The paper of Richardson et al. clearly presents attempts to constrain the timing and processed involved in pediment formation in South Africa.

The authors present different models of pediment formation based on literature then they show their sampling sites, and their chronology based on ¹⁰Be measurements and finally discussed their results.

I will have only some questions on the cosmogenic part.

• Top sample of depth profile is not at a different altitude than the other samples below (779 and 776 m); Is this correct? (table 1) Is the density of the top profile sample also set at 1.6 ?

• I was able to redo all calculations and agree with minimum ages and maximum denudation rates determined by the authors (see joined excel file).

• However, I disagree with their exposure ages determined with 0.3 m/My denudation rate. Based on ¹⁰Be only I am sure that ages up to 4.6My can be determined (table 3 last column). Using the integration time formula of Lal (1991) involving neutrons only I have an upper age of 1.04 My ( see excel All samples sheet

• Regarding the depth profile I suggest the authors to do some more investigations. From this depth profile, sample SA-LB-DP0 is an outlier compared to the other samples. Considering all samples I determined a “slope decrease” of 112 g/cm2 in disagreement with neutrons attenuation length ( ~160g/cm2); Removing the upper samples the “slope” is 155.6 g/cm2 in agreement with low denudation and neutrons attenuation.

• From this depth profile, minimum ages can be determined for each point as well as max denudation rates. This can show that the profile is almost at steady state state

• Then it will be nice if the authors could model their depth profile to determine denudation rate and exposure time. I reach a minimum age of 360ka with no denudation.

• Now because the deepest sample exhibits the higher maximum denudation rate (5.47 m/My based-on my calculation), this sample can be used to better estimate the exposure age of the entire profile. To do so, I have considered a denudation of 3.1 m/My for all samples (based on max denudation rate of sample SA-LB_DP30 ) and infinite time for all samples except the deepest one. An exposure time of 867ka was achieved for the deepest. This age can better reflect the true age of the profile.

• The authors suggest a denudation change in the past; This has already be evidenced with cosmo but with two nuclides ( ¹⁰Be and 16Al) ( Jolivet et al. 2021 https://dx.doi.org/10.1016/j.geomorph.2021.107747 ; Godard et al 2021. ⟨1002/esp.5190⟩. I think that using one nuclide this is more difficult. I tried to model this with a denudation change from 0.3 m/My to 3.1m/Ma; I can reach a n age of 3My but the 30cm deep sample from the depth profile is not well modeled.  Why did the authors not measure ²⁶Al? This can be a real nice input and help to see a denudation rate change. I will not ask for that measurements for this paper but if the authors still have some remaining fractions, I encourage them to test (eventually they can contact me if help is needed).

I suggest accepting this paper with minor revisions

R Braucher

Dear Editor,

I reviewed this manuscript a few years ago. Upon reading the recent version, I have noticed some positive improvements. Therefore, I can only reiterate what I concluded about the manuscript previously.

The manuscript is well written and scientifically interesting. I think that the presented in this study data generally well support the suggested conclusions and the presented sequence of events during the evolution of the studied geomorphological features (Fig. 14). This study brings important advances in the better understanding of the Cretaceous and Cenozoic landscape evolution of South Africa. It also takes a significant step forward in challenging the paradigm of the existence of old, singular, large-scale erosional surfaces in southern Africa.

I have annotated a PDF copy of the manuscript with some minor comments.

Finally, I would recommend the publication of this manuscript after only minor corrections.

Kind regards.

Alexandre Kounov