Dear Authors,
thank you for your detailed replies.
In agreement with the referees’ comments, I believe the manuscript needs major revisions.
On many issues raised by the referees you already state in the replies that you will add the modifications/details that are asked for. In reference to other comments you explain or detail the issue in the reply, but do not state that you will modify the revised manuscript accordingly (one of such examples is the addition of the steps from Eq 1 to Eq. 2): I suggest you to always do so (if the referees, who are experts in this research area, cannot understand the meaning of a paragraph or of a figure, the ‘average’ reader will have even more difficulties…).
1) One of the main issue is that of the data employed for calibration and validation of the hydrological model: the use of the observations and meteo forecasts of the 3 years should be clarified, as highlighted by all referees.
Reformulate ll. 13-21 of page 15815 to clarify how the hydrological is initialized (see Ref2’s comment).
Concerning data availability and the use of field measures (often not available in real-world applications), you should also better clarify which model parameters are derived from measures and which have been calibrated, see Ref1’s comment (in fact nor the caption nor the text referring to Table 1 state that such values refers to parameters that were optimised; and, if a calibration was performed, some details on the optimization procedure must be added).
And it is absolutely fundamental the addition, that you now intend to carry out as specified in your replies, to add a comparison between the simulations obtained with or without the use of the PREGI platform.
2) The second important issue is that of skill assessment and, related to that, the better definition and meaning of the thresholds, as raised by Ref2.
Ref2 suggests to use Brier Skill Score for assessing the improvement of the proposed approach in respect to an unskilled standard forecast: if you do not have climatological information on the pilot case study, you may use persistence as a standard for reference.
The meaning of the threshold is not well-defined, too: rephrase ll. 1-5 p. 15824, clarifying the period of cumulated rainfall and the meaning of such thresholds; the phrase ‘ the last two values are quite equivalent’ is indeed not clear as highlighted by Ref2. Meaning and interpretation of Fig. 7 must be clarified too (mening of BS in reference to the thresholds).

I believe, too, that some of the doubts/concerns of the Referees are due to the English syntax: a final revision of the language is now done for every article in HESS by the Editorial office, but of course if the English is improved and made clearer in the revised manuscript it would greatly help the second revision process (and also the following English editing). I warmly suggest to ask a colleague to revise the manuscript.

I look forward to reading the revised version.
Best regards,
Elena Toth

Dear Authors,
the revised version is indeed much improved, as confirmed also by both Referees.

The main comments raised by Referee#1 all require that you better clarify the final objective, applicability and merit/novelty of the work, I would ask you do so in the final version.

In addition to a list of suggestions for improving the presentation (see both Referees' comments), Referee #1 suggests also a re-organisation of the section presenting the Methods that I agree may really improve the readability and understanding of the paper by the Readers.

Thank you for having submitted your work to HESS and Iook forward to see the final version.

All the best,
Elena Toth

Dear Authors,
the presentation is further improved through your last modifications and I believe that it is now much more easy to follow.

I would suggest the additional corrections below:

Abstract: modify, “…based on ensemble predictions (20 members) …” with “…based on meteorological ensemble forecasts (20 members) …”

Title of section 2: “Case study, models and methods”

Sect. 2.2 (l. 3, p. 5): define REPS acronym here.

p. 5, ll 25-26: add that the model in the chain will be run also with forecasted weather data and not only with the ground stations data.

Add in title of section 3.1 that it refers to FEST with the use of observed weather data.

p. 11, ll. 1-7: this paragraph does not belong to this section (it refers to meteo forecasts) and should be moved/merged in 3.3

Add in title of section 3.2 that it refers to the coupling of meteorological and hydrological forecasts

p. 14, l. 9: ‘monitoring’ rather than ‘control’ of soil water content

The new Table 2 may be improved:
-add the day along with time in first column (counting the days from the start of the chain)
-In column three I would put all verbs in the present tense
-why are the REPS output uploaded after 60 hours if the time needed for running the WRF model is two days (48 hours)? please add explanation.

The English definitely needs a revision, which should be made by the Editorial Office.

Thank you for having submitted your work to HESS and I look forward to seeing it published.

All the best,
Elena Toth