Dear Authors,

thank you for submitting your manuscript to this special issue. Two reviewers have suggested improvements to your paper and you have replied proposing modifications that I think can improve your paper. I encourage you to submit a revised version following their suggestions, which will be further reviewed by myself, editor, and referees.

Best regards,
Maria-Helena Ramos.

Dear Authors,

Thank you for the revised version and the replies to the two reviewers. I think you have addressed their comments accordingly and, given their recommendations and my own revision of the revised version, I recommend your paper to be published subject to some technical corrections, which I am listing here below. I hope these final suggestions will enhance your final paper.

Technical corrections suggested for the final version:

- Page 3, lines 13, 33: concerning the citation to the unpublished work by Kidmose et al., please check if it has been already published. The authors should avoid sending the reader to this reference for further details on technical implementations if this is an unpublished work (where will the reader find it?). If it is still unpublished, considered deleting the sentences/citation or changing it to “personal communication”, but without sending the reader to it for further details. Same for Page 11, line 40.
- Page 4, line 5: change to “… 24 years. Finally, …”
- Page 4, line 23: Check if you mean Zhao et al (2011), as in the reference list.
- Page 5, lines 23-24: consider changing to “…on average, the system being evaluated does not perform better than the ESP system used as reference.”
- Page 6, lines 1-4: I did not understand if the ESP system, which has 23 members, was also bootstrapped from random selections of 15 members. Please clarify this point here, since it can play a role in the skill scores (CRPSS and SS).
- Page 6, line 18: do you mean, “The test for uniformity is not rejected if…”? Please clarify when you refer to the results of the hypothesis test.
- Page 6, lines 32-33: I do not understand the last sentence. What do you mean by “more rigorous” and why would it be “more rigorous”? Please, consider explaining it or deleting this sentence.
- Page 7, line 16: consider changing to “…value of bias for a forecast initiated on October 1 for the target month of November.”
- Page 7, lines 21 and 22: I suggest you to mention examples of months with “higher bias” and “milder bias” in parentheses to help the reader in the interpretation of the results.
- Page 7, line 27: consider changing to “… in terms of sharpness (SS)”
- Page 7, line 29: consider changing to “In our case study,…”. It is not clear to me to which “this case” is referring to.
- Page 7, lines 31-32: consider changing to “…due to accuracy (CRPSS) indicate that… of skill as expected (not shown).”
- Page 7, line 38-39: consider specifying the months and changing to: “… for winter (December, January and February) and summer (June, July, August)… at lead time 1 month.”
- Page 8, line 4: consider changing to “…the uniformity test at lead time 1 month (not shown)”
- Page 8, line 5: consider changing to “…October (not shown) show a significant…”
- Page 8, line 5: consider changing “appears to get worse” to “worsens as lead times increases” or “becomes worse”
- Page 8, line 23: consider changing to “… than the reference ESP forecasts”. I also have a suggestion here: How about adding a line on the comparison with raw forecasts?
- Page 8, lines 25-26: consider changing to “is worse than”
- Page 8, line 29: consider changing to “For spring (not shown) and …”
- Page 8, lines 33-31: consider changing to “… is clearer, with most of the months passing the uniformity test (not shown).”
- Page 9, line 5: consider changing to “…for the months of the summer and winter seasons…”
- Page 9, line 11: consider changing to “…long lead times, when the larger…”
- Page 9, line 19-20: consider changing to “…against ESP based on observations, we also include a comparison against ESP based on simulated streamflows…”. Is that what you mean? I did not fully understand in the original sentence.
- Page 10, line 2: consider changing to “… model errors are smaller…”
- Page 10, line 3: consider changing to “…is a product of…”
- Page 10, line 4: consider changing to “…errors interactions, rather than…”
- Section 3.6: I think the use of “low flow forecasting” is a bit misleading. I suggest the use of “Forecasts of the minimum daily flow within a year”. I think that the fact that it refers to the verification of a daily variable (and not of monthly streamflows as previously) should be highlighted. This clarification should be done all over the paper, and specially in this section. For instance, in line 22, “Figure 11 shows forecasts of the minimum daily flow in each year of the study period, considering a forecast issued on April 1 for the next 7 months.” Please, consider clarifying this over the paper.
- Page 10, line 23: consider changing to “… outlet station 82…”
- Page 10, lines 27-29: the use of low observed low flows, highest minimum discharge and high observed low flow is very confusing. Consider rephrasing the sentence.
- Page 10, line 31: consider changing to “…to perform better than ESP…” Same in lines 34 and 41.
- Page 10, line 33: consider changing to “…, which allows the forecasts…”
- Page 10, line 35: consider changing to “…similar effect: loss of sharpness and decrease in …”
- Page 11, line 2: consider changing to “…month lead time for the Ahlergaarde catchment in Denmark.”
- Page 11, lines 2-6: consider changing to “…the limited skill of … region (Lucatero et al., 2017). Similar conclusions have been documented…and, more recently, …for France.”
- Page 11, line 8: considering clarifying that you refer to “the minimum daily flow in each year”
- Page 11, line 13: consider changing to “… winter in the study region.”
- Page 11, line 14: consider changing to “… calibration objective functions focus on attributes that differ from the ones looked for in the final forecast quantity of interest.”
- Page 11, line 16: consider changing to “In our study, preprocessing…”
- Page 11, line 17: consider changing to “…or overestimation lead to…”
- Page 11, line 19: consider changing to “…which, in turn, helps…”
- Page 11, line 23: consider changing to “…and the ESP forecasts. This is also true for forecasts of minimum daily flows in a year, as mentioned above.”
- Page 11, line 26: consider changing to “…, in general,…decade,…”
- Page 11, line 27: consider changing to Li et al. (2017)
- Page 11, line 28: consider changing to Zhao et al. (2017). The reference is missing in the reference list. I guess you refer to: Zhao, T., Bennett, J., Wang, Q.J., Schepen, A., Wood, A., Robertson D., Ramos, M.-H., 2017: How suitable is quantile mapping for postprocessing GCM precipitation forecasts? Journal of Climate, 30, 3185-3196. DOI: 10.1175/JCLI-D-16-0652.1
- Page 11, line 29: consider changing to “The main issue discussed concern the fact that when the forecast-observation…is weak, or non-existent, …. (i.e., that have skill at least…”
- Page 11, line 32: consider changing to “… due to the limited…”
- Page 11, line 34: consider changing to “… Mendoza et al. (2017), although…”
- Page 12, line 16: consider changing to “… large skill can be an effect…”
- Page 12, line 19: consider changing to “Finally, we also estimated the …”
- Page 12, line 20: consider changing to “… minimum yearly daily discharge.” Check also for the use of “low flow forecasts” in line 21.
- Page 12, line 22: consider changing to “... show difficulties to outperform ESP forecasts…”
- References: Zhao et al. (2017) is missing; consider check papers “in preparation” and papers “under discussion” (Lucatero et al., 2017; Mendoza et al., 2017) to see if final versions are already available.
- Figures 3: I think the title should be taken out. All should be explicitly explained in the caption of the figure.
- Figures 5, 7 and 10: I suggest using letters a), b), c) etc. and explaining each graph in the caption (instead of using titles).
- Figure 6: consider indicating that it refers to lead time 1 month.
- Figure 8: consider using letters a), b), c) etc. and clarifying “raw meteo forecasts and postprocessed (QM) hydro forecasts, etc., in the caption of the figure (and not in the titles).
- Figure 11: consider reviewing the caption of the figure to explicitly mention that it is about “Forecasts of minimum daily flows for each year of the period 1990-2013, considering a seasonal forecast issued on April 1 for the next 7 months”