Thanks for inviting me to review this paper, named ‘An integrated method for assessing vulnerability of buildings caused by debris flows in mountainous areas’. The topic is interesting, and the whole manuscript is well-written. However, I feel that there are still some questions needs to be further addressed before it can be accepted for publication in NHESS. Therefore, a minor revision is suggested. The detailed comments are listed below.

• Page 7, Eq. (1). It’s better to define V_pfirst before it appears in your manuscript.
• Page 8, line 115. Please be attention with your writing. It should be ‘As suggested by Jakob et al. (2012) and Kang and Kim (2016)’.
• Page 8, line 122. I think it’s better to add a reference to support your statement that ‘the moderate slope gradient and frictional resistance would decrease the kinetic energy of travelling mass’.
• Page 8, line 129, Eq. (2). Please clarify what hstands for? It is mean flow depth or the maximum flow depth? And also the debris-flow velocity, v.
• Page 8, line 134. Same question that you need to define Q_pand then use Q_p to represent a physical parameter.
• Page 9, line 145. I think the final equation of P_tis missing, which is represented by debris-flow volume and channel gradient.
• Page 12, line 220. Please explainthat how you estimate the impact pressure of 30 kPa to the damaged buildings? Additionally, how do the authors define that ‘buildings stay undamaged’? Do you mean there is no damage to the building wall and the structure ? Do you involve the furniture or household appliances and equipment into your analysis since they also belong to properties.
• Page 16, line 263. Is there is difference between the maximum flow depth (h) here and the flow depth in Eq. (2) ? Please be clear with the definitions in your manuscript.
• Page 16, Line 265. Please indicate that whether the height of building is involved into the elevation calculation.
• Page 18, line 290. Please indicate the ratio of a training set and validation set.
• Page 20, line 327. Please indicate the satellite image that you used to extract the buildings since readers need to know the resolution of your satellite images, which may cause uncertainty of physical vulnerability estimation.

The paper titled "An integrated method for assessing vulnerability of buildings caused by debris flows in mountainous areas" presents an interesting approach. The authors employ a machine learning model to estimate the potential impact pressure of debris flows on buildings. Additionally, they develop a physical vulnerability matrix that accounts for factors including building structure, spatial positioning, and impact pressure. The machine learning model produces reliable results with relatively low error. The study's estimates of economic loss offer valuable scientific insights for decision-makers and residents. I recommend a minor revision to further enhance the paper.

Specific comments:

• Page 5 L82-84 “The formation of terrain in this area …” Please provide relevant references.
• Page 6 Figure 1. This flowchart outlines the core of your methodology and presents a significant amount of information. However, the current figure lacks clarity. For instance, what do the different colors represent? Additionally, terms such as the GridSearch algorithm, n_estimators, and learning_rate, etc., are not clearly explained, either in the figure caption or elsewhere in the manuscript.
• Page 7 Figure 2. It would be clearer for readers if different colors were used to distinguish the debris flow events in each dataset.
• Page 8 Eq. 3. How do you determine channel gradient J as it varies along the channel.
• Page 9 L154. The elevation resolution for VD is also 0.8 m?
• Page 12 Eq. 13 and Page 19 L296. How do you obtain the actual value of impact pressure for the calculation of MAPE?
• Page 20 L314. Where is the data source of the economic loss?
• Page 21 Figure 9. It would be helpful to indicate the precise locations of sub-figures a, b, d, and e within sub-figure c. Why is the color representing buildings brown in sub-figure c, while it is yellow in the other sub-figures?

Technical corrections:

• Page 9 L150. Viscose debris flows should be viscous debris flows?
• Page 12 L213. y_ire should be y_ipre?

Thank you for inviting me to review the paper titled "An Integrated Method for Assessing the Vulnerability of Buildings to Debris Flows in Mountainous Areas." The paper introduces an intriguing approach, utilizing a machine learning model to predict the impact pressure of debris flows on buildings, while also creating a physical vulnerability matrix that incorporates factors such as building structure, spatial location, and impact pressure. The study's estimates of economic losses provide potential for valuable insights for both decision-makers and local communities -- suggesting a possible application for real-world impact. That being said, I have a few minor revisions to help enhance clarity of the paper:



Line 105: Please explicitly define damage degree. Is it the vulnerability matrix? A more explicit statement of that in the beginning of 2.1 would be clarifying. Line 155: Please elaborate on the fishnet tool and what it does briefly, beyond it’s purpose for your methods

Page 6, Figure 1: please explain variables and their names more explicitly, especially in the GridSearch algorithm section. Also, please indicate choice of colors and corresponding meanings

Figures 7- 12: Please offer a more descriptive caption of the figures