| 20 Oct 2025
Intra-annual grounding line migration and retreat: insights from high-resolution satellite and in-situ observations over Milne Glacier in the Canadian High Arctic
Abstract. Changes at the boundary where marine terminating glaciers transition from grounded to floating, known as the grounding line, are critical for glacier stability and ice discharge prediction. We explore changes in the ice flexure zone (FZ), which spans from the inland limit of tidal flexure, known as the hinge line, across the actual grounding line, to where the ice tongue becomes freely floating. FZ movements on Milne Glacier, Nunavut, Canada from 2023 to 2024 were measured at a spatial resolution of 10 m using 4-day RADARSAT Constellation Mission (RCM) repeat acquisitions from four different orbits. The Double Difference Interferometric Synthetic Aperture Radar (DDInSAR) technique was employed to delineate 116 FZ positions, which were then examined with respect to tides, glacier velocity, surface and bed topography. Tides modulated by atmospheric pressure were the major driver of changes in the FZ with the correlation coefficient between DDInSAR and sea surface height displacements ranging from 0.81 to 0.94 for different RCM orbits. The lowest error (<0.1 cm in the SAR line-of-sight) occurred during the ‘stable regime’, when changes in glacier velocity were minimal, with ascending RCM tracks at incidence angles below 30°. During the ‘unstable regime’, the glacier velocity became more variable, and the error increased by up to 4 times. In summer, ice in the FZ flowed ~2.7-fold faster than the grounded section upglacier. We discovered a new ~2.4 km grounding line retreat associated with a pinning point at the eastern margin of the glacier. This rapid change is likely due to water intrusion and subglacial melt near the pinning point and shows how these bed features alter the grounding line short-term migration and its long-term retreat.
Review for Intra-annual grounding line migration and retreat: insights from high resolution satellite and in-situ observations over Milne Glacier in the Canadian High Arctic
The article “Intra-annual grounding line migration and retreat: insights from high resolution satellite and in-situ observations over Milne Glacier in the Canadian High Arctic” by Antropova et al. presents a well organized narrative of both methodologies and and local results exploring the intra-annual changes of the flexure zone of a marine termination glacier in the Canadian High Arctic. Ground and airborne data are combined with a dense time series of SAR images to unpack what is happening not just at a high resolution throughout a year but also looking along three longitudinal transects that exhibit variability that the authors do a nice job of unpacking and quantifying. Finally, they put this work into a wider context by looking at the bed topography for where future pinning points might be and how they might affect subglacial melt.
Below are my main comments followed by in line minor comments.
The authors spend some time exploring the importance of including the inverse barometer effect (IBE), but I think this as been well established by Padman et al., 2003, Padman et al, 2008 and e.g. Chen et al., 2023. While quantifying and partitioning the relative terms is very useful, I’m not sure if IBE needs to be demonstrated as important against a simulation excluding it.
The authors use GEM to fill data gaps in the pressure record, this is somewhat of a footnote. I’m not sure if there are other studies that do this, but since the authors have a dense timeseries of supporting data it could be a useful result for the community to use only GEM data and see if the 15 x 15 km grid 9 km away can reproduce the same tidal model.
There seems to be some inconsistencies throughout with regard to FZ and GZ, both in the terms used and what was calculated and how. I would ask the authors to read through this carefully and make sure terms are consistent throughout and clearly derived if used later in the paper.
There is a use of stable and unstable which I found to be imperfect terms for an intra-annual time series. In Section 4.2 the authors relax this language and talk in terms of season. I think the stable/unstable delineation might be more easy to follow and possibly more accurate to label in terms of seasons. A possible title could be “Seasonal grounding line migration (I’m not sure if there is a long enough time series here to identify permanent retreat?). Also in the title is “high resolution” high temporal resolution? I think this is really the key factor that make the study unique.
The paper is largely a collection of substories, one of which is the notch which I had the hardest time following. Maybe there’s a figure that could help the reader along.
Finally, as an intra-annual study, I think a nice addition to the discussion section could be a recommendation to future researchers, who might be limited to a more sparse dataset, on how they might be confident or cautious comparing data from different instances in the seasonal cycle.
L14-15: The language isn’t quite correct here, either something like "Knowledge of changes at the boundary…is critical for…” or “Changes…the grounding line control glacier stability…”
L15: Is there a more succinct way to build directly off of the grounding line established in the first sentence to flexure zone rather than seemingly developing a new term and then at the end of the sentence connecting it back to grounding line? Within a system with a flexure zone, isn’t there not an “actual grounding line” location but rather a moving will occupy throughout an observed tidal maximum and minimum?
L20-L21: measured or modeled tides? Was atmospheric pressure implicit or explicit?
L22: 0.81, 0.94 m?
L22-25: Is the vertical displacement a key result of your project? I would think the key point to get across is sensitivity in z to determine a meaningful x,y. The following sentence is unnecessary detail for the an abstract in how it’s presented here. As it reads, here it should just be +/- error bounds for 0.81 and 0.94, but I would argue that these numbers aren’t so fundamental to the study and the sentence of interest here is z accuracy as it relates to x,y accuracy.
L25: Maybe you mean the downstream edge of the GZ, presumably there is a gradient somewhere in the upper boundary of the GZ.
L25: How should a reader interpret this result, is this abnormal? A change? Did this seasonal cycle drive the retreat in the following sentence?
L25: new meaning previously unreported or newly developed within some amount of time?
L26: Why grounding line here and not GZ?
L27: “...shows how bed features can alter the grounding…”
L31: Add citation(s)
L31: Establish this connection: Rising atmospheric and ocean temperatures can cause marine-terminating glaciers to speed up, which then leads to …
L32: Check formatting of IOCC ref. Is it 2021 or 2023?
L33, try to establish the location for the reader, something like, “In the Arctic, …”
L33: What was the change? Faster and smaller?
L38-39: Suggest: …exhibit dynamic behaviour at two key scales, ysub-day variability from tides and [insert some time scale, e.g. years to decadal] changes as a response to climate and glacier dynamics.
L41: maybe effective rather than efficient
L45: Yes, Fricker at al, uses grounding zone for flexure zone, Freer et al, in their Fig. 1 do differentiate their points, F, H as the flexure zone proxy but also F_max and F_min, the limits of the GL. I would encourage the authors to affirm the language flexure zone for what DInSAR measures, the ice grounding zone for the true max and min of tidal migration of the GL (ice used here to differentiate from an oceanographers definition of the grounding zone which I believe is slightly different from a glaciologist’s perspective), finally the grounding line referred to not a fixed location but moving both diurnally with tide and over years with glacier-wide and climate changes. I think the authors are mostly holding to this, but it’s not perfectly concrete. (your L48 and 50-54 are very good and clear). I think you could just define FZ and GZ, I’m not sure what span adds.
L47: This “however” is a little misplaced I think, make sure there is clear conceptual contrast.
L49, 58, Chen et al., 2023 is missing in References
L59 GZ spans for consistency.
L61: I think the objective is to understand the controls and responses of a system that happens to be deteriorating. We hope these observations will work in a phase of glacier growth also.
L65: Generally would say model results or inputs not model datasets
L64,66 if using FZ “span”, use throughout
L70: what does marked mean? Could just say acceleration and no acceleration for a and b. Is it acceleration and deceleration that makes an unstable regime?
L71: In 64-66 you establish intra-annual as a short term, is this still the same short term? Or do you mean 1d or 4d changes?
L93-95: Does Van Wychen et al, provide a mechanism for these velocity changes? Obviously velocity context is useful, but will the speed up slow down cycle come back into the story here?
L96-98: There isn’t a citation here, is this your result? Add citation or move to results.
L123: automated or manually delineated?
L129: It’s possible resampling the DEM to a lower resolution can help maintain coherence with the crevasse pattern in the interferograms, just a suggestion for the future.
L134: It’s not clear which dataset actually defined the three transect lines. I think figure 1 could benefit from an inset with the surface and bed airborne and in situ radar transects and the IceBridge flightlines.
L138-L138, Isn’t the distance from Bw1 to Bw2 a change over time, your migration term not the GZ?
L141: Was the unwrapped phase calibrated to the surrounding bedrock? What is a phase discontinuity? A phase jump? How was this corrected?
L148: change and to are. In this study are…
L174: The notch isn’t perfectly clear to the reader at this point in the article.
L180: It looks like the unstable period is the white block between sable-blue and unstable-red. And why isn’t all of the flat velocity before the beginning of the first blue block considered stable? It seems to me these are environmental classifications that are actually instrumental classifications, where data was available and where there was coherence. I would encourage the authors to say something like “winter observation block” and late-summer/fall observation block” otherwise this is very confusing for the reader. Layer L191 admits this.
L195: It is technically “not stable” because there is change in velocity, but I wouldn't call a predictable seasonal slowdown as “unstable”
L227: A height range converted to a spatial extent using ArcticDEM?
L246: Don’t you have bed geometry and ice thickness (L220), why not explore this rather than leave as speculative “likely due to a distinctive bed geometry and ice thickness”
L283: Can this or a plot like this include the pinning point(s)? It would be cool to see how the Eastern transect changes
L294: Maybe I missed it, what is the reference height SSH is based on?
L303: Table 3, It’s essentially a new fringe or two that you’re saying is a persistent feature and thus not error. From this table it seems possibly within error but if it’s spatially consistent this is maybe a better argument. Why no middle value along with SD and max min, and the first column says averaged across all observations, but really divided between sets of notch and no notch?
L318-319: Is the loss of correlation a suggestion to future scientists that acceleration/decelation lowers the confidence of DDInSAR measurements? Is there also something we’re learning about this particular glacier? It’s read like a result for the particular glacier but I think the key result is methodological.
L380-384: I’m not sure if I recall the authors giving their method to derive GZ from the data.
L385: East and West is reversed
L386: Did you calculate this flux or is it just assumed based on a roughly parallel bed profile to cross-sectional ice surface?
L393-401: I think the role of IBE is well established in Padman et al., 2003 (cited but missing in references), also Padman et al., 2008, and as you cite Chen et al., 2023. Is/was there an assumption that the tidal effect would be different for marine terminating glaciers in the Canadian Arctica?
L410-412: The grounding line migration of 800m wasn’t explained or reported earlier. This should be a result and I’m not sure I know how the grounding line analysis was done.
L420: Do you mean for future predictions of pinning points? Because for a general understanding of active pinning points we don’t need detailed information of bed geometry, just good DDInSAR.