This study explored the sources of uncertainty in different components of the model chain and investigated their contributions to two climate indicators and three hydrological indicators. The variation in performance was evaluated based on different regional characteristics. I consider the motivation of this paper very good, especially in the context of using ensembles for climate projection. The structure of the paper is well organized, and the presentation is good as well. However, I have a few concerns about the calculation methods that need to be resolved before the paper can be accepted.

1. What is the purpose of applying cubic splines to the projection and what are the effects on the trend analysis (Line 583)? What is the meaning of the smooth trend denoted as CRi(t)? Please elaborate on the calculation method. Additionally, is the smoothing suitable for precipitation and hydrological indicators (especially max1D)? The authors may need to showcase some results from this step.
2. In the estimation of internal variability (Lines 593–600), why does the method first estimate Di(t) as the difference between the raw projection (Yi(t)) and CRi(t), rather than directly simulating variability from the raw projection over the target period? Does this step reduce or increase the internal variability? Based on the results, the internal variability is super large—could this be because the smoothing is not applicable?
3. Please elaborate on the calculation of ESi(t), using one example (e.g., RCP, s). Why is a linear regression model applied, and how is it used (Line 608)?
4. For Equation (A7), does this equation still work if incomplete or unbalanced ensembles are used? How are the effects of incomplete ensembles reflected in the results? Authors failed to explain this in detail since this is the second major question to be solved.
5. What is the difference between IV, RV, and FV? Should they use the same definition but with different superscripts/subscripts?
6. Does the selection of the time span length (i.e., 30 years in this study) affect the results, since a longer time span would likely lead to larger internal variability?

The manuscript “Uncertainty sources in a large ensemble of hydrological projections: Regional Climate Models and Internal Variability matter” by Evin et al. investigates the outcomes of several climate-to-hydrology modeling chains over the area of France. The perspective is to quantify the uncertainty introduced at any level of the chain by the various combinations of models that can be used at any level of the chain. This study is of great interest because it highlights the value and limitations of future hydrological projections, providing advice on which hydrological indicators can be predicted more robustly and where within the study area.

The work is a great effort of synthesis but leaves incomplete a fundamental point that I believe should be extended and integrated in the work. In the Discussion section (and recalled also in the Conclusions) the authors note that, in some cases, some models/combinations of models are discordant, then contributing significantly to the overall uncertainty. They suggest (page 26) to “look carefully at these deviating models and understand the reasons behind the atypical behavior”; then, if the model deviates for “wrong reasons (numerical artifacts, bugs, model inadequacy)” should be discarded, otherwise the analyst should investigate the reason of the dissimilarity further. In this case, the authors acknowledge the possibility that the model is superior to others. This point is extremely important, and the authors also acknowledge it at the end of page 26 and in the Conclusions section.

However, I would have expected a more thorough investigation of this point, i.e. a deeper analysis of “what to do” in these cases. Clearly, it is not possible to provide general instructions valid for any case, but the study provides a very good base to identify a few specific cases to develop in more detail. For instance, one could look at a specific catchment/climate regime and decide whether one model/model chain should be discarded or, on the contrary, should be preferred because more reliable in a specific context (see e.g. the cases reported in section 5.4.3). I understand that this is not an easy task, but it is necessary to make the paper much more than a descriptive product.

The manuscript titled Uncertainty sources in a large ensemble of hydrological projections: Regional Climate Models and Internal Variability matter by Evin et al. is well written and presents a comprehensive and carefully structured discussion, along with clearly articulated limitations and conclusions. I particularly appreciated the depth and clarity of the discussion section. The proposed methodology provides a robust framework to quantify uncertainties arising from multiple sources, including GCMs, RCMs, HMs, and internal variability, across different hydrological regimes, which I found both insightful and highly relevant.

The discussion provides a comprehensive interpretation of the results and highlights key findings regarding the relative contributions of individual models to overall uncertainty. In particular, the finding such as that one or a small number of models can contribute disproportionately to the total uncertainty is both interesting and well-supported. A deeper investigation into the underlying reasons for inter-model differences or the identification of hydrological conditions under which certain models may perform better could form the basis of a separate study. Nevertheless, by focusing solely on model output (data alone), the present work opens up a wide range of future research directions.

Overall, the manuscript is scientifically sound, clearly presented, and well-motivated, and I wholeheartedly recommend it for publication.

Cheers,

Prajwal Khanal