The brief communication by Kiesel et al. provides evidence that the numerical models must be improved to simulate extreme wave inundations. This is true because the numerical models still do not accurately accommodate the temporal and spatial variation in macro-roughness during flow-structure interaction.  

The communication only discusses the water levels (depths). However, water velocity is an important factor for high-energy wave impact. The mismatch between the predicted and real hydrographs would be due to the flow velocity effect but not completely attributed to the assumptions. The authors need to include a section on the significance of the flow velocity of this event and how it could affect the shape of hydrographs.

The authors selected a 200-year return period for simulating water levels.  Was the tidal effect included in the simulation?

What is the sensitivity of the water level to the return period?

The authors have presented the challenges of coastal hazard and risk modeling through a case study of the storm surge levels and losses observed in one recent event, for the purpose of motivating future research agenda. 

While the authors’ analysis of the recent Oct 2023 event with respect to available hazard estimates is novel, their highlighted gaps have been extensively discussed in the broader disaster hazard, risk, and resilience literature. As a Brief Communication, the manuscript can be valuable to the scientific community and policymakers with some improvements.

    1.    The title and the abstract were not clear and did not fully reflect the goals of the manuscript. “Insights from a [event]…” is very broad and could be narrowed down, for example, as “Modelling Gaps…”. Similarly, the last statement of the abstract mentions proposing “strategies for improving flood risk modelling…”, however, the manuscript appeared to focus mainly on highlighting gaps rather than proposing strategies for bridging those gaps.

    2.    In Section 2.1, the authors highlight the differences between the 200-year hazard scenario with observations from the Oct event. In order for them to match better, the authors propose inclusion of spatial correlation, and using 60th percentile duration for the 200-year event. However, these modifications would not guarantee that the updated hazard would match any future event. This is commonly observed in hazard modelling as hazard only represents site-independent exceedance levels, and is necessarily different from individual scenarios over a wide geographic region. Inclusion of spatial correlation can in fact help in providing better cumulative loss estimates, however, the authors did not include any comparisons between cumulative loss or damage risk estimates with the observed losses in order to support their proposals.

    3.    The intent of the manuscript will be clearer if a section on fragility and vulnerability functions was included, especially for dikes, and how these functions are used for loss and risk estimations. This would further support the authors’ arguments in Section 2. It will then be clearer that Section 2.2 highlights the need for developing accurate fragility functions for dikes based on their current deterioration state; and Section 2.3 highlights the need for developing conditional fragility functions based on the current damage state of dikes from previous events. This discussion can also improve Section 2.1 by highlighting that the fragility functions are dependent not only on peak storm levels, but are functions of the temporal evolution, and other factors such as flow velocity.

    4.    Line 139 - Minor language - …hydromorphodynamic *modelling* (Hinkel et al., 2021; Vousdoukas et al., 2018a). Ways forward *are* ultimately depend on the availability…