| 02 Jul 2026
Spatial and temporal patterns of fracture in confined landfast sea ice
Abstract. Fractures in sea ice form when internal ice stresses exceed ice strength. The locations of recurring fractures in landfast sea ice are well known to Arctic Indigenous communities, but recent observations suggest changes in their timing and frequency driven by changes in the Arctic climate. Fractures affect sea ice stability, travel safety, and ocean-atmosphere exchanges, so understanding why, when, and where they occur is increasingly important with climate change. To improve predictions of fracturing and support safe travel, we analyze fracture patterns in Admiralty Inlet, Nunavut, Canada, a semi-enclosed waterway. Using satellite imagery from 2018 to 2023 and guided by Inuit Knowledge of recurring fracture zones, we examine how geographic, atmospheric, and oceanographic factors shape fracture behaviour. This study provides the first assessment of fracturing in a semi-enclosed landfast ice environment in the Canadian Arctic Archipelago, extending previous work focused on open-coast environments. We find that seasonal changes in ice properties such as ice thickness, temperature, and porosity largely control fracture timing. Our results further indicate that episodic high-stress events such as wind storms are not primary drivers; instead, stable geographic features such as points of land concentrate stress and govern fracture locations. These findings highlight the need for in-situ measurements of ice strength, under-ice currents, and sea surface height, as well as higher temporal resolution fracture monitoring. By integrating remote sensing, environmental datasets, and Inuit knowledge, this work advances understanding of landfast ice stability and improves predictive capabilities relevant to both community travel safety and Arctic climate modelling.
General comments
The manuscript Spatial and temporal patterns of fracture in confined landfast sea ice by Loewen et al. investigates fracture patterns in the Admiralty Inlet of the Canadian Arctic Archipelago. The manuscript builds on a fracture dataset developed in the authors’ earlier study. This manuscript summarizes the spatial and temporal aspects of the fractures, and examines potential drivers of these. Local knowledge is incorporated with scientific methods, the study has clear motivation and user groups that can benefit from the results. In these respects, the manuscript is an excellent continuation and application of the previously developed dataset. There is novelty both in the focus on a confined landfast ice environment and in the integration of local knowledge into the analysis. The methodological choices are generally well justified, and limitations are appropriately discussed.
However, the manuscript could use a bit of condensing, and I have several minor comments and suggestions that are detailed below. Overall, I consider the manuscript a valuable contribution that improves our understanding of the spatial and temporal behaviour of fractures in confined landfast sea ice.
Specific comments
The study is partly motivated by the needs of the local people, and Inuit knowledge is heavily incorporated in the early sections. It would make sense to include a concluding sentence in the conclusions as well.
Section 3 effectively includes the local sea ice terminology, but adds cognitive load to the reader. To ease reading, consider repeating the English translation in parentheses more often, even if it feels a bit repetitive, especially when the word has been introduced a paragraph or more ago. This would be particularly helpful in Section 7.4.
In Section 5.2.4, were all the manually identified points identified just by looking at Sentinel-2 imagery, or was additional data used, particularly for the river mouths?
How representative are the years in 1959 to 1966 and 1970 of modern sea ice thickness conditions? At line 222 it is stated that the ice is thinning in the Admiralty Inlet. Are there Canadian Ice Service thickness data for other months? If so, you could compare those with the SmartQAMUTIK data (e.g. for January), to assess how applicable the old data is of the current conditions in this location.
In Section 6.1.4, what could explain the formation of east shore fractures in both early and late season, while west shore-originating fractures form mainly in late spring? Related to this, Section 6.1.5 on fracture growth appears somewhat qualitative (which the authors acknowledge). Given the apparent uncertainty in tracking the evolution of individual fractures, the authors could consider merging this section with another, and making it shorter. Table 3 might be more appropriate for the supplementary material, if retained. In addition, the table presents precise percentages despite the apparent uncertainties of the method. E.g. fractures with initial category short (east) end to the category short (west) and vice versa, highlighting potential ambiguity in how fracture continuity and growth direction were determined.
The Discussion section nicely addresses the applicability and limitations of the study!
Fracture widening is an important factor for traveling on ice. However, since the methods/data in this study do not address it, does it deserve its own section (7.5)? Additionally, Figure 13 seems redundant in this context. Could this topic be saved for a possible future paper on fracture widening?
I expected some discussion of the floe edge and variation/changes in fractures near it, as this was raised as a major concern (lines 50-59). While one cannot address everything in a single paper, would it make sense to include the ranges of floe edges (e.g. in Figure 5) and add a sentence or two on the distance of the first fractures from the floe edge, or whatever aspect the authors find makes the most sense? Alternatively, this could be left for another study, in which case its importance should be less emphasized in the introduction.
Technical corrections
L72-74: Are all the points listed in these lines covered by the manuscript? The paper delivers well on the recurring-pattern and process-understanding components, but less convincingly on the deviations and future climate sensitivity. It may be worth rephrasing these statements to better reflect the study.
L186: me who?
L260: Since the airport is mentioned, even though its data are not used in the analysis, its location (or that of Adams Sound) could be pointed from the map of Figure 2. E.g. referring to 14 in the text.
L265-267: Is the citation correct? Currently the DOI points to an analysis and forecast product.
L271: A verb missing?
L287: Every 100 m?
L460: Should it be Table 3?
L524-527: Nicely formulated!
Figure 2: What is the relevance of Iqaluit in the left map? The map on the right is rather busy, and many of the place names do not appear to be used in the text. Consider limiting the labels to locations that are mentioned in the manuscript. In addition, this figure could be referenced when mentioning Ikpiarjuk when it is first introduced (L50).
Figure 5: This figure (or Figure 4) seems like a strong candidate for a highlight figure. If so, would it be appropriate to note in the caption as well that in April-May the fracture occurrence frequency shown here might be too low due to methodological limitations discussed in the manuscript?
Figure 9b: This plot is quite difficult to read for months other than June and December, and it does not appear to provide substantially more information than 9a. Would it be sufficient to replace the points and error markers in 9a with boxplots and omit the plot with histograms?
Figure 10: Please check that the plots and the figure caption are consistent. For example, I do not see any blue in 10a. In addition, the variable v_b in 10b is v_v in the caption.
Figure 11: I suggest reducing the height of plot 11a to allow December in 11b to be displayed fully. Have you considered using split violins? That could make the figure to be slightly narrower and potentially easier to interpret.