Reply to RC3

The authors applied this model on a profile perpendicular to the Himalayan wedge. Results make in evidence a good correlation between predicted ages and observed ages using an uniform model without effects of individual tectonic structures. Authors concludes that the principal implication of these results is a good reproducibility of the thermochronological data with a simple model and a possible steady state evolution of the Himalayan orogenic wedge.

The authors applied this model on a profile perpendicular to the Himalayan wedge. Results make in evidence a good correlation between predicted ages and observed ages using an uniform model without effects of individual tectonic structures. Authors concludes that the principal implication of these results is a good reproducibility of the thermochronological data with a simple model and a possible steady state evolution of the Himalayan orogenic wedge.
The manuscript is well written and we have all the details on the model and the analytical procedure, however major points have to be clarified and discussed, see my general comments.

General comments
1/ The main topic of the article have to be clarified l. 154-155 " Note that the goal here is not to provide a detailed geological case study, but to demonstrate the use of the equation to calculate deformation and exhumation rates" but authors highlight in abstract and conclusion "The results also imply that this part of the Himalayas may be in steadystate." (l.7-8) and "This indicates that the Himalayas may be in steady-state and that, at a large scale, the exhumation of mature mountain belts may be approximated by a relatively simple model of uniform and steady-state deformation, accretion and transport" (l. 239-241). This conclusion may be true but it is not relevant for the Himalaya Mountain range using one cross section.
Reply: This is a good point, and the reviewer is right that based on our study alone, we cannot make inferences for the entire Himalayas. Our results are however valid for the particular region. We have modified the statement accordingly. Batt et al. 2001 (JGR), a comparative study of the 2 models or at least a discussion on the main differences between the models must be developed. "However, to our knowledge no analytical solutions exist for the relation between deformation and exhumation of mountain belts." (l.23-24), this sentence is partially true and must be modified to show the specificity of this new model.

Reply: This oversight has also been raised by reviewers 1 & 2. A sentence on this previous model has been included in the introduction in the new version of the manuscript.
3/ Constant geothermal gradient is a big assumption in the models and it is not realistic according studies of Coutand et al., 2014 (JGR) and McQuarrie and Ehlers, 2015 (tectonics). This point and the impact on the model must be developed in the discussion. New kinematic models are not implemented (see Ault et al., 2019, tectonics for a synthesis) specify a closure temperature is possible and easier to implemented in the models but residence time in the Partial Annealing Zone (PAZ) and Partial Retention Zone (PRZ) need to be low to assume a closure temperature.
Reply: Agreed, we have updated the thermal and thermochronology model in the revised manuscript. See also our replies to comments by reviewer 1 and 2.
4/ Discussion of the results must be more detailed using differences between the predicted and observed ages and to discuss potential fault activity along the cross section. The comparison with study of McQuarrie and Ehlers, 2015 (tectonics) based on the better fit of the new model seems a bit complex. Differences between the 2 models are not specified in the text.
Reply: Agreed, we have expanded the discussion of the comparison between our results and the previous work by McQuarrie and Ehlers (2015).
L.167-169 Use references for the different "closure temperature" used and not "resetting temperature", numerous models showed that is a quite large range of temperature in particular if exhumation rates are low (see Ault et al. 2019, Tectonics).
Reply: We have revised the thermochronology model and have removed this part of the manuscript.
L.169-172 I agree but it can be an important bias on the model and you can have an idea calculating time residence of each sample in PAZ for AFT and PRZ for ZHe. It can be a sentence to justify that samples do not stay long time in PAZ and PRZ, in this case proxy on closure temperature can be used.

L.179 Define MAE.
Reply: There is a definition in the same sentence (mean absolute error).
L.186 It can be helpful for the reader to locate the different ages and samples on the cross section and having the possibility to link ages with the different tectonostratigraphic units. Scale of the Fig. 7a seems false, it is not 200 km long with this scale.

Reply: The modeled cross-section was larger than the geological cross-section to avoid boundary effects in the model. The sample locations have been added to the crosssecttion
L.198 More discussion about the data and differences between observed and predicted ages can be useful. I am very curious to see if AHe and AFT which not fit come from a particular tectonostratigraphic unit and If it is the case it can be associated to a major fault activity ?
Reply: We feel that this is out of scope of the manuscript. The model that we propose does not resolve individual faults. Alhtough it could be adjusted to do so, that would warrant a separate publication. Reply: The general consensus on assessment of model performance is that all else being equal models with less parameters are preferred over more complex models.
L.209-210: Digitizing from figure is not acceptable for publication. Ask the data.
Reply: After a repeat request we have now gotten the data.
L.223-229 As suggested previously, it can be relevant to discuss the differences between observed and predicted ages to discuss the reason of the differences geothermal gradient, fault activity ?
Reply: We feel that this is out of scope for this manuscript and would require a separate more detailed study.
L.239-240 This sentence is for a profile in the southern flank of Himalayas and cannot be in the conclusion.
Reply: Agreed, we have revised this sentence to state that this section in the Himalayas may be in steady-state.    Reply: The R2 is actually negative here. Note that the coefficient of determination (R2) can be negative if the variance of the model error exceeds the variance of the data, which is the case here.