The ability to predict water levels on a sub seasonal scale is obviously desirable in terms of coastal flood inundation risk. This paper is partly a follow on from the 2022 paper “A novel statistical approach to predict seasonal high tide flooding” by Dusek et al  Whereas in that paper they used damped persistence derived from monthly non-tidal residuals to predict seasonal flooding, in this paper they use two operational forecast systems to make predictions of the non-tidal residual component of coastal water levels on United States coasts. They find that the skill of the forecasts exceed that of damped persistence but also important is the inclusion of sea level trends themselves.

Overall the paper is well written, the work scientifically sound and the conclusions appropriate. The figures are good but perhaps take a little bit of time to understand and interpret.

Minor questions/comments

Page 2 Line 50 Sweet and Zervas 2011 is not in the references.

Page 3 can you explain the forecast models a little more. For example what are their time resolution, how many forecast steps are there or how far ahead do they predict. Later on we see that its up to six weeks.

Page 4. Line 113

You say that “All reforecast and verification datasets are determined using seven day running mean anomalies” This to me is not very clear. What is you basic time step in this. The water level data is generally 6 minute or 1 hour or have you gone to daily values? Are you taking these values ad producing a seven day running mean or are you removing a seven-day running mean?

Line 114. What do you exactly mean by 20- or 25- year climatologies? You say in the supplement the first four harmonics (plus the mean). So I would assume that would mean Sa and Ssa but what is generally the other two?

Page 9, Line 151. Sweet et al 2022 seems to be missing in the references

Page 10. Equation 2. You talk about IBE and VLM corrected but I think you should be clearer on what you mean. Are you adding them on to one thing or removing them from another. I presume you are adding on to the forecasts to make the predicted NTR as close to the observed NTR as possible.

Page 16. Line 278. I don’t know “increase from north to south” seems better than “decrease from south to north”! Or even just “increase towards the south” or “decrease to the north”