I agree with the 2 reviewers who highlighted that this is an interesting manuscript nicely highlighting the case study of Hurricane Irene. I enjoyed reading it and only could find minor issues, which have been also highlighted by the reviewers. The response by the authors to the reviewer is adequate and will answer the remaining questions.

I thank the authors for their comprehensive resubmission. One of the reviewers has unfortunately missed the deadline for submitting a response to your comments. He has emailed them and I believe that they are very valid. Can you please address them with a small paragraph in your paper:
Comment by Mohammed Al-Arag:
Your work concerning uncertainty is interesting. Your response seems to suggest that your current work can replace traditional hydraulic modeling/methods and I take exception with that for a couple of reasons:

1) Adequate/detailed river cross section geometry is generally available for streams/rivers in NJ/NY and is the case for a well-studied river like the Hudson River. With the advent of LiDAR, river geometry is obtainable and with a greater degree of accuracy then conventional survey methods that have been traditionally used and at a far less expense.

For one of your responses you wrote: “hydrodynamic modeling would require detailed representation of river cross section geometry that is not available at regional scale”. This detailed representation of river cross section geometry is needed to create an accurate/valid hydraulic model. Your approach reported in this current study is not adequate to replace hydraulic modeling.

2) Indeed boundary conditions (BCs) are needed for hydraulic modeling. BCs can be a rating curve or known water surface elevations (WSEs). USGS maintains (calibrated) gage stations throughout the region and the Hudson River, data is readily available for known WSEs during storms of record in addition to 100-, 500-year discharges.

Thank you for responding so well to the additional comment. It is a very nice manuscript now which reads well and should be published in HESS.