The paper “Bayesian extreme value analysis of extreme sea levels along the German Baltic coast using historical information” discusses the inclusion of historic extreme events, often disregarded as outliers, into extreme value analysis and adapts an established approach to the German Baltic coast. This new methodology is able to reduce uncertainties at some of the locations of the study and to shows that current design approaches might underestimate extreme sea levels significantly due to limited availability of long tide gauge records. The paper further discusses large scale variability found in the Travemünde record, which remains undefined but might be the reason for the underestimation of extreme sea levels in other records.

The authors present their research in a precise and easy to understand manner. Their work highlights the next step in possible improvements to the coastal protection in their study area and given a well-founded recommendation on the treatment of outliers in EVA.

I therefor recommend the publication after some minor revisions, mainly focused on some clarifications and extended discussions of its important findings.

Please find further general and specific comments in tha attached pdf.

[General comment]

This manuscript presents the evaluation method for the extreme sea level by performing extreme value analysis using tide-gauge observations with historical information. The records of exceptionally large events, which might be excluded as outliers, are used to perform EVA. The results indicate that the extreme sea level may be underestimated in current designs and the proposed method can reduce uncertainties. This topic is interesting and the manuscript is generally well-written. Utilizing exceptionally large event data in a limited duration for evaluating extreme sea levels is very important. Therefore, the manuscript is valuable to be published after some minor revisions.

[Specific comment]

1. For Fig. 1, it might be easier to understand if the place names are written on the map.
2. For the size of the cycle in Fig. 2, it would be easier to understand by adding the legend of magnitude.
3. The “exp” in Eqs. (3), (4), and (10) need to be modified to be normal, not italic.
4. Some abbreviations, such as NHN at p. 8 in line 171 and MLE at Fig. 3, are used without explanations. Please check for missing explanations of abbreviations.
5. For Line 192, how much impact does the threshold variation or the number of used historical information have on the extreme value analysis?
6. In line 228, why 70 is adopted for years of the moving window?

The manuscript by MacPherson et al. presents the importance of considering historical information, primarily related to large events, in the extreme sea level estimation. Including such data, instead of treating them as outliers, it can reduce the uncertainties to better plan for coastal adaptation. I find the article to be of very good quality, well written, with a detailed methodology and with a good discussion. However, I do recommend some improvement in the results section.

The findings presented here are not only of great interest to the coastal research community but also to coastal managers. In this sense, my overall impression is positive, and I recommend the publication. I nonetheless do have some minor comments that may help to improve the manuscript by a bit.

Please refer to the attached pdf for further specific and minor comments.