The authors responded appropriately and thoroughly to the positive reviews, so I decided not to send the revised version back out to the referees. To ensure that a final set of eyes other than the authors passed over the article, I have now read it and reviewed it thoroughly. Almost all of my comments are minor -- sometimes to the point of copyediting -- aside from the need for a user's manual and a question about porosity in landslide deposits.

Overall, I find this to be a very exciting piece of work. I am glad to see an explicit treatment of mechanistically based sediment production in an LEM, and conservation of that sediment through the remainder of the system. This is a good add-on to the SPACE work by Shobe et al., and a major dose of realism beyond the stream-power-plus-hillslope-diffusion simple equations that typically are used to simulate landscape change over time.

Following the authors' treatment of these minor concerns, I believe that the article will be fit to publish as a strong contribution to landscape-evolution modeling.


Comments:

User manual: This could not be found at https://github.com/BCampforts/pub_hylands_campforts_etal_GMD or the (broken) links on that page, and must be completed prior to publication.

eq. 1: consider using \text{} for fluv, sed, hill

161-162: water discharge per unit width

180. Perhaps add a sentence as to why the failure plane bisects these two angles, to inform readers who do not understand why the failure should not just happen at the angle of internal friction. (I do not know this intuitively either, but guess that it may relate to the distribution of mass above the plan in the wedge.)

Eq. 9: perhaps give "rnd" a single-letter definition; multiple-letter variables can look like multiplication or functions.

231. I think that few beyond me care about this, but a "volumetric flux" should have units of length^3/time/length^2, with the cross-sectional area normalization making it a flux. Volumetric discharge is L^3/t. This is so routinely ignored that it probably doesn't matter to use the terms right, as much as it pains me.

230-231. Is this for the full hillslope area or just for the landsliding portion of the slopes? The discrepancy between the sub-sub-section title and the next sentence has me a bit confused.

235. as suspended sediment --> in high suspension. (And perhaps replace "instantaneously" with "rapidly". I think that your key point is that you want the sediment that still interacts with the bed. Because you use "sediment" here in both senses (all produced by the hillslope, that which you care about), I suggest that you simplify.

238. This section leaves me wondering about whether you take into account porosity. Landslide material -- especially from bedrock -- may increase in volume after it is released due to the addition of pore spaces. It seems like a reasonable idea to have this be the same value that you use for river-bed sediments, in order to not have to deal with varying porosities all across the landscape. Of course, adding porosity will create the problem that you have more volume of material to deposit than was eroded. Hopefully this can be implemented without too much trouble.

239. This paragraph could use a topic sentence.

250. which --> that

263. text, not code :) So $\infty$ not $inf$ to use the symbol

264-266. grammar/clarity

273. Perhaps note $V_\text{Lake}$ here (and/or refer to the table)

281. detachment --> detachment-limited

300. Since this is an issue of choices made within SPACE, and not something germane to HyLands, no need to address. But I would point out that sediment input equaling U * A implies that there is absolutely no clast breakdown! This is not internally consistent with your term to remove the fine fraction from landslide inputs. There are a few relevant articles that describe this issue:

* Attal et al. (2015): Impact of change in erosion rate and landscape steepness on hillslope and fluvial sediments grain size in the Feather River basin (Sierra Nevada, California)
* Sklar et al. (2017): The problem of predicting the size distribution of sediment supplied by hillslopes to rivers
* Wickert and Schildgen (2019): Long-profile evolution of transport-limited gravel-bed rivers

Importantly here, we (W&S) show that a linear Q_s,in = U * A relationship predicts channels with long profiles that are straight to convex.

There's nothing that you have to do here to address this point, but I thought it might be important to be aware of it, and the more general way of thinking of your "fine fraction removal" from landslides. Furthermore, the two hillslope papers (above) give some useful insights in to the grain-size distribution reaching the channel, which might be of interest to you.

365. Assuming bed sed porosity = 0 makes sense for this test (and may help you to address my comment on line 238), but neglecting an easy multiplier that adds realism doesn't make much sense to me. Still -- proof of concept -- you're okay in my book.

505. A more recent reference for river width and sediment supply is:
Pfeiffer et al. (2017), Sediment supply controls equilibrium channel geometry in gravel rivers. There are also good papers on alluvial rivers by Colin Phillips and Kieran Dunne, but these are probably less relevant to your high-relief and rapidly uplifting landscape. (The Phillips paper is: "Self-organization of river channels as a critical filter on climate signals".)

559-565. An excellent reference for this would be Tofelde et al. (2018): Effects of deep‐seated versus shallow hillslope processes on cosmogenic 10Be concentrations in fluvial sand and gravel. (Please feel no pressure to include it, though: I'm a coauthor, so would usually avoid noting the reference out of concern over conflict of interest, except that in this case it is so precisely geared towards your point.)

Excellent and appropriate revisions. The only tiny change that I might make is to move the Tofelde et al. reference to the next sentence (after the Wang et al. reference) because it is more focused on understanding the impacts of landslides on 10Be concentrations, and not just on the straightforward application of the method. Beyond this very small detail, I have no specific comments and think that this is an excellent piece of work.