| 15 Dec 2025
Sub-seasonal variability and multi-year trends in glacier area change and ice speed on the Antarctic Peninsula
Abstract. As the climate continuous to warm, glaciers on the Antarctic Peninsula (AP) are experiencing rapid dynamic changes, including accelerated rates of thinning, terminus retreat, and ice flow, all of which have significant implications for global sea level rise and climate feedback mechanisms. These glaciological changes unfold over varying timescales spanning from months to decades, necessitating temporally and spatially detailed monitoring. Here, we used sub-seasonal records of terminus area change (2013–2023) together with high-resolution satellite-derived ice surface velocity measurements (2014–2024) to investigate the evolution of 42 key outlet glaciers on the northern AP. We found that during the past decade, these glaciers underwent substantial but spatially heterogeneous changes in terminus position and ice flow velocity. Cumulative ice loss amounted to ~279 km², with 73 % of this loss occurring on the eastern side, particularly within the Larsen B embayment. By contrast, western glaciers showed smaller, more variable responses. Overall, 71 % of glaciers accelerated, though most eastern glaciers displayed slight trends of slowdown – except in the Larsen B region, where major calving events beginning in early 2022 were followed by drastic increases in velocity, with some glaciers more than doubling their flow speed. In addition, a seasonality analysis revealed widespread inter-annual variability: two-thirds of glaciers showed strong to very strong seasonal fluctuations in flow speed, and about half exhibited comparable signals in terminus change. Cross-correlation analysis further indicated that, for most glaciers, terminus area changes and ice velocity dynamics had only minimal influence on one another. These results highlight the spatially and temporally heterogeneous nature of glacier dynamics on the northern AP, suggesting that glacier change in this region is shaped by a combination of environmental drivers and glacier-specific factors operating on widely varying timescales. They underscore the need for detailed, high-resolution observations and continuous monitoring to improve the understanding of glacier evolution under ongoing climatic and environmental change.
In this manuscript, Leibrock et al. analyze time series of glacier area and velocity along the Antarctic Peninsula. The manuscript is well written and easy to read. The authors' rationale for each step of the analysis is explained clearly and the authors have nicely placed their results in the context of previous work via a thorough literature review. Despite its virtues, I do see opportunities for significant improvements to the manuscript.
MAIN COMMENTS:
After reading the paper, I am not sure what the main finding is. I know what the paper is *about*, and I think I understand the methods that were used in the analysis, but I could not confidently recall the main takeaway if someone asked me about the manuscript tonight at a cocktail party. To bring the main findings into clear focus, I recommend reducing the length of the abstract by about 50% and focusing on one main finding. Possibly include a secondary point or two if they're profound or give helpful context, but keep the messaging tight. The journal Nature provides a formula for writing an abstract that works remarkably well, and a similar approach might benefit the present manuscript (https://www.nature.com/documents/nature-summary-paragraph.pdf).
The Introduction section motivates paper, saying observation-based studies like this one are needed to understand ongoing changes in the AP, but the Discussion says that understanding any correlations or lack of correlations between velocity and glacier area is beyond the scope of this study, and the Conclusion states that future work is needed to understand the observations reported here. The glacier area and velocity data are both from other studies, and no new understanding have been derived from them, so I am wondering what exactly is new here.
Overall, I find it difficult to fully understand or believe in the conclusions, because I haven't been able to see where the conclusions come from. Readers are told about the data, we're told about the analysis, and we're told what the authors' conclusions are, but we're not *shown* much along the way. As a result, we're in the position of trusting the authors' interpretation of the data, rather than being able to see the data and come to the same conclusions on our own. To give readers an intuition for where the conclusions come from, consider adding a plot that shows the raw data, the processed data, and the natural grouping of the different glaciers. Consider showing all 42 time series of velocity and area, with the raw data (i.e., the image-pair level velocity values) in a light color, overlaid with the filtered monthly average time series in a heavier, darker color. And identify, perhaps with line color, the Category that each glacier is determined to be. Adapt these suggestions as necessary, but the main idea is to bring readers along and show us the major steps from raw data to final conclusions.
I find the seasonality strength metrics to be somewhat suspect. It's hard to be certain, because we are not shown most of the data, but in the four Area time series that are shown in Figure 3, I see no indication of seasonal variability whatsoever, yet if I am reading Figure 4 correctly, it indicates all of these glaciers have "moderate" seasonal strength. Is a residual from the once-in-a-decade calving events of 2022 being interpreted as a seasonal signal?
I don't think the simple correlation analysis is appropriate for this work, because it conflates many different potential seasonal signals with response to rare calving events. Seasonal velocity variability may be driven by seasonal calving, or by seasonal basal melt near the grounding line, or by surface melt that impacts basal hydrology, particularly at these northern latitudes. Longer term changes in velocity might be driven by major calving events or by gradual changes in driving stress due to surface elevation or SMB. Simply correlating two time series--one that accumulates over time and another that represents an instantaneous response to forcing--without consideration of different timescales or driving mechanisms, is unlikely to reveal any meaningful patterns. The manuscript points to Ultee et al., 2022 as justification for doing a similar analysis here, but an important difference is that Ultee et al. made a concerted effort to separate and understand the processes at play, while the present manuscript ends the analysis when no clear pattern is discovered in a simple correlation analysis.
The conclusions are framed around the authors' own classification scheme that categorizes glaciers as either accelerating, decelerating, advancing, retreating, fluctuating, or "change point". I understand the need to make sense of the world by clustering similar behaviors together, but these categories are somewhat squishy, they rely on arbitrary thresholds of statistical significance rather than incontrovertible properties of the glaciers, and without seeing the data, it's hard to know how distinctly clustered these categories really are, or how easily a glacier's identity might switch between categories with one calving event. If these categories truly represent distinct characteristics that might indicate common driving or governing mechanisms within each category, then I'd like to see evidence for that and a discussion of the physical processes that might be at play. If not, I'm not sure what value the categories provide.
MINOR COMMENTS:
L20: I think "inter-annual variability" should be "intra-annual variability"?
L23-26: The last two or three sentences of the abstract essentially say that nature is complicated and future work is needed to understand it. This sentiment makes the abstract fall flat and implies the present study does not offer any satisfying conclusions. Reword the end to come to a stronger conclusion about what this study DID find rather than what it didn't find, and delete the vague statements about the need for future work.
L49: Check those units. Davison et al. presented units of Mt yr^-2, not Mt yr^-1 in their abstract. In my opinion, presenting units of acceleration in the Davison paper wasn't a great choice on their part because interpreting changes in acceleration rates isn't very intuitive, and it's easy to mistake for simple flow rates (as has happened here!). I recommend rewording in a way that doesn't repeat Davison's units of acceleration.
L56: The acronym EO is defined here, but never used.
L87-90: I think this entire paragraph can be deleted because the sentence that begins "This study focuses on 42 marine-terminating glaciers" is very similar to the sentence on Line 64 reads "In this study, we...investigate...42 outlet glaciers" and the "Of these glaciers..." sentence doesn't lead to anywhere.
Figure 1:
* The caption says units are km yr^-1, but the colorbar says m yr.
* It is common to show ice velocity in a log color scale. I'm not suggesting it here, but only noting a potential for ambiguity that could be mitigated by including more tick labels on the colorbar to show linear increments.
Fig 1 caption, L127, and L143: Reword to remove references to Slater (In Prep). If any details in the unreleased manuscript are necessary to understand the present study, describe those details here.
Table 1: It would be helpful to actually see all 42 time series, either colored or somehow clustered by Category. Because the table provides a suitable reference, but doesn't provide intuition for what shapes of time series to imagine when we think of each category.
L153-163: The description of the processing would be easier to understand if we could see what is being described here. Point readers to Figure 3 here (and reorder the figures as needed) so we can see the 3x3 km boxes, etc.
L171: "Seasonal" and "Trend" don't need to be capitalized.
L183-197: Show, don't tell. At a minimum, point readers to the time series shown in Figure 3, but to understand where the conclusions come from, it would also help to see the normalized cross correlation analysis.
L206: "initially advanced" is not very descriptive. Upon first read, I thought it meant a rapid acceleration that served as a precursor to significant retreat, but after looking at Figure 3, I think it means a steady advance rate over a decade?
L256 and Figure 4c: Why doesn't the correlation coefficient include the sign of the correlation? The strength of the correlation is meaningless without knowing the sign.
L289: Typo "norther" should be "northern".
L379: This may be a term I am not familiar with, but I'm not sure what is meant by a "consolidated" ice front.
L330: There is no need to define the acronym SIE, as it's only used in this sentence.
L340: The new findings don't necessarily "contrast with" Seehaus et al. Perhaps "update" or a similar sentiment would be more precise?
L372: Same typo "norther" should be "northern".
L379: I'm not sure what it means to "consolidate" a glacier terminus.
L642: The Wallis citation is incomplete. Needs a doi, journal name, or other identifiers.