Automated Detection of Instability-Inducing Channel Geometry Transitions in Saint-Venant Simulation of Large-Scale River Networks

Round 1

Reviewer 1 Report

This study proposed an automated detection method to identify the modeling instability caused by the channel geometry transition. The study is interesting, and the method is novel, I provide the following comments for your reference.  

1. My first opinion is the research purpose should be more clearly addressed. If the objective of this submission is to identify the non-convergence of SVE routing due to the channel geometry unusual transition, this information seems can be very easily obtained through the direct conducting model simulation. If the purpose is to fix the instability of simulation, the fixed approximation seems to be omitted in this study. Identification of problem seems not an issue, how to fix that is the key point, and the paper does not give enough solution on that.
2. From above, if this study is going to provide the solution of automatically correcting the geometric section of the river channel, it may be significant. However, the adjustment of simulation is not just only relevant to the convergent of simulations but should be revised according to the hydrological observations.
3. In addition, there are some ambiguous comparisons from the results of this submission. It actually means nothing that the simulation costs are better than the physical modeling. Because each model has its applicability and functionality. The models are chosen based on their purpose, such as used for flood forecasting, engineering design, machine learning for future patterns..., etc., and to compare the superior among modeling is meaningless only if modeling purpose has been clearly defined. To define the purpose of research and the choice of appropriate tools is the meaning of simulation.
4. It is still vague on the theoretical part. The schematic diagram from Fig.3 fails to address how to deal with the confluence especially junction node, which will result in circumstances of flow slowing down. That should be explained in the paper.
5. This article has defined in the application of Large-scale river networks, it is supposed to consider the advantage of a large number of channel calculations. But for the set of governing equations applies from the large river networks to single channels is basically the same. Any scales of modeling, local-scale, river-basin, large-scale, will face the same instability problems. That argument makes your motivation of research weak. Please clarify your research objectives.

Author Response

Authors are grateful to the reviewer for taking time to review the manuscript. Our responses and the revised manuscript are in the attachment. 

Author Response File: Author Response.pdf

Reviewer 2 Report

The manuscript is well organized. The developments concerning the use of the De Saint Venant Equations are interesting.

In any case, some methodological aspects must be clarified on the basis of the utilization of the basin scales adopted.

I think it is necessary to better define the state of the art on this approach also in relation to the recent contributions by Costabile et al. (Hydraulic Characterization of River Networks Based on Flow Patterns Simulated by 2-D Shallow Water Modeling: Scaling Properties, Multifractal Interpretation, and Perspectives for Channel Heads Detection in Water Serources Research, 2019).

I suggest that the scaling processes must be integrated to better characterize the approach performed, above all to define whether the methodology used also reflects the channel scales as well as the basin ones.

Author Response

The authors appreciate the reviewer for taking time to review the manuscript and make suggestions. Our responses and the revised manuscript are attached herein for reviewer's reference.

Author Response File: Author Response.pdf