, where the thalweg position of the considered cross section is determined by x; n means the variable changing along the cross section. Such explanation is added after formula 8.
Figure 9 does not give a clear graphical representation. It would be better to represent the sections on a plane rather than in the 3D space.
Response: Thank you very much. 2-D graphs can give a clear graphical representation, but it cannot visually show the section positions which also represent the accuracy of the constructing methods, while it can be achieved by 3-D graphs.
In the text and in table 4 the acronym of the Root Mean Square Error (RMSE) is reported as EMSE.
Response: Thank you for pointing it out. It is a mistake and has been changed into RMSE.
In conclusion: a minor revision is recommended before the manuscript is accepted. The revision has to be performed on the abovementioned points.
Response: Thanks for your constructive comments and valuable recommendations.
Author Response File: 
Author Response.docx
Reviewer 2 Report
Basically, I would like to see this paper published in water, but still I have some concerns on several issues in this paper.
This paper presents the methodology developed to construct 3D river beds information between cross-sections. The main objective of this method is to improve the thalerg interpolation between cross-sections by introducing the Quartic Hermite Spline Parameter.
General opinion
As the authors pointed out, the method to construct 3D river bet geo-information is important. While I feel that this method is seems to provide important contribution to a hydrodynamic model, the manuscript is weak in its presentation and in particular in the description of the erroneous interpolated thalweg by CHS and the parameter settings for Quartic Hermite Spline in the application. The authors pointed out the erroneous river trend constructed by CHS methods when the river section data is rare or the east-west direction value of the river sections used for interpolation is increasing. The author need to explain why the erroneous trend happen by CHS.
From this point, findings the main contribution of the manuscript would be introducing the parameter of Quartic Hermite Spline method in order to fit actual thalweg trajectory.
I recommend the authors focus on the parameter all the text.
I hope these comments will be helpful.
C1 Line 49: It is difficult to understand why the authors could get the result of Fig.6 (a) by use of a cubic Hermite spline method when the x value is decreasing. According to Caviedes-Voulli et al. [20], in their method to create thalweg trajectory interpolation by a cubic Hermite spline, the thalweg function must pass through thalweg points, and must be perpendicular to cross section Si and Si+1.
C2 Fig.6 (a) have no satisfaction that the thalweg trajectory interpolation is perpendicular to cross section. As they define the CHS to be written in parametric form of a piecewise interpolator, I think the direction of the flow has no relation to the interpolation.
If you use another method of CHS function, the authors have to describe why do you chose such method.
C3 Line 45: special > spatial?
C4 Line 102 “especially when the river section data is rate”
Is this a reason why unrealistic river trend happened by the CHS method.
C5 EQ. (2), What is ”h” stands for?
C6 EQ. (4) tantan > tan
C7 Line 169: How do you chose the value of adjusting parameter μ in the River thalweg construction? And do you use the same value to each case for all along the thalweg?
C8 Line 234: There is no description how to decide a QHS Parameter.
C9 Figure 14. Please add the description of DEM, what kind of product is this.
Please add the scale bar.
Can you show us ME and RMSE of verification sections?
C10 Table 3. Do you use DEM to calculate ME and RMSE?
Author Response
Responses to Reviewers
Dear editor and reviewers:
We would like to thank the editor very much for giving us a chance to revise the paper, and also thank the reviewers so much for the constructive suggestions which help us significantly improve the quality of the paper. According to the comments and suggestions, we have made a thorough revision and resubmit the new version of the manuscript entitled “An Improved Method Constructing 3D River Channel for Flood Modelling” (ID: WATER_426279). In the revised manuscript, we highlighted all the changes in red. Should you have any query for the revision, the authors are pleased to make the further revision.
With best regards
The Authors
Basically, I would like to see this paper published in water, but still I have some concerns on several issues in this paper.
This paper presents the methodology developed to construct 3D river beds information between cross-sections. The main objective of this method is to improve the thalweg interpolation between cross-sections by introducing the Quartic Hermite Spline Parameter.
General opinion
As the authors pointed out, the method to construct 3D river bet geo-information is important. While I feel that this method is seems to provide important contribution to a hydrodynamic model, the manuscript is weak in its presentation and in particular in the description of the erroneous interpolated thalweg by CHS and the parameter settings for Quartic Hermite Spline in the application. The authors pointed out the erroneous river trend constructed by CHS methods when the river section data is rare or the east-west direction value of the river sections used for interpolation is increasing. The author need to explain why the erroneous trend happen by CHS.
Response: Thank you for suggestion. The detailed description of the erroneous river topography is added in the Introduction (line 57) as:
In the actual situation, the river trend is complicated and it may be meandering. That means independent variable, x, is not monotonically increasing, and there may be a regression of x value (a fixed x may correspond to two y values) as shown in Figure 1 (a). In this case, it is prone to cause constructing errors by using CHS approach as plotted in Figure 1 (b) in section A
(a)
(b)
Figure 2. Meandering river channel (a) and erroneous river trend constructed by using the Cubic Hermite Spline (CHS) method (b).
From this point, findings the main contribution of the manuscript would be introducing the parameter of Quartic Hermite Spline method in order to fit actual thalweg trajectory. I recommend the authors focus on the parameter all the text. I hope these comments will be helpful.
C1 Line 49: It is difficult to understand why the authors could get the result of Fig.6 (a) by use of a Cubic Hermite spline method when the x value is decreasing. According to Caviedes-Voulli et al. [20], in their method to create thalweg trajectory interpolation by a cubic Hermite spline, the thalweg function must pass through thalweg points, and must be perpendicular to cross section Si and Si+1.
Response: As in last response, in the actual situation, the river trend is complicated and it may be meandering. That means independent variable, x, is not monotonically increasing, and there may be a regression of x value (a fixed x may correspond to two y values) as shown in Figure 1 (a). In this case, it is prone to cause constructing errors by using CHS approach as plotted in Figure 1 (b) in section A
(a)
(b)
Figure 2. Meandering river channel (a) and erroneous river trend constructed by using the Cubic Hermite Spline (CHS) method (b).
Yes, of course the thalweg function have to pass through the thalweg points in the control sections according to Caviedes-Voulli et al. [20].
C2 Fig.6 (a) have no satisfaction that the thalweg trajectory interpolation is perpendicular to cross section. As they define the CHS to be written in parametric form of a piecewise interpolator, I think the direction of the flow has no relation to the interpolation.
Response: Thank you for suggestion. The flow direction has no relationship with to the interpolation. The reason why erroneous trend happens by CHS are explained as follows:
In the actual situation, the river trend is complicated and it may be meandering. That means independent variable, x, is not monotonically increasing, and there may be a regression of x value (a fixed x may correspond to two y values) as shown in Figure 1 (a). In this case, it is prone to cause constructing errors by using CHS approach as plotted in Figure 1 (b) in section A
(a)
(b)
Figure 2. Meandering river channel (a) and erroneous river trend constructed by using the Cubic Hermite Spline (CHS) method (b).
If you use another method of CHS function, the authors have to describe why do you chose such method.
Response: Thanks for your suggestion. The author has consulted a large number of literatures and found that such CHS method presented in Caviedes-Voulli et al. [20] was widely used in some other research.
C3 Line 45: special > spatial?
Response: Thanks for your careful reading. Please excuse this clerical error. The “special” should be written as “spatial”.
C4 Line 102 “especially when the river section data is rare”
Is this a reason why unrealistic river trend happened by the CHS method?
Response: Thanks for your question. It is not the reason causing unrealistic river trend, the true reason is explained in the above-mentioned responses
C5 EQ. (2), What is “h” stands for?
Response: Thanks for your careful reading. Please excuse this clerical error. The “h” should be written as “δ”.
C6 EQ. (4) tantan > tan
Response: Thanks for your careful reading about this manuscript. This is also an error.
C7 Line 169: How do you chose the value of adjusting parameter μ in the River thalweg construction? And do you use the same value to each case for all along the thalweg?
Response: The value of μ is calculated by the trail method according to the measured value. When the calculated result is closest to the measured data, then the value of μ will be determined. The μ is different at different sections according to the above method.
C8 Line 234: There is no description how to decide a QHS Parameter.
Response: Thanks for your question. In the interpolation method of QHS presented in detail in Caviedes-Voulli et al. [20], there is no adjustable parameter μ.
C9 Figure 14. Please add the description of DEM, what kind of product is this.
Please add the scale bar.
Can you show us ME and RMSE of verification sections?
Response: Thanks for your kind consideration.
Firstly, a Digital Elevation Model (DEM) is a 3D representation of a terrain surface and created from elevation data of the terrain.
Secondly, the scale bar of this picture has been added in the Figure 14.
Figure 14. DEM and available cross sections for Wangmaogou catchment. Unit: m.
Thirdly, the value of ME and RMSE at different sections has been shown at Table 2 line 243.
C10 Table 3. Do you use DEM to calculate ME and RMSE?
Response: Yes, the ME and RMSE are calculated by interpolated DEM data and measured one.
Author Response File: Author Response.docx
Reviewer 3 Report
The reading gave me an impression that the paper was prepared in a hurry and is a bit messy as a result.
The Wangmaogou catchment is not characterised in the text so the reader has no idea what are the real channel dimensions and, as an effect, is the 15% better accuracy significant or it is within the margin of error.
Fig. 1 is repeated as part of Fig. 6. Is the Fig. 1 really necessary in introduction?
In Fig. 2 points marked along the cross sections lines are barely visible
Is RMSE (equation 10) presented as EMSE in Table 4? Check the rest of the text for the same.
Table 4 is mentioned as Table 1 in the text.
No scale in Fig. 14, 15 and 16. Please consider constructing regular key for those figures for readers comfort.
It is not clear in Fig. 15 what in fact is presented, I mean what are the coloured values representing. Why are the colours inverted (red for lower values and blue for higher).
In line 260, did you mean the 8 interpolated sections marked red in Fig. 14? If not, which and why did you choose? Number them in Fig. 14 for better understanding.
Lines 271 - 278 should be rephrased.
In my personal opinion the phrase "river terrain construction" repeated few times in the text should be replaced by "river terrain model construction"
Language of the paper requiers an improvement.
Author Response
Responses to Reviewers
Dear editor and reviewers:
We would like to thank the editor very much for giving us a chance to revise the paper, and also thank the reviewers so much for the constructive suggestions which help us significantly improve the quality of the paper. According to the comments and suggestions, we have made a thorough revision and resubmit the new version of the manuscript entitled “An Improved Method Constructing 3D River Channel for Flood Modelling” (ID: WATER_426279). In the revised manuscript, we highlighted all the changes in red. Should you have any query for the revision, the authors are pleased to make the further revision.
With best regards
The Authors
The reading gave me an impression that the paper was prepared in a hurry and is a bit messy as a result.
The Wangmaogou catchment is not characterized in the text so the reader has no idea what are the real channel dimensions and, as an effect, is the 15% better accuracy significant or it is within the margin of error.
Response: Thanks for your suggestion. The dimension of the catchment is about 5 km2. It has measured high-resolution terrain data and measure flow discharge, the good approach could construct more accurate terrain data and produce more reliable results, the authors believe the 15% better accuracy is able to show the performance of the proposed approach.
Fig. 1 is repeated as part of Fig. 6. Is the Fig. 1 really necessary in introduction?
Response: Thanks for your suggestion. The author assume the Fig. 1 is necessary in this part which can clearly show the problem by using the existing QHS method. A picture is some tiles much important comparing with a lot of words.
In Fig. 2 points marked along the cross sections lines are barely visible
Response: Thanks for your suggestion, we mark the points in a darker color in Figure 2 and Figure 4.
Figure 2. The planar view of a river channel with measured cross sections.
Figure 4. River channel construction using the QHSP method.
Is RMSE (equation 10) presented as EMSE in Table 4? Check the rest of the text for the same.
Response: Thank you for pointing it out. It is a mistake and has been changed into RMSE in line 225.
Table 4 is mentioned as Table 1 in the text.
Response: Thanks for your suggestion, I am so sorry that here is an error of serial number. The Table 4 in Line 225 should be Table 1.
No scale in Fig. 14, 15 and 16. Please consider constructing regular key for those figures for readers comfort.
Response: Thanks for your suggestion, the scale is added in Figure 14, 15 and 16.
Figure 14. DEM and available cross sections for Wangmaogou catchment. Unit: m.
(a) | (b) |
Figure 15. Constructed river channel at Wangmaogou catchment by using CHS (a) and QHSP (b) methods. Unit: m.
Figure 16. Flood simulation on the DEM constructed by using QHSP method at t=2.
It is not clear in Fig. 15 what in fact is presented, I mean what are the coloured values representing. Why are the colours inverted (red for lower values and blue for higher).
Response: We correct the scale bar as your suggestion (blue for lower and red for higher), thank you so much.
(a) | (b) |
Figure 15. Constructed river channel at Wangmaogou catchment by using CHS (a) and QHSP (b) methods. Unit: m.
In line 260, did you mean the 8 interpolated sections marked red in Fig. 14? If not, which and why did you choose? Number them in Fig. 14 for better understanding.
Response: Yes, the 8 sections marked red are the selected sections under consideration. According to your advice, the authors numbered all of them in Figure 14 for better understanding.
Figure 14. DEM and available cross sections for Wangmaogou catchment. Unit: m.
Lines 271 - 278 should be rephrased.
Response: Thanks for your suggestion, the paragraph in Line 290 - 297 has been rephrased by the author. The revised paragraph has been listed as follows:
Figure 17 plots that the simulated flood peak of the QHSP method is closer to that on the surveyed 3D bed than the CHS method. Besides, the computed hydrographs of the QHSP method are always in better agreement with those on the surveyed 3D bed. The phenomenon is also expressed in Table 5, demonstrating the computed errors by using the QHSP method are much lower than the CHS method at the considering sections. The accuracy of the QHSP method in terms of ME is increased by 18.5% and higher than 30% than the CHS method at section 6 and other sections, respectively. The similar result can also be indicated by the RMSE factor. The proposed QHSP method is therefore more suitable than the CHS method for the river bed construction.
In my personal opinion the phrase "river terrain construction" repeated few times in the text should be replaced by "river terrain model construction"
Response: Thanks for your suggestion, the authors has discussed that “river terrain model construction” is more suitable for the expression of this paper and replaced with it the paper.
Language of the paper requiers an improvement.
Response: Thanks for your suggestion, the language of this paper has improved by some experts.
Author Response File: Author Response.docx
Round 2
Reviewer 2 Report
Please refer to attached file.
Comments for author File: 
Comments.pdf
Author Response
Responses to Reviewers
Dear editor and reviewers:
We would like to thank the editor very much for giving us a chance to revise the paper, and also thank the reviewers so much for the constructive suggestions which help us significantly improve the quality of the paper. According to the comments and suggestions, we have made a thorough revision and resubmit the new version of the manuscript entitled “An Improved Method Constructing 3D River Channel for Flood Modelling” (ID: WATER_426279). In the revised manuscript, we highlighted all the changes in blue. Should you have any query for the revision, the authors are pleased to make the further revision.
With best regards
The Authors
New Comment
Comment 1：
I do not agree your opinion. I would ask authors to give us an explanation how you get the result in Fig. 1
When I used eq (9) in Caviedes-Voulli et al. [20], no trouble happened as you show in Fig.1.
What kind of CHS equation do you use for the result of Fig. 1.
If you use another equation, I think you cannot use this reference. Because readers would misunderstand the situation.
Response: The CHS equation used in this work is exactly same as Eq. (9) in Caviedes-Voulli et al. [20]. In order to describe this problem, a case study in Ningmeng Reach in Yellow River has been applied for these two different approaches. The problem shown in Figure 1(b) in the manuscript has been found in the following pictures. It is obvious the problem can only happen in the area where a same x value has two y values, see Fig 2 (blue lines denote the same x corresponds two y values).
Fig. 1 Erroneous river trend constructed by using the Cubic Hermite Spline (CHS) method in Ningmeng Reach.
Fig. 2 River construction by using Quartic Hermite Spline with Parameter approach.
In addition, all the source files and description are uploaded to the system.
Comment 2：
Please check Equation (2) b1 value in the manuscript. b1=－δ3+3δ2, when μ=0. Is this correct?
The b1would be δ3-δ2 when μ=0.
Response: Thank you for your kind reading. We are so sorry about the typing error on equation (2), the error may happen when presenting b1 from copying the equation of a1. It should be written as follows and the code is programed in the same way, we have revised it in the manuscript. Thanks for your correction again.
(1) |
Author Response File: Author Response.docx
