Please see the attached document.

Review for „Modelling the effect of submarine iceberg melting on glacier-adjacent water properties“ by Benjamin Davison and co-authors 

The paper by Benjamin Davison and co-authors combines a general circulation ocean model (MITgcm) with a submarine iceberg melting module, in order to investigate the impact of iceberg melting and cooling, and their vertical distribution, on fjord water properties close to Arctic glaciers. The work is to my understanding an extension of the 2020 study by Davison et al. that was focused on a single Greenlandic tidewater glacier, by studying different likely iceberg “scenarios” and different simplified geometric fjord configurations in order to cover the wide range of Greenland fjord configurations. The work has potential implications also for efforts to project Greenland Ice Sheet behaviour with models, which usually can make use only of far-field properties beyond the fjord’s mouth to force these ice sheet/ice shelf components, so the scientific relevance of the study is high; and the paper should, in my opinion, be published soon.

I think that the model setups are defined very elegantly in order to answer how iceberg melting affects glacier-adjacent water properties. While almost all simulations show a cooling in the upper 60m or so, below that level either warming or cooling can occur depending on the “icescape” and configuration.

Specifically, the paper implies that projections for the large fast-flowing Greenland glaciers that contribute most negatively to the mass balance are potentially affected by the lack of iceberg effects on fjord water properties (hosting numerous and large icebergs), and this is very clearly shown with simple model configurations. These “details” can potentially matter a lot for the “large-scale” mass balance of Greenland. Notably, the authors even provide a first idea for simple parameterizations (l.434-436) in their paper and I hope that these ideas will be picked up in the community promptly.

The last paragraph of section 4.1 attempts a comparison to observations with some success. Here, it would have been great to close the circle by saying more clearly (or even plotting in the same panels) which simple six model configurations can mirror panels a-f in Figure 8. Or in other words, which assumptions are needed to model profiles similar to the observed profiles (e.g. presence of a sill) with the simple fjord geometry used in the study.

The paper is written and organized excellently (with no obvious typos, which is rare), and the arguments are easy to follow. All results are clearly described and discussed and the conclusions are based entirely on the model results. The quality of the figures is also okay. Below, you can find a short list of line-by-line comments that the authors could still work on. I suggest to accept the paper with (very) minor revisions.

Thomas Rackow

##################

Line-by-line comments:

l.89/90 Here or somewhere else, I think the high-impact study by Schaffer et al. (2020) should be mentioned who conclude that near-glacier sill-controlled ocean heat transport can play a crucial role for glacier stability.

Reference:

Schaffer, J., Kanzow, T., von Appen, WJ. et al. Bathymetry constrains ocean heat supply to Greenland’s largest glacier tongue. Nat. Geosci. 13, 227–231 (2020). https://doi.org/10.1038/s41561-019-0529-x

l.94-109 What is the surface boundary condition at the atmosphere-ocean interface? I missed that somehow and it would be good to know whether this could influence the surface representation of the profiles (e.g. holding them close to some value).

l.129/130 I think it would be good to also cite the much earlier Hellmer & Olbers (1989) study for the three-equation formulation, which is often forgotten:

Hellmer, H., and D. Olbers (1989), A two-dimensional model for the thermohaline circulation under an ice shelf, Antarct. Sci., 1, 325–336, doi:10.1017/S0954102089000490.

l.396-398 The constants are also from Jackson et al. 2020?

Section 4.2: You tried to explore the relative change in submarine melt rate quantitatively, which is great. To my understanding, this is a simple diagnostic and the model does not see the different melt rates. I was wondering whether any feedbacks are to be expected, or whether your conclusions might be different in a model setup that would account for the iceberg-induced melt changes?