Dear authors of “Assessing the influence of water sampling strategy on the performance of tracer-aided hydrological modeling in a mountainous basin on the Tibetan Plateau”,

I had the pleasure to read your paper and in the following I provide some comments and suggestions aimed at improving your paper in a potential revision. I think that your paper has potential to advance the application and hydrological assessments using tracer-aided models in large-scale, mountainous catchments with an important glacier and snow contribution to streamflow. You clearly outline the challenges associated with working in such extreme environments and the difficulty to account for spatio-temporal variability of model inputs and outputs used for model evaluation. In this context, the use of isotope-enabled GCMs as input data for a hydrological model is a key result to me with potential for other applications elsewhere. Having said that, I have some doubts and suggestions you could consider incorporating into a revised paper:

• The model experiments are not clearly presented, which makes it in parts difficult to follow up on the results and conclusions.
• The model THREW-T, setup and data is poorly described raising questions about how the model treats isotope transport within model compartments. For example, I kept wondering throughout if and how the model treats glacier and snow isotope processes. To me, the glacier meltwater contribution (water and isotope) to streamflow is a simulated process within the model and part of the model evaluation. Your statements about a fixed glacier meltwater isotope signature to force the model (experiment 1) is not clear. Further, are the no model parameters associated to isotope transport and mixing? How did you arrive at the single (optimum?) benchmark parameter set for model comparison? The model calibration and how uncertainty was treated is also not clearly explained and no posterior parameter ranges after calibration are presented.
• The snow cover seems a crucial information for model calibration, but how was it derived and then simulated is not explained. Consider further sub-dividing the effect of snow cover on the model calibration compared to only streamflow and streamflow plus stream isotope calibration.
• The water source contributions to streamflow (Table 5) don’t seem to add up to 1? Also, many of the boxplots do not show a difference compared to the benchmark. Maybe consider a different visualization or quantification of differences and/or similarities might help to support the conclusions.
• The bias-correction of the iso-GCM is described in equations 1-3, but the results not shown. How does the result perform against an inferred or measured isotope-elevation gradient and is the lapse the dominant driver as opposed to spatial variability over such a large catchment area? The results of model experiment 2 (Figure 8) should be evaluated against streamflow isotopes and not streamflow.

The paper should also be thoroughly edited for language, as I detected many odd wordings and grammatical errors. I attached an annotated pdf with comments and suggestions for your information. My overall recommendation is that major revisions are required for potential publication in HESS.

Best wishes,

Christian Birkel

This paper used an isotope-aided hydrological model to assess the water sampling strategy in the mountainous region. Under the background of climate change affecting water resources, the isotope-aided model becomes a very useful method to see how does the water source contribution, especially the glacier-melt water contribution with the temperature change. The accuracy of such a method depends on the input of parameters, and thus the sampling strategy is highly important. The work tried several different sampling strategy and gave suggestions using the sensitivity analysis of existing hydrological model. So I agree the paper to be published, but there are also some points the authors should clarify.

My major comments are:

1. Data input. I did not see the description on the water isotopes original data. So what is the variability of precipitation isotopes? And how about snow-melt, glacier-melt and groundwater?
2. The conclusion: ‘Using a set of glacier meltwater δ18O that were 2‰~9‰ lower than the mean precipitation δ18O resulted in only small changes in the model performance and the quantifications of contributions of runoff components’ was inconsistent to the existing findings. Most of the previous studies attach great importance to glacier melt and snow-melt water isotope change, because they think this will lead to great bias of hydrograph separation (See the following references). Please discuss more on this.

Laudon, H., Slaymaker, O., 1997. Hydrograph separation using stable isotopes, silica and electrical conductivity: an alpine example. J. Hydrol. 201, 82–101.

Tao Pu, Ke Wang, Yanlong Kong, Xiaoyi Shi, Shichang Kang, Yonghui Huang, Yuanqing He, Shijin Wang, Jeonghoon Lee, Matthias Cuntz, 2020. Observing and modeling the isotopic evolution of snow meltwater on the southeastern Tibetan Plateau. Water Resources Research, 56, e2019WR026423. https://doi.org/ 10.1029/2019WR026423.

Altaf Lone, Ghulam Jeelani, R.D. Deshpande, Virendra Padhya, Estimating the sources of stream water in snow dominated catchments of western Himalayas, Advances in Water Resources, 10.1016/j.advwatres.2021.103995, 155, (103995), (2021).

3. The authors gave some suggestions on the sampling. For example, they concluded ‘It is highly recommended to increase the number of stream water sampling sites rather than spending resource on extensive sampling of stream water at a sole site for multiple years’. But I think this is highly up to the research purpose. If one wants to see the seasonal variation of water source contribution related to climate change, the conclusion should be inverse. So I suggest the authors to draw the conclusion more seriously, or add some preconditions.

Dear Authors,

This is a very interesting manuscript dealing with the tracer-aided hydrological models in the mountain basin. This topic contribute to developing knowledge in the catchment scale studies and lay down some basic points regarding the sampling strategy in the hydrological studies.

Overall, the manuscript is well organized, results are sound and most of them are well supported although there is a methodological issue (see the points below), and graphs well prepared. In my opinion, there are some moderate issues that requires a throughout revision.

I appreciate the description of the numerical experiment in Tab. 3. This is, certainly, a good way to summarize the faced setting.  However, I think that the authors need to improve the model description in the section 2.4. In particular, it is necessary to integrate the experiment 1 section because is not clear how to justify the benchmark parameter and then, the calibration is not clear explained. I think that many information can be deduced from Nan 2021 but it is better for the reader to have an integration. Similar considerations can be achieved for the correction of precipitation (c.ca Line 601, iGCM).

Fig. 1 I suggest to remove the connection lines between the three figures.

Then, I suggest to include the geographical info of Fig.1a and c inside the figure (similar to fig. 1b).

Line 297-298 maybe you miss the word “precipitation” after the first “isotope”.