Operational Estimation of Landslide Runout: Comparison of Empirical and Numerical Methods
, Anne Mangeney 1, Gilles Grandjean 2, Clara Levy 2, Yannick Thiery 2, Jérémy Rohmer 2 and Antoine Lucas 1Round 1
Reviewer 1 Report
Please see attached.
Comments for author File: 
Comments.pdf
Author Response
Please see the attachment
Author Response File: Author Response.pdf
Reviewer 2 Report
This study compared methodologies for estimating runout distance of debris flows based on empirical approach, numerical simulations with back-analysis, and combination of both approaches. It discussed analysis procedure and results of each analysis in detail. It also showed interesting results and discussions.
If the authors add the results more cases of landslides and/or debris flows than three cases for further studies, it would be more reliable to understand the characteristics of runout distance. Moreover, it is recommeded to analyze the runout distance with non-parametric models to estimate runout distance under various conditions in the future.
I agree with this paper to be published in the current contents of the manuscript. However, I suggest to review the manuscript to upgrade English grammar and expression.
Author Response
We agree that more case studies would be needed to quantify more thoroughly the added value of using thin-layer models to estimate travel distances, in comparison to purely empirical methods. We add a comment about this point in the conclusion, l.555-556. However, we think such a systematic analysis is beyond the scope of our study, whose focus is on presenting the methodology of runout estimation with a first application on well-constrained case studies. Even though we use only three case studies, they span different contexts and volumes, which allows testing our methodology in different contexts. Testing more case studies would require the identification of landslides where the necessary data are available (in particular, Digital Elevation Models and the geometry of the initial mass) and/or acquire or infer these data, which is not the purpose of this paper.
We also agree that non-parametric models could help improve the results, as already mentioned in the perspectives of the conclusion (l.558-559). For a first application of our method, we chose power laws because their implementation is relatively easy, and because the interpretation of results (in terms of dependence to input data and of uncertainty propagation) is also more simple.
Finally, typos and grammar corrections were made throughout the article, along with minor modifications to enhance readability. They are highlighted in red and blue in the version with modifications highlighted. We list below the lines (in the article version where changes are highlighted) where such modifications are done:
l. 68-69, 84-86, 89, 124, 128, 131 135, 160, 165, 229, 287, 312, 363, 379, 555-556
