Dear authors,
We now received the comments of the two referees, as well as two independent “short comments”, to all of them you provided an author’s response.
I will now summarise the most important questions raised by these four colleagues:
(1) The EPIC model was used to predict daily water stress. However, the EPIC model was not seriously introduced at all. An independent section and the corresponding literature reviews should be added to provide sufficient information of EPIC.
(2) In conclusion, what will be message to the general public by the proposed work? What could be the limitation(s) of the proposed work? What might be improved in the future based on the proposed work? The conclusion should be substantially revised accordingly.
(3) The description of ecotone in 2.1.1 should be shortened and most of the section should be moved into the section of introduction.
(4) China meteorological data sharing service system of China only offers sunshine hours. So, how to transfer the sunshine hours to global radiation?
(5) The authors should offer the genetic parameters of maize used in the EPIC model for the three sites in Fig.3. Moreover, please provide the station name of six validation sites.
(6) Please provide sowing date of maize, planting density of maize, fertilization amount used in running EPIC model at three representing sites under sufficient and no irrigation conditions.
(7) The linear regression curve seemed more appropriate to fit the data than the logistic curve. So, why you select the logistic curve as the physical vulnerability curve?
(8) The discussion section should be strengthened by comparison with previous studies, including the impact of drought on spring maize in the farming-pastoral ecotone, the measures used to adapt to climate change, etc.
(9) It is important to include in some part of the introduction the differences about hazard, vulnerability and risk, because sometimes are used indistinctly. For example, the author can use as basis the terminology used by UNISDR (https://www.unisdr.org/we/inform/terminology).
(10) It can be seen in figure 5 that the drought hazard intensity index has a cyclic behavior with a return period of 20 years aprox. How is this considered in the risk assessment?
(11) It could be illustrative to include in the conclusions the weaknesses and limitations of the approach.
In your answers you promised to adjust your manuscript accordingly, even if this was not in detail shown for all of these questions. So please be aware that you follow step-by-step these suggestions and provide us with a revised version of this manuscript including a track-change version, as indicated in the author guidelines of NHESS. The revised version will be further reviewed, preferably by those colleagues who did review also the first round in the underlying discussion paper.
Wishing you good success I am looking forward to receiving your revised piece of work.
Kind regards, Sven Fuchs (Editor NHESS)

Dear authors,

based on the referees comments I am happy to inform you that your revised manuscript is accepted for publication.

Kind regards,

Sven Fuchs (Editor NHESS)