Dear Authors,
reviewer reports on your submitted manuscript are available now. Based on these, I agree your paper can potentially be published in NHESS subject to minor revisions as recommended by both reviewers and detailed in the attached referee reports.
To proceed with the closed peer-review process, please revise your manuscript following the comments of the reviewers and resubmit your paper accompanied by a detailed point-per-point reply on the reviewer comments, and a manuscript version in track change mode highlighting the applied changes. After resubmission, your paper will be reviewed again by the editor.
Best regards, A. Günther

Dear Authors,
Regarding your revisions, I feel you have mostly well replied on the reviewer´s concerns, but many of the additional information given in the reply letter is not provided in the revised manuscript. Please reorganize your paper incorporating all of your reply at appropriate positions in the text.

Referee #1 pointed out the high similarity between RS of site b and c, and in your reply you´ve indicated a mistake in the DSHA calculations for site b that should have been fixed in the revised manuscript. However, when comparing Figures 4 of the revised and original submissions I can see no difference in the RS of the sites. Please clarify this, together with a more in-depth explanation on the criteria used to assign specific earthquake locations, the general similarity of the RS, and more discussion why a source model developed for PSHA is suitable for DSHA. It should be explained why specific RS for sites a-e are required and not only one because of their similarity. Additionally, more explanation on the specific seismotectonic site characteristics of the six locations is required in the manuscript. Moreover, it is important to explain properly in the manuscript why local earthquake records were not exclusively used for this study. Adding a paragraph on this in the discussion is not sufficient, a justification why global earthquake records have been searched and why these are applicable to the Taiwan case (and site characteristics) using RS based on a local GMM should be provided in Section 3 of the manuscript.

In addition to the above general observations on your revisions in the light of the referee comments, I have some additional specific comments on the presentation listed below. Please revise your paper following these and the general comments above and resubmit your paper together with a detailed point-per-point reply and a manuscript version highlighting the applied changes. After resubmission, your paper will be reviewed again by the editor.

P5L110: Here you should also refer to Figure 2 where the sites are plotted as stars (although not indicated in the Figure caption). Please also denote site letters a-f for easy identification throughout the manuscript.
P5L118: If for all sites the area source is the controlling source, why do you provide the line sources and their parameters? Please explain.
P5L120: Isn´t the source-to-site distance zero in all cases since the sites are located in the specific controlling source areas? Please explain.
P7L181: In Figure 6, it cannot be observed that something is highlighted. Do you refer to the red rectangle in the Figure? This is not indicated in the caption.
P8L200-P9L222: Section 4.1: I am not sure whether a general discussion on pros and cons of PSHA vs. DSHA is really required here. It would be better to argue why DSHA have been chosen for this study and how it can be developed further in the Taiwan case.
Figure 1: This map has no scale. Reference to plotted stars (sites) is missing in the caption. Maybe you can plot more information from Table 1 in this Figure (e.g., source areas colored with max. magnitude classes).
Figure 2: Map has no scale. Is it needed since your DSHA only requires area sources? Please check.
Figure 6: The screenshot of the search engine is of very bad quality. Specifically, the search criteria are not visible, and the red box (probably highlighting MSE-ranges) is not indicated in the caption.
Table 1: Is it necessary to provide the parameters of the line sources?

Best regards, A. Günther

Dear Authors,
I am pleased to inform you your revised manuscript can now be published in NHESS.
Best regards, A. Günther