Dear authors,

Thank you for your responses in the interactive discussion. The major point at issue in some of the referee comments was the novelty and relevance of your work. Appreciating all of the comments together, I can overall see your manuscript fit into HESS in the sense that it provides an important contribution to applied research. At the same time, there is need to thoroughly address the further points raised by the reviewers, and thus to take the opportunity to strengthen the manuscript and emphasise its relevance for a broader readership. To this end, I suggest to also

1) emphasise why using GCMs can be considered appropriate in your case, and actually has advantages – see your short comment SC2.

2) discuss the generalisability of your study, not least also considering that you use GCMs and not RCMs. Moreover: What is particular in the Columbia River Basin, what can potentially be transferred to other regions without major effort? What lessons learnt can be useful for application to other regions?

3) include referee comments that go beyond the scope of your paper as indications of further work necessary.

Kind regards,
Daniel Viviroli

Dear authors,
Thank you for your revisions. As already noted earlier, your manuscript could make a valuable contribution to HESS. However, the main concerns of Referee #4 (i.e., points 1 and 2 in their initial referee comment, https://www.hydrol-earth-syst-sci-discuss.net/hess-2019-474/hess-2019-474-RC5-supplement.pdf) have not yet been sufficiently addressed. You have added a section on verification, which is an important step in this direction, but because your work leads to results different from many other studies, the referee emphasises the need for being more explicit in your comparisons, and helping the reader understand the uncertainties in the results. These points have to be addressed adequately before a final decision regarding publication in HESS can be made.
Kind regards,
Daniel Viviroli

Dear authors,

Your revised manuscript has now been seen again by Referee #4. The referee acknowledges that you have put considerable effort into the paper and the revision, but on the basis of the additional validations points to a substantial level of cascading uncertainties that might be inherent in the methodology used. In personal comments, the reviewer expressed that these uncertainties could question the significance of the changes in flood ratio you identify. The reviewer also expressed that your approach and model setup seem to do better with longer duration runoff and that it might therefore not be sufficient for extracting daily flows and analysing flood ratios up to 100 yrs return period, due to the calibration on 7-day means and uncertainties in downscaling.

These concerns are important because your results are contrary to prevailing literature, and even though you do make mention of limitations, there is strong need for clarification. In this sense, discussing potential weaknesses and their impact on your findings clearly would be a considerable added value to the paper. However, it might not be possible to accept the manuscript for publication in HESS unless that is done comprehensively and throughout the manuscript (including abstract) or, alternatively, if you can provide a clear rebuttal of the concerns, also to be reflected in the manuscript appropriately.

Kind regards,
Daniel Viviroli

Dear authors,
Thank you for the revised manuscript. The limitations and possible shortcomings of the study are now indicated more prominently and brought to the attention of the reader. This is not only important for the present study but – together with the comparisons added in the previous review round – can be useful as background for similar studies in other regions. After appraisal of the interactive online discussion, the reviewer comments made in the following revision rounds and your responses and revisions to the manuscript, I can accept the paper for publication in HESS subject to a few minor points:
- Figure 1: The numbers are hardly readable in some regions (e.g., Upper Willamette) and should be set apart better. Also, the signature for “streamflow location” (black dot) does not seem to appear in the figure itself. And for easier reference, the figure should reveal the letter-codes of the streamflow stations as well, not only the numbers (see e.g., Figure 6). Overall, please make sure the letter codes and numbers used are consistent across all figures and tables and are intelligible to readers not familiar with dataset and region.
- P1, L19–20, “more possible differences”: Differences in/to what? Do you mean a larger set of methodological variations?
- P6, L148–149, Please integrate “See Chegwidden et al (2019) for details.” into previous sentence.
- P14, L387–388, rather write “Third, our use of a large ensemble samples…”
- References: Some entries do not seem to be referred to from the text (e.g., Addor et al., 2014; Gutmann et al., 2014).
Kind regards,
Daniel Viviroli