I have taken a great pleasure to read the manuscript entitled “A geomorphic-process-based cellular automata model of colluvial wedge morphology and stratigraphy”. This manuscript investigates, using the GrainHill cellular automaton model (Tucket et al., 2018, 2020), the role of geomorphological processes in the development and geometrical structure of the sedimentary wedges that develops at the toe of normal fault scarps. The manuscript is overall well written, pleasant to read and, in my opinion, is original by efficiently linking geomorphological laws and processes to the facies and geometry of sediment wedges. It offers some outcomes that could have an impact in geomorphology (dynamics of the scarp-wedge system), but also in paleoseismology (interpretation of wedge structure and facies in terms of earthquake events) and sedimentology (by representing a neat study of the links between geomorphological processes and sedimentological facies). I congratulate the authors for that. In turn, the paper could be published as it is, as I have not noticed any methodological flaws or incorrect interpretation. However, I strongly believe the manuscript could be strengthened by better clarifying and structuring the results using physical dimensionless numbers (see my main comment below), which could also help to reduce the length of the paper (some parts are a bit long). I also give below some minor comments.

I also wish to mention that I am not a sedimentologist and that I was not able to properly assess the quality of the results in terms of sedimentological description.

Main comment:

The results of this manuscript, that are based on numerous models that were performed after a sensitivity analysis in the parameter space (e.g., disturbance rate, lateral collapse rate, weathering rate), could be described in a more structured (and clearer) manner by classifying some main types of models. This would really help to better understand the conditions required to produce different main types of wedges. In this case, the classification is not binary (e.g., transport- or detachment-limited) but considers three main driving processes which are defined by the weathering rate Wo, the disturbance rate D and the lateral collapse rate LCR. Three physical dimensionless numbers could be defined as Wo/D (i.e., the Peclet number already mentioned in the manuscript) to characterize the ability of the model to transport by disturbance (or diffusion) the produced regolith, Wo/LCR to characterize the ability of the model to transport by gravitational processes the produced regolith, and D/LCR to characterize the dominant mode of transport. Each model could then be classified using these three physical dimensionless numbers and discussed in the light of this classification. I strongly believe that this paper, in particular the Results and Discussion, would benefit from adopting this classified approach, which could really help to clarify the description of the models. In turn, wedges and scarps could be characterized by these numbers, in turn offering a clear physical description of the resulting wedge facies and structure.

I also believe that the main findings could be summer in a synthetic sketch using a kind of ternary-like diagram (with the axes Wo/D, Wo/LCR and D/LCR – even if the sum of these numbers is not necessary one) showing the main types of wedges and scarps.

Minor comments:

Line 38: “we desire this predictive power” – feels a bit strange in a scientific paper

Line 39: “do you preserve a post-earthquake colluvial wedge” – could be replaced by “is a post-earthquake colluvial wedge preserved” as the use of “you” seems unusual in a scientific paper.

Line 40: remove the capital letter to “3) How”

Line 167: It would be interesting here to mention the typical earthquake magnitude required to generate a 2m tall faut scarp using classical scaling laws between displacement and moment (or magnitude) [Leonard, M. (2010). Earthquake fault scaling: Self-consistent relating of rupture length, width, average displacement, and moment release. Bulletin of the Seismological Society of America, 100(5A), 1971-1988.]

Title section 2.2. I suggest rephrasing the title of this section for clarity "Facies definition and transport metrics based on cell tracking" or something else that suits the authors

Lines 216-220: The choices of the velocity thresholds between the different facies seem rather arbitrary. Could you please justify these values? I was also wondering why not using thresholds on the transport time, which intuitively appear as a more natural choice to divide the different facies and has the benefit of displaying some better resolved clusters (on Fig. 8) that would also likely be detected with classical clustering approaches (e.g., dbscan).

Lines 244: “both scarps both” – issue with the use of the word “both”

Figure 8: LCR is not defined in the caption.

Figure 11: This figure could go in supplementary as the fact that the volume of the wedge increase with Wo, D and LCR is relatively obvious, given the configuration of the model.

I would first like to thank the authors for this contribution, it was very interesting to read and opened my mind to new ideas. This manuscript attempts to decipher the geomorphic processes that control the evolution of a colluvial wedge following normal faulting and rapid surface deformation. The research of fault scarp evolution has been studied for many years, mostly as a tool to reconstruct paleoseismic activity. The authors use a relatively new physical process based cellular automata numerical model, which is an interesting alternative to the commonly used continuum-style models and by that avoid several limitations and over-simplistic assumptions. Similar models have been recently and successfully used to model various natural landforms, but not fault scarps. The results of this study give new perspective on the evolution of colluvial wedges but also on the much studied and poorly understood processes of soil covered hillslopes evolution.

The authors chose not to compare the model with real-world examples but rather run sensitivity analysis using very generalized settings. This limits the application of this study but gives profound basis for further research, which I hope to see in the future.

The manuscript is well written and original and could be published with minor changed. Below are suggestions that I think would benefit the manuscript, some might require moderate changes and I leave it to the editor and authors to decide if they are required.

Main comments

The resulted grouping, as shown in figure 8 is very interesting and deserves a dipper discussion. Several model runs show distinctly different grouping then the ones used (e.g. figure 8 C,G,K). It might hold valuable information on the evolution of colluvial wedge morphology and stratigraphy. It would be very interesting to connect the observed groups with the different facies, instead of using fixed cutoff values. Reading the manuscript, I was expecting to find a convincing, physically based, argument for the chosen threshold values of the different facies, but did not. I strongly suggest using a simple grouping method, or at least discuss the physical logic in using the same threshold values in mean velocity for different model scenarios.

The authors chose to use fixed morphological parameters to focus the analysis on geomorphical processes. Looking at the results, it seems that the comparison between 60 and 90 degrees dip shows high sensitivity and results that are sometimes trivial and related to bigger accommodating space and steeper initial dip. It might be better to use fixed dip angle and focus on the process-based parameters, which are less known (covering several orders of magnitudes each) and very challenging to decipher with any other method (unlike dip angle that can be measured directly in a trench).

According Wallace 1977, which his work is fundamental for this study, the debris-controlled phase ends when the fault scarp reaches the angle of repose. This was also the basis for initial conditions for several models that were used to morphologically date fault scarps. I believe that your results could shed light on the validity of this assumption if addressed more spesifically.

I suggest adding another parameter to the analysis - the time that has passed since a particle last moved. It could give valuable information on the timing of facies formation and give a more complete picture of the geomorphic evolution of the colluvial wedge. It would also make results more comparable with luminescence dating data.

Minor comments

Lines 166-167: How small is the is the initial sediment layer? This should be more methodically defined and described. It is not unlikely that a pre-faulted surface is covered with up to tens of centimeters of mobile regolith. I assume that the thickness of the initial sediment cover could influence results.

Lines 184-185: The classic definition of a peclet number, to my knowledge is different, and the referenced paper is about soil mixing, a process that is not described by the model.

Line 185: In figure 11 you show combinations of D and W0 values that range over 3-4 orders of magnitude. To my understanding of your definition of the peclet number (D/W0), this must result on much wider range of values.

Line 189: Why choosing the specific 4 orders of magnitude?

Lines 206-208: It is not clear to me where the threshold value of √3 radians comes from.

Line 331: Nelson 1992 is repeatedly and rightfully cited in this study, however it is worth addressing the fact that his pioneering work was limited to arid regions.

Line 351: I expect that the collapse dominanted stage will end when the surface angle will approach the angle of repose. It would be interesting to test it. It will validate a physical basis of the model and also contribute to studies of long-term modeling of fault scarp evolution that commonly assume rapid evolution until angle of repose is reached.

Technical corrections

Lines 103-104: Concider citing BenDror and Goren, 2018, JGR: Earth Surface.

Line 153: I believe it was Culling, 1963 who first used the term ‘diffusivity’ in soil transport.

Line 215: add reference to figure 8.

Line 340: Give references to these observations.