The authors present a comprehensive biomarker data set from three NE Greenland shelf records together with instrumental/observational data for sea ice fluctuations over the last century. A well-balanced and interesting manuscript with plenty of new datasets certainly of interest for the readership of EGUsphere. I would like to highlight a few critical points that might be addressed before the manuscript can be accepted for publication:

• First, the authors have access to bulk organic information including TOC, TN, and d¹³C_org While the bulk organics give you a comprehensive overview on the organic matter sources, the biomarkers cover only a tiny fraction of it. You could use the data better to inform the readership of dominant organic matter source in the records. You may even consider a rough semi-quantification of marine and terrestrial organic matter and use is more actively for your interpretation. The d¹³C_org data vary between -23 and -27 permille implying quite a bit of variation in terms of terrestrial organic matter supply to your shelf system.
• The authors (desperately) try to argue that the near-surficial deposits are less influenced by bio-degradation compared to the climate signal preserved within the biomarker records. Rontani et al. (2018) is often referred while only the ration of epi-brassicasterol and 24-methylenecholesterol is shown. Why don’t you analyse the autoxidation products of IP₂₅ in some of your samples? You have the co-authors to do this experiment. It would strengthen your dataset immensely and avoid mis-interpretation of your data.
• Clearly, from the discussion, core 109R is affected by biodegradable products. (from the Bra/24-Me) ratio. You may run some of your fractions again for potential prevalence of autoxidation products of IP₂₅ as well. Also, the gradual decline in brassicasterol concentration in all records could be interpreted as a result of diagenesis. Perhaps the application of PIP₂₅ is here rather speculative and taken the uncertainties of biodegradation into account, I would suggest to leave it out. You have a visually good correlation with declining sea cover from your observational data set. According to Rontani et al. (2018) this is your strongest argument against significant bio-degradational control.

Minor comments

• You mention X-ray fluorescence scanning and grain size analysis in the methods, but you hardly use these data for your interpretation. Consider showing the data actively or omit. X-ray fluorescence data can also provide you with information on diagenesis (redox boundaries).
• You may provide more details to your bulk analysis including d¹³C_org measurements, uncertainties, errors, standards etc.

General comments

I read the article from the standpoint of someone familiar with sea ice data sets, but not at all familiar with biomarkers for sea ice.  The authors state that the outcomes “support the reliability of biomarkers for sea-ice reconstruction in this region”. They present biomarker evidence that a “polynya-like feature” may have been forming in the westernmost Belgica Trough sometime before mid-century. This is an interesting finding, and it illustrates how these proxy data can be used when other sources fall short. The work contributes to understanding the history of sea ice area and extent off NE Greenland, and this, as the authors note, contributes to understanding the dynamics of two important glaciers that are buttressed by sea ice. 

The historical sea ice data set that biomarker data are compared with (Walsh et al., 2019) has large uncertainties, but I think it is used appropriately here. The authors are not correlating sea ice concentration percent values from the historical data with bioindicator values, rather, they are considering only relative amounts of sea ice cover and adding strength to their interpretation with change-point analysis.

Overall, the paper is well constructed and well written.  Terminology needs clarifying in places, if only to help cross-discipline readers. I’ve called out those places below.  

Specific comments

In the Introduction, it would be helpful to give a few descriptive words when IP25 and PIP25 first appear. This would be a kindness to those of us who know nothing of biological proxies for sea ice but want to learn how they can be used along with the satellite and other observational records we’re familiar with.

Ln 28.  Please define sea-ice cover here.  “Sea-ice cover” can be confused with sea ice extent, when what I think you mean is sea ice concentration or area.  (NSIDC has a short piece on "What is the difference between sea ice area and extent?", by the way.) “Sea-ice cover” is fine to use if how it is being used is made clear.

Beginning line 59 is: “Northeast Greenland is an area characterised by several sea ice types and features; it is thus a region of interest to understand the impact of climate changes on sea-ice extent. These features include land-fast sea ice (hereafter ‘fast ice’), seasonal sea ice and the Northeast Water (NEW) polynya.”

I’d like to better understand how the authors are using “seasonal sea ice”.   Usually, the term refers to broad expanses of ice that form in the winter months but are ice free in the summer; that is, the region between the ice edge in winter and the ice edge in summer.  The Belgica Bank area has not typically experienced this type of seasonal ice.  Looking quickly at the monthly extents in passive microwave satellite data, August and September of 2021 are the only times I see the sea ice extent retreat north of the Belgica Bank, although ice retreats well to the west in 2017.  

In typical usage, “seasonal” means ice is there in the winter but not in the summer.   But ice off the NE coast of Greenland is always there (except, notably, in summer 2021).   If the authors are thinking of seasonal ice as ice cover interrupted over time by polynyas, or just by variable areas of open water between pack ice floes moving south as shown in Figure 1, I recommend using a different term.

Ln 57. It’s not necessary to include this but I want to note that the Divine and Dick data are available at NSIDC:

D.V. Divine, C. Dick. March through August Ice Edge Positions in the Nordic Seas, 1750-2002, Version 1  NSIDC: National Snow and Ice Data Center, Boulder, Colorado USA (2007), 10.7265/N59884X1

Around Line 90, suggest you reference Fig 1(b) in the same sentence that first mentions the two marine-terminating glaciers.      Why is one glacier labeled NG when first introduced, and labeled “79NG” on the figure and in the text in later mentions?

Ln 126-129 Curious as to why this X-ray fluorescence step was not carried out for the other cores. Consider adding a sentence as to why.  Also for the grain size analysis step. Were these steps done just to check the match between Rumohr core and gravity core results for roughly the same location?

Ln 209. Please reference Fig. 7, where the sea ice cover (a.k.a. concentration in this instance, for Walsh et al.)  data are used.

Section 4.5 beginning Ln 293 on “Sea ice cover observational record”:

Given the uncertainty in the Walsh et al. record, it might be good to run the change detect routines for Aug and Oct just to see if there is a material change in results, although perhaps this is unnecessary given that 5-yr running means are used.   

Paragraph beginning Ln 390:

Please rewrite this sentence: “The positive correlations between IP25 and brassicasterol, and IP25 and dinosterol at all three sites (Fig. 5) can best be explained by the fact that under more extreme sea-ice conditions, both biomarkers show low values but with decreasing sea-ice, indicating more open-water and ice-edge conditions. “

Here is what I think is meant, but I am not sure about it:

“…can best be explained by the fact that when sea ice is preset more of the time, both biomarkers show low values. When open water conditions prevail, because the concentration of sea ice is low, or the ice edge moves shoreward of the location, both biomarkers show higher values. “

Ln 396. Suggest referring to Fig 4 here.

Ln 415. Consider replacing “seasonal sea ice has”  with “areas of open water have” in this sentence:   “The presence of IP25 in most of the samples in 90R and 109R suggests that seasonal sea ice has been present for the last ∼120 years in the coastal and mid part of the Belgica Trough.

Seasonal sea ice generally refers to broad expanses where ice forms in the winter and melts or moves out in the summer, “an area of ocean that extends from the permanent ice zone to the boundary where winter sea ice extent is at a maximum; here, sea ice is present only part of the year; this zone primarily consists of first-year ice.” (from the NSIDC glossary). Here, I believe you’re referring to what biomarkers are indicating could be a fairly regular occurrence of polynyas in an area that is more often thought of as ice-covered.

Ln 419.  Same comment as for ln 415, although in this sentence, could you replace “absence of seasonal ice” with “presence of sea ice” or “absence of periods of open water”?

The sea ice edge in this region, as defined using satellite passive microwave data, retreated north of the Belgica Bank area in 2021 (see https://nsidc.org/arcticseaicenews/2021/09/ )  but I believe that may have been only time that has happened in the satellite record.

Technical corrections

In Table 1, the longitude in the last row is missing a minus sign.

Ln 375  There is a missing “are”.