A Model Predictive Water-Level Difference Control Method for Automatic Control of Irrigation Canals
Round 1
Reviewer 1 Report
The authors integrate a method to control and improve free surface canals for irrigation. The topic is interesting, although not particularly novel. The MS has a moderate quality. The paper has many merits, but also some minor flaws that suggest its improvements. To summarise, postponing more details in the attached file:
(i) there is not a description of the study area
(ii) all the figures should be improved in their quality
(iii) there is a lack of comparison of relevant literature
(iv) some grammar mistakes are present and English deserves a careful control.
Comments for author File: 
Comments.pdf
Author Response
Dear viewer,
Thank you for your useful comments of our manuscript. We have corrected several mistakes in our previous draft and modified the manuscript accordingly. The detailed corrections are listed below point by point. Also, we mark the important changes in red in the revised manuscript.
Point 1: There is not a description of the study area
Response 1: The description of the study area is added to the manuscript in lines 85-91.
Point 2: All the figures should be improved in their quality
Response 2: We have improved the clarity of the figures and the figures 2-7 were enlarged. The abscissa time was shortened to 35 hours in order to better show the difference of different results.
Point 3: There is a lack of comparison of relevant literature
Response 3: We use the LQR water level difference control method in literature ( Clemmens A J. Water-Level Difference Controller for Main Canals) for comparison. There are two kinds of comparison, one is the comparison of the former difference control strategy with the difference control strategy in this paper, with the scenarios No.1 and No.2. The other is the comparison of LQR method together with difference control with MPC method together with difference control. We have explained this comparative relationship more clearly in the manuscript. Also, in order to get a fair comparison of MPC control and LQR control, a MPC control method with no future disturbance information used was added, and the Q in LQR-II method was changed to 30, to get a better water level difference control and a better comparison of LQR method with MPC control. Then the conclusions are improved with new results.
Point 4: Some grammar mistakes are present and English deserves a careful control.
Response 4: We have checked the grammars carefully and modified the grammatical and temporal errors
Author Response File: 
Author Response.docx
Reviewer 2 Report
The method used to solve/model the open channel equations must be stated. All that is required is that the solving method is stated and a reference is given on how this method works to solve the open channel equations.
Allowing the MPC system to use future disturbances in its routine is not a fair comparison to the LQR method that does not have this function. In order for this to be a fair comparison future disturbances must be used in all methods.
This is a very interesting paper. Please review the comments that I have made on the transcript. I encourage the authors to make the corrections that I have stated and submit the paper.
Comments for author File: Comments.pdf
Author Response
Dear viewer,
Thank you for your useful comments of our manuscript. We have corrected several mistakes in our previous draft and modified the manuscript accordingly. The detailed corrections are listed below point by point. Also, we mark the important changes in red in the revised manuscript.
Point 1:The method used to solve/model the open channel equations must be stated
Response 1: The simulation model of open channel and the method to solve it were added to the manuscript in lines 98-103. And some references were cited on the method to solve the equations.
Point 2: Allowing the MPC system to use future disturbances in its routine is not a fair comparison to the LQR method that does not have this function. In order for this to be a fair comparison future disturbances must be used in all methods.
Response 2: In order to get a fair comparison, we decided to add a MPC method with no future disturbance information used. Because to our knowledge, the LQR method does not have this function to use future disturbances in its routine, in some literature a feed forward method was used together with LQR method when future disturbance information is known in advance, but when to use feed forward and when to use the LQR method is still depends on experience. So, we added a kind of MPC method with no future disturbance information used. And the Q in LQR-II method was changed to 30, to get a better water level difference control and a better comparison of LQR method with MPC control. We have improved the clarity of the figures and the figures 2-7 were enlarged. The abscissa time was shortened to 35 hours in order to better show the difference of different results. Also, the conclusions have improved with the new results.
Author Response File: Author Response.docx
Round 2
Reviewer 1 Report
The authors have replied to my observations and addressed all the comments. I am satisfied about the work done.
Author Response
Dear viewer,
Thank you for your useful comments of our manuscript. After carefully check, we have modified the manuscript. The detailed corrections are listed below point by point. Also, we mark the important changes in blue in the revised manuscript.
Point 1:
A first sentence was add at the beginning of the Abstract explaining the technical problem that will be studied in the paper.
Point 2:
On line 89-90 a verb“is”was added
Point 3:
The explanation of friction slope term is clearer with another equation 2
Point 4:
The previous sub chapter 2.1 Test Canal and scenarios was divided into two sub chapters：2.1 Test Canal and scenarios and 2.2 Simulation model
Point 5:
The two terms "control horizon" and "prediction horizon" were defined at the beginning of the sub chapter 2.4. Model Predictive Control
Point 6:
In Table 3 add a third digit after the comma and the introduce in the chapter Results was also changed
Point 7:
In figure 2、4、6, the symbol ej was used in the ordinate axis, and DQj in figures 3、5 and 7. The letter a) to d) were used in the graphs
Point 8:
In the conclusion, a sentence at the beginning with a short description of paper topic and methodology before the discussion of the main results was added.
Reviewer 2 Report
I believe that the paper is ready for publication.
Author Response
Dear viewer,
Thank you for your useful comments of our manuscript. After carefully check, we have modified the manuscript. The detailed corrections are listed below point by point. Also, we mark the important changes in blue in the revised manuscript.
Point 1:
A first sentence was add at the beginning of the Abstract explaining the technical problem that will be studied in the paper.
Point 2:
On line 89-90 a verb“is”was added
Point 3:
The explanation of friction slope term is clearer with another equation 2
Point 4:
The previous sub chapter 2.1 Test Canal and scenarios was divided into two sub chapters：2.1 Test Canal and scenarios and 2.2 Simulation model
Point 5:
The two terms "control horizon" and "prediction horizon" were defined at the beginning of the sub chapter 2.4. Model Predictive Control
Point 6:
In Table 3 add a third digit after the comma and the introduce in the chapter Results was also changed
Point 7:
In figure 2、4、6, the symbol ej was used in the ordinate axis, and DQj in figures 3、5 and 7. The letter a) to d) were used in the graphs
Point 8:
In the conclusion, a sentence at the beginning with a short description of paper topic and methodology before the discussion of the main results was added.
