Geometry-Based Preliminary Quantification of Landslide-Induced Impulse Wave Attenuation in Mountain Lakes

Round 1

Reviewer 1 Report

The paper provides a simplified methodology for assessing the magnitude of potential attenuation landslide-induced impulse waves in mountain lakes. Through the use of hydrodynamic modeling the metrics ‘wave decay potential’ was computed to evaluate wave dissipation in the lakes given its geometric properties that are used to estimate expected flow depth at the shoreline. The methodology can be used to assess lakes of different dimensions and characteristics to determine the vulnerability to landslide-induced tsunamis.

Through a description of the modeling procedure and application to natural environments authors are able to identify the strengths and weakness of the methodology. They do recognize this is the initial step in development of an assessment of landslide-induced waves and this should generate additional research. I wonder if additional work on the methodology could assess the different materials/lithologies that fail into the lake (rock and sediment). This would be of particular interest for studies of moraine-dammed lakes where sediment masses fail. Also, it would be interesting if the topography surrounding the lake shore could be modeled. With the advancement of drone technology and bathymetry mapping with UAVs it may be possible to reach inaccessible area and perform closely timed repeated surveys.

See the attached version of the manuscript for specific comments. The paper is well written, but there are a couple of places where the text could be more succinct (see lines 87-88 and 230-232). Also, the text in some of the figures is too small.

Comments for author File: Comments.pdf

Author Response

Dear Reviewer #1,

thanks a lot for taking care of our paper.

Please find in the attachments the file with the answers to your comments.

Hopefully, to have satisfied your expectations, we wish you all the best.

Kind regards

 

Andrea Franco and co-authors

Author Response File: Author Response.pdf

Reviewer 2 Report

Line 102- “The only variable required are…”

Lines 141-143: A Newtonian-like fluid of higher density than water is used to simulate the slide material.

-This does not seem to be a good model for the rock type of slide in the Chehalis-lake you use for validation. Why did you use this slide model as opposed to a solid slide model?

-In your numerical modeling are the slide and water layers coupled or uncoupled? Please specify. If uncoupled, some kind of justification would be appreciated.

-Figure 2.  For clarity,  for the right axis label, I would suggest not to use ˜Dw-Dw maybe ˜Dw or Dw, something that does not involve a mathematical operator. Also correct in Table 1 and elsewhere.

-Lines 167, Please define L and W, before you start using the L,W nomenclature.

-Figure 3: Please, explain what the xlabel on the plots of Figure 3 “Bin” mean.

-Lines 213. Please be consistent with the nomenclature for grid resolution. Does “(2m) mesh” mean (2x2x2)? And is the order (x,y,z)?

-Lines 215-217. The rational for the selection of the appropriate resolution does not seem very logical and seems to be driven by the inconvenience of long computations given the available hardware.

-The appropriate resolution level for the type of problem you are modeling comes imposed by two factors:

1-The complexity of the geometry which might require high grid resolution to accurately define the geometry of the problem. This is not your case, as your idealized geometries are very simple.

2-The wavelength of the smallest wave frequencies you expect to resolve in your calculations. This should be the driving factor in your analysis and not the size of the computational domain. Unless you are assuming that the highest frequency waves can only occur in small lakes, which is not the case, as this is determined by the size of the slide itself, which you have not changed throughout your simulations.

-Chances are you might be resolving all relevant wavelengths at the resolution you are computing in both small and large lakes, perhaps you might even be computing at higher resolution than necessary, and this would be good, but you should run a quick grid convergence test to ensure that your solution for at the lowest resolution is converged. You can do this by performing a simulation with a coarser grid than 10x10x5 and confirming that the solution only loses detail but does not change significantly by the introduction of numerical dispersion and/or dissipation.

Question: Why would a large lake require lower resolution than a small one? The grid resolution level in the idealized geometry that is being used for the numerical experiments (that is, the geometry of the problem being well-captured even at very low grid resolutions because of its simplicity), depends on the frequency components of the wave that need to resolved, but not on the size of the domain

-Figure 5. Please, make the small font labels larger. They are almost unreadable on an 8.5 x11 in or A4 paper size.

-Lines 298-302, seem to indicate that a correlation value of r=0.665 was obtained between lake length and mean flow depth at the shoreline.

Further in line 302 and Figure 6 a correlation between WDPP and length is stated.

Question: How can there be a correlation of length with any computed value, if your numerical experiments have been conducted with open boundary conditions away from the generation region (lines 221-223)? In other words, if your numerical experiments have an infinite length. Please, clarify.

-Figure 6. I find all text in that Figure to be almost unreadable due to small size on a standard size print out.

-Line 316. Please define what R squared, “Coefficient of Determination” is.

-Figure 7. Please, explain with more clarity what is the meaning of the correlation coefficients, “r=0.69” printed in red and in black “r=0.899”…What case do they represent?

-Line 330-331 . Subsequent analysis..consider …scenarios with a max depth greater than 20 m, but Figure 8 still seems to show results for depth=20 m. Also if the x axis in panel “b” is ShpP (m^2/3), shouldn’t the x axis in panel “a” be the log of that, please specify.

-Line 335: Why does the length go into the definition of Shop, when all your numerical simulations consider lakes of infinite length (open flow boundaries at both ends). Please, explain.

-Line 369; From Line 369 and line 378 it is unclear if the Chehalis Lake rockslide is 3 Mm3 or 2.2 Mm3. Please, clarify.

-Lines 415-418: “”There is nothing you can infer about the new methodology being more or less conservative  than the ETH one based on a single case comparison. You are as likely to be right as to be wrong, so you should not make such statements.

-My main concern with this study is the lack of real data to validate any of the conclusions. While the Chehalis Lake event is used in an attempt to do that, it seems that the validation is based on the comparison with the results that an analogous study has provided for that event, but now with real data collected by a field survey of the and post event analysis of the findings. This does not provide a good basis for validation. This is similar to validating a numerical model by comparing it to another numerical model. Just not the way it should be done.

I really encourage the authors to try to ground truth their formulas and equations with real data if at all possible. Alternatively, if the Chehalis Lake event could be modeled numerically with the same FLOW3D code they are using, they could use numerical simulations of a real event to validate their conclusions. Certainly, not ideal, but better than validating with another set of empirical, ad hoc, equations.

Author Response

Dear Reviewer #2,

thanks a lot for taking care of our paper.

Please find in the attachments the file with the answers to your comments.

Hopefully, to have satisfied your expectations, we wish you all the best.

Kind regards

 

Andrea Franco and co-authors

Author Response File: Author Response.pdf