Developing an Approximation of a Natural, Rough Gravel Riverbed Both Physically and Numerically

Round 1

Reviewer 1 Report

The authors are commented on addressing a very important topic for the physical and numerical modelling of wall bounded flows: the effect of bed surface roughness. They do so via the use of novel methods for reproducing both physically (eg via a CNC machine) and numerically (via digital techniques), the channel's bed surface roughness.

The manuscript is in general written and presented in accordance to the standards of this journal and I believe has sufficient novelty to interest the broader readership of the journal. I would suggest that this submission is approved for publication after a few minor edits aiming to improve its presentation.

Initially it is assumed that the goal of this work is to develop an approximation of a naturally rough gravel bed surface (eg see title). The characteristics of this surface are meant to be matching the natural case by design, thus any findings that there is a good degree of matching should put emphasis on the fact that the methods offered here allow for that to happen (eg that this is the outcome of this study, not merely what is designed to be the case). See for example lines 23-25 (and onwards similar cases, until the conclusions!): they read as if this is a statement (eg this is done so anyway by design), instead of like the major finding (e.g. need rephrase to something like "These observations mean that the methods developed in this study, can offer a physical approximation of a gravel bed surface, that is comparable to natural ones ...".).

The role of developing a (hydraulically) rough surface that is water-worked, is of fundamental importance, as the authors acknowledge themselves. The authors present some results around this but do not put sufficient emphasis to eg quantify and further discuss their results and implications. As the stated purpose of the methods developed here is to use them in CFD modelling, among others, it is important to know that the surface can be prepared to resemble a water-worked one or if not possible, offer suggestions to mitigate any effect on the results produced (could the authors expand a bit on their interesting discussions on this?).

Relevant to the above, can the authors provide any thoughts where their technique could be used (in combination with other sampling or photogrammetric methods) to provide a more water-worked riverbed surface, particularly for cases this may be deemed more important to model?

I am not convinced on the need of Figure 1 (as well as if the authors have permissions to reproduce it). It may be easier to remove.

Lines 52-56: could I suggest adding here the below reference relevant to coherent energetic flow structures and entrainment of bed material: Entrainment of coarse particles in turbulent flows: An energy approach M Valyrakis, P Diplas, CL Dancey - Journal of Geophysical Research: Earth Surface, 2013

Is there a reason why skewness and kurtosis are capitalised? 

Replace "natural gravels" with "natural gravel-bed surfaces" 

Stating numerical results (eg 3.806mm!! for standard deviation of gravel bed surface) should happen with a level of precision that is relevant to the practical applications. 

I would suggest remove the word "outlier" used to describe irregularities in bed surface elevations, as from a technical viewpoint these are not (statistical) outliers.

In the discussion and conclusions it would be desirable if the results were quantified (eg lines 409: by how much lower? comparing with their own data or from the literature).

Author Response

Please see uploaded document

Author Response File: Author Response.docx

Reviewer 2 Report

Dear Authors,

your work is very interesting and definitely worthy of publication, being very clear and well-written. The description of the experimental is made in detail, as well as the presentation and the discussion of the results.

Aside from the text, the figures need some work, mostly looking at improving the readability of the reported images. I suggested a few improvements:

What mean the two axes in Figure 1? From the text, I can understand that is the longitudinal and vertical velocities but could be better to report it directly in the figure.

In Figure 2, please add the letters (a,b,c,d) as reported in the caption. Same thing for the other figures.

In addition, please rearrange Table 1 to highlight what are the a) and b) columns.

Finally, is the Hurst coefficient equal to 2.94 (Figure 9) or 2.93 (lines 373 and 407)? Please, be consistent.

Author Response

Please see uploaded document

Author Response File: Author Response.docx