General comments

The presented manuscript introduces a new approach determining the influence of granular flows on mast-like structures. It addresses an important question in the structural designing process: How high does an avalanche run up and what pressure acts on a structure? The concept of separating run up height from the hight of relevant pressure is described and well justified. Although simplifications were necessary, these are stated on the one hand and, on the other hand, the results demonstrate an improvement over existing approaches. Relating to this only Swiss guidelines are referenced, while a similar approach is also defined in a recently published Austrian standard (OENORM B 4801). Including this standard could further emphasize the importance of the topic.

Specific comments

Title: The authors could reconsider the title of the publication to clearly indicate that their results are based on numerical simulations.

Line 16: In this sentence the authors define snow avalanches as natural granular flows. In my opinion granular flows are conceptual models that provide an approximaton to nature. 

Line 110; 391: How was the value of 25kPa choosen? 

Line 118: What is meant by "...sufficiently large to minimize boundary effects and adjusted based on the compressibility"? It would be helpful to clarify the underlying assumptions.

Figure 3(a): It would be helpful to explain the pronounced drop of the upper envelope of FR=5 at t/tmax = 0.55. 

Technical comments

Some full stops are missing at the end of figure captions. 

Parts of the text in figure A1 ist hard to read ("2.5 dx"). Adding a halo or outline around the text could improve readability. 

Figure 8: Since lambda is mentioned before zeta in the text (Line 293), the panels a and b could be swapped for consistency. 

Line 340-341: Please check sentence structure

Figure A2/A4: The variable phi is presented in different styles to Figure 1. A uniform notation is recommended. 

I appreciate the opportunity to comment on this very interesting paper.

General

As a practicing engineer, I appreciate the author’s efforts to improve on methods to calculate pressure height distribution on narrow objects. I like the definition of pressure height. The rheology-flow regime parameters seem reasonable. 

Since the design of tall narrow structures subject to avalanche impact is often determined by the overturning moment rather than the total force, the heights of impact are critical for design. The simplifed point-mass methods overestimate impact pressure heights and can lead to overly-conservative designed, unless modified by engineering judgement. 

Specific

The graphical abstract (adapted from Margreth, et.al., 2015) shows a curious pressure distribution with the highest pressure at the base. This differs from Fig. 1 in the reference.

Constitutive model oversimplifies,  but is reasonable.

Results for the inertia-driven regime (Fr=5) represent common conditions for design. However, Froude numbers up to 10 occur for dry dense flowing avalanches. Why were higher Froude numbers not modeled?

Fig. 3 illustrates findings nicely and is consistent with full scale avalanches.

An example calculation and comparison with traditional method for the proposed empirical method might be helpful. One possibility is Example E.2 , Load on a 20m mast in the 2009 EC Design of Avalanche Protection Dams.

There is no discussion of temperature or wet avalanches flows. This might be stated directly or describe limits of temperature and/or water content that the methods are limited to. Conversely, state the “ideal” conditions where confidence is highest.

On line 242, “the spirit of the Swiss guidelines” is referenced. What is meant by this? 

My experience and judgment are consistent with the finding that pressure height becomes “saturated” and empirical methods are off at high Fr. Numbers. Fig. 5 illustrates this nicely.

Perhaps the most interesting finding of the numerical approach is the w/H affect for narrow structures. Lines 332-333 are relevant for design purposes. I calculate a 23% underestimation of f-r(w/H) for dense flow at w/H=0.5 based on Fig. 10. This is very interesting and important for design purposes. I look forward to seeing this finding tested, verified or modified as new data becomes available.