NHESS 1512

This paper is in the context of flash floods, loss estimation models, and flood preparedness. The paper introduces the FLEMOflash model, using data from past German flash floods; methodologically, it combines machine learning and Bayesian networks to estimate probabilistic losses and their uncertainties. In terms of topics, the paper is relevant for and aligned with NHESS.

The paper is well-written and -organised. Comments are mostly minor (even typos). The only major comment is about preparedness. From the paper, I do not understand what is meant by preparedness, and in specific what ‘high’ and ‘low’ preparedness mean. What are the assumptions behind ‘preparedness’? e.g. that people with more knowledge of risk will act in a certain way (which way?)? At page 14, it is said: ‘…doesn’t knew what to do’. For high preparedness, what people know about what to do? The model seems suited to derive the predictive density of losses, however I have doubt about the effect of preparedness. I would be very cautious to include this part in the paper.

A secondary comment is that I would add some background about the previous/traditional version of FLEMO (e.g. https://www.gfz.de/en/section/hydrology/projects/4-flood-loss-model-flemo-for-residential-and-commercial-sectors); there is none at the moment I think.

Specific comments (P for page, L for line):

• Valid for all direct citations: coma is not needed before the year, e.g. Smith et al. (2000) - and not Smith et al., (2000)
• Valid for the whole paper: equation factors, such as rloss, need to be in italic in the main text of the manuscript
• Valid for the whole paper: do not use contracted forms like ‘doesn’t’
• P2L42: double parenthesis in the citation
• P2L50: double space before ‘significant’
• P4L101: double space before ‘The percentage’?
• P7L146, P11L236: ‘This’ what? Add a noun, specify
• P9L190: remove the dot before the parenthesis of Fig. 1d-e

See attached document