the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 20 Apr 2026
Freshwater sources and circulation in northern Greenland fjords from a multi-tracer analysis
Abstract. Northern Greenland fjords regulate the interaction between the ocean and two major marine-terminating glaciers, yet the sources and pathways of freshwater that transform fjord waters remain incompletely understood. During the GEOEO-North of Greenland 2024 Expedition, hydrographic observations together with water oxygen isotopes (δ18O) and noble gas tracers (helium and neon) were collected in Petermann and Victoria fjords to characterize water-mass and freshwater composition and to identify glacial meltwater pathways during late summer. The observations reveal pronounced differences between the two fjords; Petermann Fjord exhibited a relatively well-mixed surface layer under largely ice-free conditions, consistent with active ventilation and exchange with shelf waters, and in contrast, Victoria Fjord retained a shallow and strongly stratified surface layer beneath persistent sea-ice and dense iceberg mélange, conditions that likely limited air-sea exchange and enhanced freshwater accumulation in the upper layer. The combined hydrographic, isotope, and tracer framework indicates that freshwater in the upper layer of both fjords was dominated by meteoric input mainly derived from local glacial melt, while contributions from Arctic-sourced freshwater were comparatively small. Noble gas anomalies further identify glacially modified waters below the mixed layer, revealing a subsurface layer influenced by meltwater between roughly 100 and 250 m depth. Tracer distributions suggest that meltwater enters the fjords through multiple pathways, including subglacial discharge that influences the upper water column, and submarine melting that affects deeper layers by mixing with Atlantic-origin waters. These observations provide new constraints on meltwater pathways and circulation in northern Greenland fjords and help elucidate how ocean forcing can influence ice--ocean interactions in this rapidly changing region.
- Preprint
(10257 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
- RC1: 'Comment on egusphere-2026-1366', Anonymous Referee #1, 15 Jun 2026
-
RC2: 'Comment on egusphere-2026-1366', Anonymous Referee #2, 04 Jul 2026
Here is my review of ‘Freshwater sources and circulation in northern Greenland fjords‘ by Camille Hayatte Akhoudas et al.
Greenland’s glaciers experience widespread retreat, in part driven by warming subsurface waters entering the fjords. The present study presents exciting, novel, tracer-based datasets in two fjords of north Greenland, Petermann Fjord and Victoria Fjord. The multi-tracer analysis allows for a differentiation of meteoric freshwater from sea ice melt (dO18) and different glacial meltwaters in the meteoric pool (subglacial discharge, submarine melt, ancient melt). The study applies proven methodologies, and the authors demonstrate they know their art. In my view the study will represent a very valuable contribution to the particularly sparse observations of ocean-glacier interaction in North Greenland. I congratulate the authors for putting together this insightful study, and I apologize for the late submission of the review. The paper should be published in my view, however, subject to a number of clarifications, that I hope will increase the impact.
Scientifical, my main criticism is that the authors’ conclusion that freshwater enters the fjords in different ways lacks precision. A clearer interpretation of the glacial contributions and their pathways would strengthen the manuscript (see detailed comments below). The schematic of Fig. 1b could serve as a guidance for this.
Another general point is the treatment of uncertainty. In several places, comparisons between PF and VF are made, but whether differences are significant is not evaluated convincingly. I appreciate thaz the sample size (particularly in PF) is sometimes small, but still this point should be given more consideration (see detailed comments below). In this context, I would like the authors to state more clearly, how the two approaches (OMP and S-d18O) are compatible with each other in estimating water mass fractions. AW fractions are provided by both methods, PW is provided by one, but how is PW then implicitly taken care of in the other one (see comment below)? The S-d18O method yields AW fraction > 100 %. How should that be interpreted (what are the uncertainties) and what does this imply for robustness of the negative SIM fractions.
Another point that would make the manuscript clearer is an improved separation between results and discussion. Sometimes, an exciting observation / result is followed by a number of references to literature without the latter being discussed thoroughly. This reduces the clarity, in my view. I’d suggest to present the observations in a clear way first, and come back to important discussion points further to the end of the manuscript (see comments below).
Detailed comments
7 I could not identify a well-mixed surface layer (see below) in PF. PF is less stratified than VF, but a well-developed mixed layer is not apparent.
13-15 The conclusion that freshwater enters the fjords in different ways is certainly correct, but also lacks precision. A clearer interpretation would be desirable (see below).
General comment to methods: There is the S-d18O framework and the OMP analysis. I am not an expert of this. Can the authors explain, why it is OK to exclusively distinguish between AW, SIM and MW in S-d18O space. Or maybe asked differently, why is it justified to exclude PW here? Mathematically, one can only solve for 3 source water fractions, I guess. But is it explicitly assumed that a certain mixture of AW, SIM and MW fully describes PW? In the real world, PW also contains contributions of Pacific Water that are not accounted for. Is that a problem? Maybe another question related to this is. You have two separate estimates of AW fraction (both approaches yield one), one shown in Fig 5 the other in Fig A2. Are they compatible? The Fig. A2 suggest that AW fractions consistently exceed 100 % in the deep ocean. This seem to represent an issue with the endmember choice, I would think. At the same time negative SIM fractions are shown in the deep ocean down to 1200m. Is that physically plausible?
157-158 The choices of endmember values and error bars should be justified better.
221-222 I do not see a mixed layer. What criteria is used? Visually, waters are clearly stratified below 25 m or so, and even slightly stratified above that depth. Please adjust.
223 Certainly, waters in Victoria Fjord are more stratified ... Still, PG does not have a pronounced mixed layer. Therefore, I’d say the point made here on a pronounced role of mixing is not well-supported by observations.
228 ff Here the authors start with an interpretation before presenting the data / results. I think this should be done in reversed order.
230 I guess the authors refer to ‘exchange‘ with open ocean rather than ‘ventilation‘. Also, why is the exchange (i.e. the estuarine circulation) affected by fast ice?
234-235 It seems that temperatures at the surface are above the salinity-defined freezing point. Is this a signature of surface warming by solar radiation?
243 Only the shallowest measurement level shows a strong depletion, not throughout 50 m. Please adjust.
246-247 I do not see a major difference in 18O between VF and PF. Can the authors quantify the difference and and state an uncertainty?
247-248 Do the authors imply there is a stronger overturning circulation in VF than in PF (by entrainment of AW in a buoyant meltwater plume)? Later we see there is more SMW in PF at those subsurface depth layers. Is this a contradiction? Can the SMW inventory in PF be larger than in VF but the entrainment of AW into the plume ‘less efficient’. I don’t know the answer, but it would be an interesting discussion.
279 Replace matches by matching.
306 Where does the assumption come from? The authors should state why it is relevant and what it implies?
312-313 There are only two samples that this trend-estimate of 36% is buillt on, if I see this correctly. Aren't the error bars just too large to make such a precise estimate?
317 I can't follow. Time scales have not been mentioned. The authors should state what can be said about the different time scales?
320 There is no mixed layer in my view.
Section 4.2 I had to read the last paragraphs several times to memorize the results. Fig. 3 is also quite difficult to access because the symbols can't be differentiated where symbols cluster. Sometimes Victoria fjord is discussed alone sometimes both systems. Is there a way to present the results clearer? I suggest the authors maybe add a table where they list the findings (key results, implications). In general, the authors chose to interpret the results straightaway along with each result, rather than separating results and discussion a bit more (at the same time giving little room for precise comparisons to the literature). It makes it hard to understand at times, what is interpretation and what is not. Also, I found it difficult to grasp what the errors are. When differences are discussed (e.g. between the two fjords) how significant are the differences. More care should be taken to state uncertainty on statistical grounds.
330 In my view there is no mixed layer visible in the data provided.
333-335 Here the significance of the findings should be discussed. If the authors looked at median values, I guess there would not be a significant difference between PF and VF (imagine removing the largest and smallest value from the VF data set – both sitting outside of the fjord). Also results and interpretation are interwoven, hard to know what is what. Can the authors make the distinction a bit clearer?
336-337 The sampling size is low. Is it sufficient to support the authors statement?
341-364 The authors show negative SIM throughout the whole water column – even at 1000 m. I wonder what this means (along with AW fractions exceeding 100%). Could it be that the interpretation of the fjords being effected by sea ice formation rather than melting stands on shaky grounds, and that a slight change in the endmember choices would give different results? The authors should comment.
347-349 Can the authors be more precise in their statement. What do those other studies show? Not sure what the previously documented seasonal sea ice cycles refer to. Also, how dependent are the results on the endmember definition? Maybe in the fjords, the ice salinity could be quite low, and melting would provide less salt than assumed (by the endmember choice)?
352 The authors should state whether the difference compared to PF significant?
353-364 Does this explicitly refer to results shown in the study? The authors should state which spatial gradients is the reader supposed to look at? Or is this a discussion point (but not directly supported by the dataset provided). I am having a hard time to find evidence in the figures. If there is, the authors should guide the reader better.
375-376 How comes the SGD is only present at the surface samples while SMW has a subsurface maximum? I thought that SGD enters the fjord from land at the glacier base. Then it would initiate a buoyant plume entraining AW and melting the calving front (I guess). If then this plume settles (becomes neutrally buoyant at a certain depth) wouldn't that also arrest the SGD? If the latter sits at the surface, could that then represent surface runoff rather than subglacial discharge? The signal is very confined to the surface. The authors should try to exploit this exciting dataset more to actually disentangle the glacial freshwater pathways (see main criticism above).
385-389 Same question as before. The authors should state whether that means the SGD does not originate from the base of the grounding line. Does it enter the fjord at shallower depths? Maybe at the surface?
391-393 I find the statement quite vague. The authors should lay out more clearly what these multiple pathways might be. Listing a selection of paper did not make it clear to me what the authors' interpretation is.
396-398 There are a very view samples in PF near the surface. Is the difference to VF significant at those levels? I find it interested that there is considerably larger amounts of SMW below 100 m in PF than in VF (this looks significant too me) . If the authors did an overall SMW inventory between PF and VF, is there more SMW in PF? If so, does that mean the authors expect higher submarine melt rates in PF (having a floating ice tongue) than in VF? I guess the circulation strength in VF is essentially unknown. Still the authors should develop a view on this, particularly since the hydrographic properties show that T and S are higher in VF than in PF at subsurface levels.
407 ‘0.025–5 Sv‘ looks wrong to me. Please correct, if appropriate.
423-424 This statement needs to be explained in more depth, as vulnerability (meaning sensitivity to certain forcings?) of the glaciers has not been discussed. Are the authors maybe saying that their dataset provides a useful baseline for assessing future changes?
Caption of Fig. 4 ‘a-c‘ and ‘b-d‘ should be replaced by ‘a, c’ and ‘b, d’.
Citation: https://doi.org/10.5194/egusphere-2026-1366-RC2
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 233 | 148 | 20 | 401 | 18 | 19 |
- HTML: 233
- PDF: 148
- XML: 20
- Total: 401
- BibTeX: 18
- EndNote: 19
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
See attached