This is an interesting paper on the optimization of change detection and the analysis and discussion on the rockfall magnitude-frequency relations on a cliff site. This manuscript has already been improved after implementing the recommendations expressed by the reviewers. However, there are several key points that still need to be addressed mainly related to the structure and writing of the manuscript.
My main concern about this manuscript is the lack of a clear structure. It does not follow the classic I-M-R-D-C approach which makes the reading sometimes very confusing. In addition to this, the topic is challenging and not easy for an audience not highly expert on this field, which makes using a very clear and consistent writing strategy absolutely essential. Reviewer 2 expressed also his concerns on the structure several times. In its present form, the manuscript is missing some of the potential scientific impact, not because the quality of the research (which is excellent) but how the information is presented and organized.
Overall, this can be considered as a major revision regarding the appearance of the final manuscript, but moderate considering the amount of work to be done, which has more to do with rearranging the information that already is on the paper than providing new one. Please find the details in the attached file with my comments over the revised manuscript.
The main aspects that need improvement are the following:

1. Clarify the objectives at the end of the introduction, i.e. 1) methodology for change detection; 2) rockfall M-F relations. Remove what sounds like methods or results. Include references for both objectives. Currently, it is not balanced, most of the introduction deals with objective 2, while most of the manuscript deals with the methodology.

2. I recommend using the objectives to give a clear structure to the manuscript (one section for each objective).

3. Use subsections for the different steps in a systematic way. This has already been implemented, but not systematically. Figure 3 should include all the steps that are included. Try to give a name to each step. There are stages without a specfic title such as removing scans due to bad weather conditions.

4. Try to be consistent with the terms, e.g. you entitle Fig 3 as rockfall inventory compilation. To me this is a diagram of the optimized change detection methodology, which is the first objective. I would recommend to use whenever possible the same terms for clarity purposes.

5. Try also to be specific, e.g. the deviation filter should be named waveform deviation filter. Deviation is too general.

6. I strongly recommend following the canonical research structure. Please rearrange the information to follow it. What are the methods applied? What results you obtained by applying them? Then you can discuss these results in the light of previous research.
This will imply copying and pasting complete sentences or paragraphs in the suitable sections and subsections. Most probably, the authors would not be aware of the need to proceed in this way, since being the designers of the whole manuscript, everything must be completely clear for them. However, the classic structure is meant to provide a common ground and order in the scientific communication, research quality assessment and repeatibility of the methods. To achieve this, the authors must examine carefully the sentences and decide to which section they belong. I recommend using the objectives of the paper as sections both in methods and results.

7. In the methodology, for each stage start by: 1) firstly mentioning the objective or problem you are addressing ; 2) summarizing previous approaches, and 3) only later describing or specifying separately the modifications proposed. Currently, everything are mixed in the same paragrapgh and your specific contributions are not clearly mentioned. It must be absolutely clear what has been already done in the past, what you borrow and what you contribute (and the reasons why).

8. In connection with some of my previous remarks, please reduce the length of several paragraphs. Some of them are more than 20 lines long (page 8, 13, 18). Try to provide one main idea per paragraph.

9. In addition to this, regarding the figures (which in general are excellent, but numerous, 13), many of them are packed with information (several subplots, several types of data in one subplot, very long captions). This produces unrest in the reader which already has to deal with complex mathematic equations and long paragraphs. Most of the information is useful, but try to be critical whether to reduce part of the content or at least reduce the caption comments, which can be made in the main text. By improving Figure 3, the authors might produce an excellent summary of the methodology.

10. In my view, section 4 should not be considered an application but the analysis of the M-F relations, which is the second objective.

11. I think that the conclusions could be separated in 2-3 paragraphs and made more exciting by providing a more general view and perspective, rather than providing a mere summary of the manuscript. Future research needs could be included.

The authors have addressed all the suggestions made by this AE. The manuscript looks now much more clear, readable and structured. Only a couple of possible technical corrections are provided.

1. The achknowledgements section has been included after the figures. Typically this should be included as a section after conclusions. If this is journal´s policy, then not consider.

2. I would suggest using correlative numbers in equations and avoiding letters such as 18a, 18b..., even though these equations were related.

Dear Authors,
Thank you for the corrections you have made to the paper submitted to ESurf. I am delighted to inform you that your paper has now been accepted for final publication subject to the two minor changes suggested by the associated editor (as listed in this message). If you could move the acknowledgement section and the equation numbers as suggested this would be great.
All the best and thank you for your paper,
Best wishes,
Tom Coulthard