Summary

This paper presents observational hydrographic data from Disko Bay, Greenland, which sits near a large, fast tidewater glacier draining the Greenland Ice Sheet (Sermeq Kujalleq). They combine existing data sources from a nearby monitoring station (GEM site), two profiling floats, and re-analysis atmospheric & satellite data over two seasonal cycles (the years 2022-2024) along with weekly repeat stations taken over Fall 2023. In the end, they describe a seasonal cycle that is consistent over both years, showing both the modulation of the mixed layer and when the warmest and densest Atlantic-origin waters arrive at the mouth of Ilulissat Icefjord (which leads to Sermeq Kujalleq, SK hereafter). Overall, they present a detailed account of the hydrography in the bay’s water mass layers, which should be beneficial to future studies looking to oceanic change as a driver for mass loss from Greenland. Although many pieces of the study are already known, I find the most novel results to be their discussion of the deep-water renewal and lifting of the Atlantic-origin waters above sill depth at the mouth of the icefjord. However, the paper seems overly complicated for the conclusions it reaches, and the figures could be clarified and improved given all the disparate sources of data. For example, explicitly stating what is novel here and what is validation of previous conceptual models of the Bay’s exchange with the coast and the icefjord would be helpful. One way to do that would be to update or revise the schematic from Gladish et al. part 2 (his Fig. 15). But there are other ways too. I was also not convinced of the Fall along-isopycnal warming signal they attribute to outflowing glacially modified waters (GMW). Finally, a new paper (Picton et al. 2025) came out in JGR-Earth Surface recently that uses some of the same data (floats and GEM site), although their focus is slightly different.

More on these comments below, as well as line by line comments. I certainly believe these data and this write-up to be publication worthy, but their impact will be significantly improved through revisions.

Major comments:

1) During the discussion period, another paper came out that studies the same region with some of the same data sources (Picton et al. 2025). It would be good to include this reference in the revision. For example, that paper shows temperature vs time at 240 m depth and seems to show a similar Fall warming and seasonality as described in this current study. I think discussion of this new publication will only strengthen this one. In terms of other references, I believe the authors could certainly expand their sources. I point to a few examples in the comments below.

2) In terms of the along-isopycnal warming trend they document and discuss, I am not convinced it’s entirely from the outflow from the icefjord. I think what they are saying is that the summer outflow signal is delayed and lagged as it transits Disko Bay, so the warming they see through the Fall is from that entrained WGIW. However, the melt season at the glacier (e.g., look at a subglacial discharge time series for SK) is relatively short and the warming seems to continue beyond the length of a typical melt season. Although they calculate advective time scales for other processes in the paper, they don’t estimate the time scales here. Are there other mechanisms that could cause this warming trend through the Fall?

3) The presentation quality is average. I think the figures could be improved for clarity and I mention specific instances in comments below. Also, in all figure captions, it would be helpful to specify what data sources were used, as it was hard to follow along and keep things straight, given there was the ‘monitoring’ site, the floats, and the 2018 spatial survey all used for different things. Simplifying the results section would help here, as a lot of discussion seems to occur in the results, in addition to the formal discussion section.

4) The conclusion that GMW entrain Atlantic-origin water in Greenland tidewater glacier fjords and produce outflows with temperature anomalies relative to their depth is not new. This has been shown over and over again in many different systems (e.g., Straneo et al 2012 looked at several sites around the ice sheet). Many more references since have examined this dynamic.

5) In the results and abstract they mention that renewal can occur in Fall given upwelling winds. However, it appears like there were consistent upwelling winds in the second year as well, yet no similar renewal occurred. Was there a reason for that? It seems hard to claim that it is part of the seasonal cycle if you have it observed for 1 out of 2 years of data collected, i.e., some other process(es) must be relevant.

Line by Line Comments

Line 25: AW can enter shallower than 200-250 m too (as you show later in this paper!), but the sill sets the maximum depth of water that can flow into the icefjord.

Line 40: I am a bit worried throughout by the use of a summertime 2018 spatial survey with the time series data from 2022-2024. Summer 2018 was at the end of the cool period noted by other authors. Is there some way to give context for the interannual variations in these data? Maybe the new Picton et al. paper could help.

Line 43: Using these initial guiding questions to structure your discussion might be a way to tie the paper together more. Otherwise, they do not need to be included here. Overall, the tone of the paper is informal and often falls into the trap of (i) telling us what you are about to write about, (ii) telling us, and then (iii) summarizing what you just told us. That lengthens the paper quite a bit. I would try to cut out the preambles on all your results, etc. and just get to the point. One example is at line 79 where you could delete “Below we describe each dataset”.

Line 59: Is the WGCC relevant to Disko Bay given this uncertainly around whether it even exists at this latitude? That is, does it matter if it’s the WGC or WGCC?

Figure 1: I think you should try to add on the 2018 survey locations to this figure, so all the observations are on one map. They could be small dots. Also, it is impossible to see the crosses used for the float’s first profile. Finally, can you label Vaigat Strait here too since it is mentioned a couple times in the text?

Line 103: float data ‘were’ (data are plural)

Line 110: Not clear what CTD observations the float data were compared to?

Line 129: Unless the Semper et al 2024 paper is published, it is not helpful to cite here. I would outline the method more completely or maybe show an example or two in a supplement. It’s also not stated what threshold you use for the ‘sharp increase’ in normalized sum-of-squared errors or if the final MLDs are sensitive to this threshold. Finally, it’s a two-step method and I clearly see the ‘first’ step. Is the second step the checking with the 1 standard deviation envelope?

Line 135: The PW layering in Fig 2b is very muted, especially compared to other profiles and systems I’ve seen published. I assume this is because there are many profiles with surface layers that are not PW? I know that you lump all cool, fresher waters together in this PW layer, but it might be stated more up front before talking about this figure. In addition, the different sources/types of PW presumably have their own seasonality (and in fact some of your discussion of how the mixed layer changes over the year is linked to this).

Line 154-155: Is a mixed layer always present? Inside the icefjord and I imagine in Disko Bay at least at times there is a stratified surface layer due to ice melt or other freshwater inputs.

Line 156-162: I think this paragraph could be shortened, specifically the part about icebergs. Just say they are part of the freshwater flux.

Line 172: For the wind stresses, you show multiple equations for the drag coefficients, but never the equation for the actual wind stress (which presumably has the drag coefficient in it). Given the relatively small amount of discussion on the Ekman pumping calculation, this section seems long.

Figure 2: Note that iceberg or glacier melt would pull water masses along a line with the same slope as the SMW line, but not necessarily on that exact line shown. That is, the slope is important, but not the exact location and that is true for SGD too, as it matters where the outflowing plume reaches neutral density. All those pink dots in the GMW section seem like they are being pulled down the melt line, but this isn’t clear from the figure or the text.

Line 202: ‘Both’ is wrong, as you list three things here...and then you have a 1 sentence paragraph.

Fig 5,6,8: I wonder if these figures could be combined somehow for more visual impact and being able to compare the timing. I think this is where my confusion over the source of the deeper warming comes in, as these figures are all getting at the same thing but are separated.

Fig 8 and analysis: Float 2 seems to stay relatively stationary but Float 1 moves quite a distance. In Figure 8 and in the results, how much of the ‘Float 1, 2022-23’ variability is due to time and not a spatial gradient? It’s interesting that the second-year variability is lessened given that that float didn’t move as much.

Line 263: ‘representation’ of what?

Figure 6: The colored dots on the T-S diagram are fairly hard to distinguish. Either try different colors or make them bigger markers.

Figure 7: Labeling regions on Fig 7a would be helpful.

Line 280: But the subglacial discharge that you surmise is causing this is certainly a transient feature, as it ramps up each melt season and dramatically ramps down by end of August or Sept. If the along-isopycnal warming commenced in August, does that imply a 2-month lag time in the icefjord for waters to come out given Float 2’s position? Does that jive with estimates of the icefjord circulation/advective scale?

Figure 9: It took me a while to infer these data were from the floats (I think?). Indicating that in the caption would be helpful. Also, you mention shading for d-e, but the shaded areas are different. There is also no mention in the text of the large data gap in the second year, which precludes the specification of the start of the renewal.

Figure 9 discussion: Again, as for the float data above, I would like some justification for using the float data as time series again given the movement of Float 1. How much variability would you expect over this distance (maybe you can use the 2018 spatial survey to estimate this)?

Line 319: I don’t follow the statement that wind forcing was not evident during this period. It certainly looks like upwelling favorable winds were present at that time?

Line 323-334: A lot of this paragraph would fit better into the discussion. There is some redundancy between the results and the discussion, which I think leads to some confusion and adding to the length of the paper.

Line 333: I don’t see 1.5C, maybe 1C increase?

Line 359: You cite ‘typical’ ranges here from previous work, but not anything about the range? That is, is it reasonable for this feature to be advective given variability in these conditions?

Line 363: You cite both SGD and SMW are sources of freshwater, which is true. But they have one fundamental difference. For iceberg melt, the depth of entrainment is much much shallower than for SGD, so it’s really the SGD that is controlling the depth of the GMW layer (well, and the sill). There is some new literature on refluxing at the icefjord sill (Hager et al. 2024) that might be useful as well to explore. This whole section is relatively long for describing a process that other studies have shown already. It’s not really the novel or interesting part of this present study- that’s more on the WGIW properties in my opinion.

Line 393: do you mean ‘width’ here instead of height? As you also have draft? I’m not sure any of those papers actually have measurements of iceberg draft, but assume some geometry based on above water volume and ice/ocean densities.

Section 5.3: I am not sure this section says very much. A lot of it is speculation about how the results might be important to marine ecosystems. The last paragraph in particular is vague. Adding some meat/content would be good. For example, can you show the cyclonic circulation sense in the 2018 spatial survey (not just in T/S but in dynamic sections)?

Line 446: ‘de-seasoning’ is not a word.

Line 453: This sentence and the one before it are confusing. They say nutrients are entrained deep in the icefjord, flow out into Disko Bay and then into the icefjord? Maybe I’m missing something here.

General comment: Is it possible to update the schematic or create your own schematic (Fig 15 of Gladish et al (part 2))? This might help the reader understand what is new here or what is validating previous ideas (which is important too). I think the data here are very cool!  

Based on the comments above, I’ve scored the manuscript using the principal criteria (score 1-4, 1 is excellent and 4 is poor) from the OS website.

1) Scientific significance:

Does the manuscript represent a substantial contribution to scientific progress within the scope of Ocean Science (substantial new concepts, ideas, methods, or data)?   2

2) Scientific quality:

Are the scientific approach and applied methods valid? Are the results discussed in an appropriate and balanced way (consideration of related work, including appropriate references)? 2 (in general it could benefit from a more diverse set of references)

3) Presentation quality:

Are the scientific results and conclusions presented in a clear, concise, and well-structured way (number and quality of figures/tables, appropriate use of English language)?    2.5

This manuscript by Latuta et al. presents analyses of hydrographic dataset in Disko Bay, Greenland. It provides insights into the seasonality and spatial variability of the water masses in the bay system and gives explanations on the potential physical mechanisms that drive the seasonal cycles. Overall, I believe the manuscript would contribute to further our understanding of the hydrography in this area and the finding are significant and timely.  But I also think the presentations could be improved and some clarifications should be made before acceptance.

General comments

1. A key explanation shown in this work is the identification of wind-driven WGIW renewal via upwelling-favorable conditions at the EDS. However, this mechanism (e.g., around line 305) is complex and may be difficult to visualize for some readers. I recommend including a schematic figure showing the seasonal wind changes, upwelling over EDS and the resultant basin renewal.
2. The manuscript suggests that upwelling-favorable winds caused WGIW renewal in November and December 2022, but the upwelling effect was not as effective during autumn and winter 2023-2024 (50m uplift; line 317). It would be helpful to provide an estimate of the threshold of the wind stress or its duration needed to cause upwelling sufficient to lift WGIW over EDS. Is there any lag between wind stress and the dense water renewal? Could the authors quantify this relationship?
3. The sampling frequency in Table 1 seems a bit long to capture variability in shorter time scales. It would be great if the authors could discuss whether the variability in higher frequencies would affect the water mass exchanges in the bay system.

Specific comments

Figure 1 and lines 61-70: I notice the lack of labels for locations in the diagram, and it was difficult to follow the statement without further searching for another map online. Please include some key locations such as Vaigat Strait to help the reader navigate.

Line 165 and Figure 2 caption: I realize the mixing line definition is not explicitly indicated in the Figure 2 caption. It would be more straightforward if the reader could find the definition of the mixing line end-points without referring back to the T-S statement in the manuscript.

Figure 7a: please label the boxes with eastern, northern, and central.

Line 319: it seems that in March and April 2024, the wind forcing and upwelling were strong (the same period in Figure 9c as the orange shading in Figure 9ef). Why does the statement here say “not evident”?

Figure 10: please label the panels (a) and (b).