Thank you very much for your valuable review. Please find the authors' response below. We refer to each of the reviewer's comment, which are shown in italics.

 Main:

Transport-limited vs supply-limited hillslope systems. In lines 101-103 the authors state that the debris flows in Horlachtal occur in transport-limited hillslope systems. However, in the discussion the authors argue that highly active periods affect debris-flow activity in the following years by reducing magnitude and frequency as a result of depleted sediment storages (e.g., lines 472-482). This is a textbook example of supply-limited conditions, and therefore the statements in lines 472-782 and 101-103 are in direct contrast with each other.

Comments from the authors:

In principle, the debris flows in the Horlachtal are transport-limited.

In the vast majority of the cases, the debris flow material originates on the one hand from glacial morain material covered with rockfall debris (talus slopes). On the other hand, it originates from rockfall material deposits temporarily stored in the catchments of the debris flows.

If there is a heavy precipitation event or several events within in a short period, the rockfall deposits in the catchments may be emptied. In addition, some debris flow channels are strongly incised into the talus slopes. Thus, those debris flows can no longer mobilise the morain material that easily. Only in these occasional cases we expect a short-term change from a transport-limited system to a supply-limited system.

We will elaborate this statement more clearly in the manuscript.

Relation between rainfall magnitude and flow magnitude. On a related note, in transport-limited systems one would expect a correlation between triggering rainfall magnitude and debris-flow magnitude. In contrast, such a correlation becomes weaker and would typically be absent in supply-limited systems because flow magnitudes also are limited by the volume of sediment available. It would therefore be of interest to compare triggering rainfall to flow magnitudes, or given the data availability perhaps maximum rainfall magnitudes in a given period versus the maximum debris-flow magnitudes in the same period (this should at least be possible for sub-catchment ZT judging from the information in section 5.2.2). It may be needed to normalize by catchment size or another morphometric characteristic of the source catchment as this also affects flow magnitude.

Comments from the authors:

We have already done these analyses, which showed no correlation between rainfall magnitude and volume.

However, the debris flow triggering thunderstorms are far too local to be able to make a well-founded statement here. We think that even for ZT, the precipitation measuring station is still too far away.

Catchment morphometry versus flow magnitudes. In section 4.3 the volume of 404 debris flows is compared to a number of morphometric parameters of their source catchments. A key component of such an analysis is information on how many events are generated from each studied source catchment. If each catchment in the dataset generates multiple flows there is stronger statistics, while if each catchment only produces 1 flow this introduces uncertainty since this one flow may have been relatively small or large. It would be good to elaborate on this in the manuscript.

Comments from the authors:

Thanks for the advice!

We can include this suggestion into our analyses. If we only take those catchments into account that produced at least two debris flows in the period studied, the sample size is reduced from 404 to 296.

The correlations now are a little bit weaker, but the statements from the analysis do not change as a result. We can add these data to Tab. 3. Now, the table not only shows correlation parameter for all debris flows but also are complemented by those catchments, which produced at least two flows (n > 1). You can see the adjusted Tab. 3 in the attached document.

Conclusions. I suggest to shorten the conclusions and also remove the subsections.

Comments from the authors:

Reviewer 2 also made a similar comment. We will adjust the conclusion section accordingly.

 Details:

Lines 97-98: Please elaborate on how these type 2 debris flows in Zimmermann (1990) or type 1 in Wichmann (2006) and Rieger (1999) are defined, e.g., describe their characteristics.

Lines 162-163: Two times “The approach”.

Line 2019: Parameter should be parameters.

Lines: 270-274: “The mapping of the debris flows showed a concentration of these processes in the parallel sub-catchments GT, LT and ZT. As those debris flows show such a different picture when comparing them to the activity in the other sub-catchments, and because of the similarities in the geomorphological and geographical settings, the analyses concerning deposition volumes were carried out exclusively in GT, LT and ZT.” It is unclear how the debris flows of sub-catchment GT, LT, and ZT differ from those of the other catchment. As such, this statement raises a lot of questions. Please clarify.

Line 300. Include space in “manyevents”.

Lines 297-301. To me the most striking feature in Fig. 8 is the strong increase in flow volume around 1990. Therefore, it would be good to also describe that here.

Line 532. Include space in “adetailed”.

Comments from the authors:

We will adjust the manuscript according to these comments.

Figures:

Overall: In many figures the font sizes are too small and should be enlarged.

Figure 2: It would be more informative to not only plot mean temperatures and precipitation, but rather plot a band indicating the values range. For example, mean +- std or 25-50-75 percentiles.

Comments from the authors:

We will try to implement this and expand Fig. 2 accordingly.

Figure 5: For readability the font size on the axes should be enlarged.

Figures 6 and 8: Given the unequal intervals of the time slices the left panels of these figures are not informative. I therefore suggest that the authors only present the data on the right panels, and combine debris flow frequency (Fig. 6) and magnitude (Fig. 8) in one figure. Do not denode panels as left and right, but annotate as “a” and “b”. In addition, also for these figures the font size is too small and I suggest enlarging the font. In addition, for the magnitudes it would be beneficial to also include uncertainties with dashed lines.

Comments from the authors:

We will also adjust these figures according to the comment.

Figure 12: For comparability it would be better to present the magnitude-frequency curves in panels b and c together in one panel. Also gridlines would help interpretation of the figures. Again, font sizes should be enlarged in this figure.

Comments from the authors:

The comments on the figures can be implemented as such.

Thank you very much for your valuable review. Please find the authors' response below. We refer to each of the reviewer's comment, which are shown in italics.

Major comments:

The scientific challenge of the manuscript need to be further sorted out. The author hopes to explore the spatial and temporal distribution characteristics of local slope-type debris flows, however, there seems to be no clear rule or conclusion until the end of the manuscript.

Comments from the authors:

We will try to be more precise about the objectives at the beginning of the manuscript.

The results indicate that the slope-type debris flow activities in the Horlachtal region show three active periods. However, they seem to be artificially divided. Under this premise, whether the statistical results of debris flows in different periods, especially the quantity, are in line with the actual situation. In reviewer’s opinion, people can get good statistical results they want by adjusting the time interval. Therefore, the basis of three active periods may need clarifications and solid reference.

Comments from the authors:

The used method influences the boundaries of the active periods. As a consequence, these boundaries are not chosen randomly, but are determined in advance by the availability of historical and recent aerial image surveys.

The only way to compare these periods with alternating durations is to normalise by the number of years between the periods.

This is already described in the manuscript in lines 395-397:The calculations of ‘debris flows per year’ suggest a uniformly distributed debris flows activity throughout the respective epochs, which is far from reality, and hence these calculations should be treated with caution”.

It is therefore due to the methodology that this approach is somewhat problematic in order to be able to delineate the “real” active periods with high accuracy.

The available precipitation data do not help a lot to further limit the active periods because the heavy rainfall events occur too local.

We will try to elaborate on this more clearly in the discussion.

In Abstract, authors points out that local thunderstorms are the triggering factors of debris flows. In this manuscript, only very limited words are used to describe this phenomenon. In reveiwer’s opinion, the existing materials cannot support this conclusion. Furthermore, this conclusion does not seem to be closely related to the subject of the manuscript, and it is not the main result of the study. Therefore, reviewer does not believe it is appropriate to mention in the Abstract as the main conclusion of the manuscript.

Comments from the authors:

The thunderstorms as debris flow triggering events were not the focus of our analyses. We will therefore cut the keyword from the abstract.

The manuscript mainly focuses on the spatiotemporal statistics of debris flows. However, the analysis of the causes of these laws and their physical mechanisms is relatively limited. The susceptibility of debris flow is affected by some important factors such as soil properties and vegetation conditions. In the analysis, the influence of the above factors should be further discussed in combination with the characteristics of the study area.

Comments from the authors:

Thanks for the advices!

In the vast majority of the cases, the slope-type debris flows in the study area are generated in the hydrological catchments (consisting of bedrock) respectively at the contact zone of the catchments with the talus slope.

In Horlachtal, no trees or higher vegetation grow in the catchments. There is also hardly any soil formation there due to the altitude and the high morphodynamics. If there is any soil formation, then only shallow initial soils.

Whether a debris flow is triggered or not is therefore less influenced by the factors mentioned above, but is (since transport-limited systems are predominant) mainly influenced by precipitation event intensity and the conditions in the catchment area of the debris flow, which lies exclusively in bedrock with no noteworthy vegetation or soils.

We can address this more clearly in the discussion.

The structure of the manuscript needs to be further streamlined and optimized. For example, “Methodological limitations” are suggested to be placed after the discussion, rather than before each discussion, which will hinder readers' understanding of the research conclusions. In addition, “Conclusions” in the current manuscript need to be modified. It is recommended to refer to other literatures published in NHESS for further simplification to show the insight, impact and implication of current study.

Comments from the authors:

We appreciate the suggestions for the structure of the manuscript and we will try to implement them.

Reviewer 1 had similar comments on the conclusion. We will revise the section accordingly.

Minor comments:

It is recommended to further modify the figures:

The font sizes in Figures 5, 6, 8, 9, 10, 11, and 12 are too small, It is recommended to adjust according to the journal requirements;

For debrs flow volume, uncertainties of the calculations are presented by error bars in Figure 8, so other volume related figures may alos need erro bars?

Comments from the authors:

We will try to implement the comments on the figures as noted.