This paper is a nice analysis of marine heatwaves with depth from an ocean model, which emphasises the role of model drift, baseline choice and resolution on marine heatwaves detection and statistics. Vertical coherence and drivers are also discussed as part of a case study.

The paper is interesting, and a good contribution to the field. However, I think that recommendations around baselines and spin up periods could be clarified, and believe that the heat budget analysis, while not incorrect, may not be fit for purpose.

Major Comments

Heat budget

I have concerns about the application of the heat budget. As used in the paper, it allows the diagnosis of the drivers of changes in heat content. However, MHWs are defined as discrete threshold exceedances relative to a local climatology, not by absolute temperature tendencies. Thus, areas of persistent heat convergence are not necessarily directly comparable to discrete MHW events. The manuscript currently suggests a causal link between sustained heat convergence and MHW occurrence or vertical coherence, but does not address this fundamental distinction. In order to prove this causal link, I believe that the heat budget would have to be performed on an event-by-event basis, and vertical coherence would have to be considered very carefully in terms of the boundary conditions for the budget for each event.

While this would be a very interesting analysis, I think it would be beyond the scope of the paper in its current form. In fact, I do not think that the heat budget analysis adds much to the outcomes of this work, and so my recommendation would be to restrict the analysis to that of vertical coherence, and remove, or at least strongly tone down and place the heat budget in the context of warming, and not of MHWs.

Recommendations

The paper makes recommendations around the use of sufficient model spin up periods, as well as the resolution at which MHW statistics should be calculated. I think that these recommendations will be very useful to researchers planning experiments for MHW use. However, I think the recommendations could be clarified in the text as they are sometimes not clearly laid out.

Is it possible to include more detail about the spin up process required? How much drift is removed in, for example the 4^th cycle as opposed to the 6^th? Would a repeat year forcing spin up be sufficient?

I find the finding about resolution very compelling, i.e. that while high resolution is needed to resolve mesoscale processes, MHW statistics can, in most cases be calculated on a coarser resolution grid. I note that this finding is not emphasised in either the abstract or conclusions of the paper and would suggest that it should be.

Minor Comments

I found the terminology and methology related baselines a little confusing at times. Smith et al (2025) recently published a detailed investigation of the effects of different baselines. For consistency with future literature, I would suggest citing this paper, and adopting their terminology (e.g. ‘detrended baseline’ instead of ‘linear baseline’)

I’m confused about the difference between model drift and real temperature trends – using a detrended baseline will remove both indiscriminately, while using a long enough spin up will remove model drift, but preserve natural temperature trends – is this correct? If so, it should be made clearer in the manuscript.

References

Smith, K. E., Gupta, A. S., Amaya, D., Benthuysen, J. A., Burrows, M. T., Capotondi, A., ... & Wernberg, T. (2025). Baseline matters: Challenges and implications of different marine heatwave baselines. Progress in Oceanography, 231, 103404.

I include my comments on the attached pdf