The present manuscript discusses causes of the bimodal landscape in the Bohemian massif using geomorphic indices, basin-averaged erosion rates, and landscape evolution model. The topography in the study area consists of elevated low-relief surface upstream and deeply incised valley downstream, which roughly correspond to areas of metamorphic & granitic rock units and sedimentary rock units, respectively. The authors argue that this contrasting topography resulted from the difference in rock erodibility rather than transient response of the massif to the changes in uplift rates. I think this could be a good case study that demonstrates the controls of bedrock erodibility on landscape evolution and helps to understand the evolution of the bohemian Massif. In particular, I value that the authors combined the DEM analysis, 10Be-based erosion rates, and landscape evolution model. Although the manuscript would be of interest to the readers of ESurf, there are critical issues in the current manuscript that require significant revision of analytical methods, manuscript organization, the validity of the conclusion, and the contextualization. I provide major comments followed by line-by-line comments below, which I hope to help improving the manuscript.

The first issue is the inconsistency between the goal of this study and the presentation of the results. Although the goal of this study is to test the hypothesis that the current landscape is largely controlled by rock type, the current manuscript does not include statements nor figures that show the difference in geomorphic indices or erosion rates between rock type. Also, the manuscript does not provide information regarding the position of the elevated plateau and incised valleys, thus it is very hard to evaluate the contrasts of geomorphic indices and erosion rates between them. Filtering the topographic data according to percentile values of elevation in each catchment may help to see the difference between the elevated plateau and incised valleys, but I wonder why the authors did not analyze them separately when calculating geomorphic indices, which must be more straightforward. Also, it is difficult to understand why the Aschach catchment is representative of the study area. In most of the studied catchment, metamorphic & granitic rocks occur in the upstream part while sedimentary rock occurs sownstream. This is clearly not the case in the Aschach catchment.

The second issue is about the organization of the manuscript. The current discussion section contains paragraphs that are clearly about “method” or “results”. For instance, the section 5.2 explains the relationship between topography and bedrock lithology in the Aschach catchment. However, there is little explanation about this catchment in results, and all the information appear in this section. The section 5.3 includes model setup and the results of the model outputs. The model setup has to be in the method section, and the model outputs should be in the result section.

The third issue is about the conclusion. Although the authors argue “The resulting stepped landscape requires neither spatial nor temporal changes in uplift rate but can form by erodibility contrasts under uniform uplift conditions”, the current topography of the Bohemian Massif seems to be undergoing a transient response to uplift that formed the Massif. As the authors replicated in their landscape evolution model, the current topography of the Massif cannot be formed without the initiation of this uplift. It was very confusing to me that the authors argue the spatial and temporal change in uplift rates were not required while they provided a lot of evidence of the transient landscape. If there are other studies that propose a recent increase in uplift rates droved the transient response of the studied catchments, I suggest that you introduce those studies first and present your hypothesis.

The fourth issue is the contextualization. As a person who is unfamiliar with the geologic history of this region, it was hard to understand the significance of this study in terms of the evolution of the massif. Although the authors provide some references on geologic background of the massif, they did not clearly point out a research gap. Also, since the influence of rock strength on landscape evolution is well known, I suggest the authors explain more about the implication of their findings to put the current study in a wider context. I agree that bedrock strength strongly controls topography of the massif; however, the current study may end up with a case study without proper contextualization.

Line by line comments:

L47-50: Is the vertical velocity field based on GNSS observation?

L58–59: Influence of rock property on fluvial landscape is not exactly the same as those of uplift because the mechanisms of rock property and uplift control the erosional landscape are clearly different.

L82–83, ‘Curiously, the drainage divide…’: It wasn’y clear what you intended to say with this sentence.

L93: Please explain what the basin inversion is (e.g. spatial and temporal pattern of crustal deformation).

L106–108: In some catchment, granitic rock constitutes only a smaller part of the catchments (catchments at the western and eastern part of the study area).If the content of quartz is not similar among the rock types, resulting erosion rates are siginificantly biased towards rates in areas of specific rock type. I suggest explaining if the quartz content is similar among rock types examined.

L110 & Figure1: Please clarify the location of the physiographic transitions. This is a crucial information in this study; however, I could not understand where it was.

Fig1(a): I suggest using a continuous color scale rather than discrete scale, which helps to differeciate the elevated plateau and incised valley.

L139: Clarify ‘pre-processing’.

L153-154, ‘The choice of…’: This sentence would be incorrect. What matters in the choice of reference concavity is the gap between the concavity determined for each river profiles and the reference concavity.

L169–172: I do not think using elevation slices is helpful because you can manually separate the elevated plateau and incised valley.

L180: You may want to move some parts in section 5.3 to here.

L201: How did you determine the spatial variation of precipitation?

L205, result: I suggest rewriting this section to clarify the difference in gemorphc indices and erosion rates between the elevated plateau and incised valley or the difference between rock type.

L260-262: Geomorphc indices of the catchment 10 do not seem to be ‘exceptionally high’.

L308: I do not see arguments on the influence of rock properties in result section.

L351–353: Did you find the difference in erosion rates using nested-catchment samples (Lainitz: P03, P04; Kleine Mühl: P17, P18)?

L355–363: This paragraph reads like a geologic setting of the study area, which may be more appropriate in the section 2.

L367–369: I could not understand this sentence. Does this sentence imply the northern catchment have already been adjusted to the contrast in rock stregnth?

L431–432: How did you determine K=1? Since K strongly controls the model outputs, it is important to justify your choice.

L435–436: Does the smoothed topography retain the bimodal landscape of the Bohemian Massif? I guess the shape of this smoothed surface strongly controls the final results. Showing something like DSM of the top of layer L2 maybe helpful.

L443–444: Again, why did you set Kd of L1 10 times as large as that of L2?

Figure8: I suggest presenting in a plan-view. Also, it maybe better to show the modeled area in a large-scale map such as in Figure1.

In the paper from Robl et al., the authors use a combination of topographic analysis, cosmogenic erosion rates, and landscape evolution modeling to explore some of the nuances of a tectonically quiescent landscape with significant contrasts in erodibility and find that this erodibility contrast is likely critical in explaining the morphology of the landscape and the extent to which catchment averaged erosion rates do (or do not) follow sensible relations with the morphology. Overall I liked this paper and think it will make a strong and interesting contribution to ESurf after some revisions. Most of my comments reflect that while I feel like I eventually followed the authors arguments, it took a few more logical jumps than maybe were needed - and may have been unduly biased by my own perspectives. Either way, I think the paper could benefit from a bit more hand holding in terms of walking readers through the reason certain things are done and exactly what is being argued for. Most of this is addressed in my more line by line (or section by section) comments that follow, but I have two general comment to start with (one an actual comment and another more of a question). I look forward to seeing a version of this published sometime in the future.

General Comments:

1. I think the paper could really benefit from a schematic outlining in simple form what is being argued for in the context of the lithologic contrast. In theory, the model outputs do this (Figure 8), and while these are useful to some extent in validating the authors proposed model, in practice it’s actually hard to see what is actually happening through the detail. A simple cartoon would probably help. Along with what, it would help to emphasize (and I think is a point that is easily missed, unless I’m making it up for myself) I that the “bedrock barrier” as you describe it will persist for sometime after the overlying soft material is removed. I.e., at present, really only catchment 19 has the “setup” that you’re invoking, but critical to the argument is that (presumably) many of these catchments had molasse “sitting” on them before, but it even though it has been stripped away, the morphological impact remains as the streams haven’t fully adjusted yet. It’s an important point that’s pretty obvious from past work on lithologic contrasts in layered stratigraphy, but this is a great field example of it. Explaining this a bit more might help walk readers through the idea more and a simple cartoon would, in my opinion at least, really work well as a touchstone to do so.
2. Assuming I haven’t read way too much into what your data is showing and what you’re interpreting from it, do you think there’s more that could be done in terms of thinking about time? I.e., would parsing out which catchments (1) still have molasse, (2) don’t have molasse but have preservation of the “bedrock barrier” as a morphological fingerprint of the molasse having been there but now stripped away, or (3) have effectively fully adjusted – and the spatial relationships of these catchments to each other – tell you anything about the overall response time of the system and the potential timescale over which the bedrock barrier can persist. In other words, how long after the stripping of the molasse will its morphological presence be felt? This is a question that you can answer from landscape evolution models, but the timescale will depend on the (appropriate) values of erodibility, diffusivity, scaling exponents, etc., where as you here have an interesting natural example to maybe tease it out. Maybe not possible, maybe too much for this paper, but seems like something fun to consider.

Line-by-Line Comments:

L157-160: I doubt the results would be substantially different, but is there a specific reason you’re using geophysical relief as opposed to the more common local relief (e.g., DiBiase et al., 2010)?

L163-170: I think you need to lay out a bit more rationale for why the quartile approach is useful or appropriate here. I know that you do briefly in the sentences that follow, but given where the paper ends up, I imagine that many would expect that a more logical way to parse the landscape analysis would be by lithology. Indeed, it seems like Reviewer 1 specifically suggests this as well. This may indeed be useful, but I suspect some part of why you’re not doing this and are opting for the quartile approach instead is because your hypothesis and subsequent analysis suggests that there will be a good amount of morphological variability by lithologic unit (especially in the underlying hard unit) because of the formation of the “bedrock barrier”. I think this is an important point to make, both because it helps to explain why a morphologic analysis filtered by lithology might not be useful in this instance (and thus helps to lay out the underlying rational for your methodology a bit more clearly) but also a general takeaway for others doing these types of analyses.

L170-172: Probably a more nuanced point, but in terms of morphological responses to bedrock erodibility contrasts, these don’t necessarily have to propagate up from a stable base level, i.e., depending on the channel slope and the orientation of the contact between bedrock with contrasting erodibility, these contrasts may appear virtually anywhere along a profile, specifically in cases when a contact broadly dips in the same direction as channel slope. This is seen in a variety of simulations of fluvial erosion through contrasting erodibilities (e.g., Forte et al., 2016; Wolpert & Forte, 2021).

Section 4.1: There’s nothing wrong with the analysis here per se (and it’s certainly a very standard approach), but there’s a case to be made that given that many of the sampled basins are clearly in a transient state, simple regression of mean erosion rates vs mean topographic statistics would not really be expected to tell us much (and to the extent that they do show a pattern, there’s reason to be suspicious of it as there’s a  good chance of it being spurious to some degree or another). For example, from looking at something like figure 5c compared to 5f, I suspect that if many of these values were plotted with their appropriate uncertainties, that it would be more clear to readers that the degree of correlation should be viewed with more skepticism, i.e., some of the basins have a mean ksn that is effectively the same as the standard deviation on that value. I wonder if the depth of the analysis here ends up being more confusing than it needs to be in the sense of where it seems like you are taking the reader compared to where we eventually end up. Simply put, most of these basins are not the type that are well described with a single mean value because of the morphologic variability, so that there is a quasi linear trend in these is maybe not actually very relevant.

Figure 5: Per the above comment, it seems a bit disingenuous to show these plots without uncertainties. Even if these are not considered statistically in a meaningful way, at least visually readers could assess the extent to which a linear relationship is actually extractable. And similarly, a goodness of fit metric that includes the uncertainty (e.g., reduced chi squared, etc.) might be more relevant than a simple coefficient of determination.

Figure 7d: The yellow markers help, but it might be useful to also show where the contacts occur on the long profile (e.g., maybe colored bars marking which unit is exposed along the river).

L400-4010: While simpler than your models in the sense that they only consider detachment limited erosion and obviously not being calibrated to the specific details of your study area (e.g., must faster rock uplift rates), these topographic and erosion pattern are also broadly observed in “soft over hard” models in Forte et al. (2016). Specifically, the development of the low gradient “bench” of underlying hard rocks during and after the stripping of the overlying softer rocks but also the depression of the erosion rates upstream of the contact until the profiles have responded and steepened (e.g., their Figures 2-4).

Section 5.3.1: What values do you use for m and n? I don’t think I saw these specified either in this section or the relevant methods section on the landscape models? As I’m sure the authors are aware, in simple stream power models at least, the value of n dictates a fair bit of the type of behavior in layered stratigraphy (e.g., Perne et al., 2017) so it’s important to specify.

References cited in this review:

DiBiase, R. A., Whipple, K. X., Heimsath, A. M., & Ouimet, W. B. (2010). Landscape form and millennial erosion rates in the San Gabriel Mountains, CA. Earth and Planetary Science Letters, 289(1–2), 134–144.

Forte, A. M., Yanites, B. J., & Whipple, K. X. (2016). Complexities of landscape evolution during incision through layered stratigraphy with contrasts in rock strength. Earth Surface Processes and Landforms, 41, 1736–1757. https://doi.org/10.1002/esp.3947

Perne, M., Covington, M. D., Thaler, E. A., & Myre, J. M. (2017). Steady state, erosional continuity, and the topography of landscapes developed in layered rocks. Earth Surface Dynamics, 5(1), 85–100. https://doi.org/10.5194/esurf-5-85-2017

Wolpert, J. A., & Forte, A. M. (2021). Response of transient base level knickpoints to erosional efficiency contrasts in bedrock streams. Earth Surface Processes and Landforms, 46(10), 2092–2109. https://doi.org/10.1002/esp.5146