Characterizing snow grain size and shape evolution across 20 scenarios of dry snow metamorphism

Dick, Oscar; Calonne, Neige; Hagenmuller, Pascal

doi:10.5194/egusphere-2026-4897

Preprints

https://doi.org/10.5194/egusphere-2026-4897

Preprints

31 Aug 2026

| 31 Aug 2026

Status: this preprint is open for discussion and under review for The Cryosphere (TC).

Characterizing snow grain size and shape evolution across 20 scenarios of dry snow metamorphism

Oscar Dick, Neige Calonne, and Pascal Hagenmuller

Abstract. Snow microstructure evolves continuously during dry snow metamorphism, governing many effective physical properties of snow. Accurately representing this evolution in snowpack models requires objective descriptors of grain size and grain shape together with reliable temporal evolution laws. However, the evolution of snow microstructure has not been systematically characterized over the full range of dry snow metamorphism conditions. Here, we investigate the temporal evolution of several grain-size and grain-shape descriptors during dry snow metamorphism using a unique 4D X-ray tomography dataset comprising 20 experiments covering a broad range of temperatures, temperature gradients, and initial snow types. We analyze specific surface area (SSA), optical and microwave grain sizes, correlation length, structural anisotropy, three curvature-based grain-shape descriptors, as well as the full chord length and mean curvature distributions. Our results reveal that SSA evolution is far more diverse than generally assumed, including transient increases under strong temperature gradients. Optical and microwave grain-size descriptors often differ in both absolute value and temporal evolution, suggesting that snowpack models should predict both rather than a single grain-size descriptor. Vertical structural anisotropy does not systematically develop under temperature gradients, contrary to the common assumption. Overall, the initial snow microstructure strongly controls grain-shape evolution and, to a lesser extent, grain-size evolution, while both are also influenced by temperature gradient and mean temperature. These results provide new experimental constraints for developing and evaluating snow microstructure representations and temporal evolution laws in snowpack models. Future snowpack models could benefit from predicting the full chord length and mean curvature distributions rather than only scalar descriptors.

Received: 13 Aug 2026 – Discussion started: 31 Aug 2026

Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims made in the text, published maps, institutional affiliations, or any other geographical representation in this paper. While Copernicus Publications makes every effort to include appropriate place names, the final responsibility lies with the authors. Views expressed in the text are those of the authors and do not necessarily reflect the views of the publisher.

Download & links

Oscar Dick, Neige Calonne, and Pascal Hagenmuller

Status: open (until 10 Nov 2026)

Post a comment Subscribe to comment alert

RC1: 'Comment on egusphere-2026-4897', Anonymous Referee #1, 25 Sep 2026 reply

General comments
The paper builds on a recent publication by the same main author (Earth Syst. Sci. Data, 18, 2875–2889, 2026 https://doi.org/10.5194/essd-18-2875-2026) that presented an extensive set of time-laps micro-CT images of 20 snow metamorphism experiments spanning a range of initial snow types, temperatures and temperature gradients. Here, several shape descriptors are introduced, computed, and their evolution over time is analyzed in function of the experimental parameters. The results are put into perspective with existing work and suggestions on how shape descriptors might be used in future snowpack models are made.
While subsets of shape descriptors have been presented for similar experiments before, the paper makes significant progress through its completeness, both regarding the experimental and the shape descriptor broadness. For the latter, the combination of scalar descriptors and statistical distributions seems particularly promising, albeit the approach is not completely new.
The paper is generally well written and the approaches and results are nicely presented. Clarity could be improved for a few statements and the discussion, namely when comparing the results with other publications, could benefit by considering the specific remarks provided in the first part of attached pdf-file. For minor suggestions and technical corrections, see also the annotated paper preview in the second part of the pdf.
I suggest the paper be revised accordingly before accepting it for publication.

Reply

Citation: https://doi.org/10.5194/egusphere-2026-4897-RC1
RC2: 'Comment on egusphere-2026-4897', Takafumi katsushima, 28 Oct 2026 reply

The major strength of this manuscript lies in the unique and extensive 4D X-ray tomography dataset comprising 20 dry-snow metamorphism scenarios. Previous studies have typically investigated individual grain-size or grain-shape descriptors separately and often under a limited range of experimental conditions. By contrast, this study provides a systematic comparison of numerous descriptors within a single, coherent dataset, substantially extending the current understanding of snow microstructural evolution. Importantly, the value of this study does not stem from proposing a new descriptor, but rather from providing an unprecedented framework for comparing existing descriptors and evaluating their respective strengths and limitations using a common experimental dataset. In this regard, I would like to strongly commend the authors for the excellent work they have done.
In my view, the novelty of this work goes beyond the presentation of a large experimental dataset. The manuscript fundamentally revisits the questions of what should be considered as "snow grain size" and "snow grain shape" from an observational perspective. The results clearly demonstrate that different descriptors capture different aspects of snow microstructure and may evolve in contrasting ways during dry-snow metamorphism. This has important implications for both the interpretation of experimental observations and the development of future snowpack models.
I particularly appreciate the authors' effort to compare traditional scalar descriptors with the full chord-length and mean-curvature distributions. This approach provides new insights into the multi-scale nature of snow microstructure and challenges the common assumption that a single grain-size descriptor is sufficient to characterize snow evolution.
Although the observed metamorphic behaviors are highly complex, the authors provide clear interpretations and well-supported discussions throughout the manuscript. The paper is well organized and written in a clear and concise manner, making a complex topic accessible to readers. I found the manuscript both scientifically rigorous and enjoyable to read.
Overall, I consider this study to be a significant contribution to snow microstructure research and believe that it will be of considerable interest to the readership of The Cryosphere. Subject to the comments provided below, I strongly support its publication.

Major comment
The authors convincingly demonstrate that the initial snow microstructure exerts a strong control on the evolution of grain-shape descriptors. However, my main concern is that the term “initial snow microstructure” is treated as a single explanatory variable throughout the manuscript, while its underlying components are not explicitly disentangled.
In this study, differences in initial microstructure are represented through different snow types. However, these snow types differ not only in grain geometry, but also in several other fundamental properties, including bulk density, porosity, coordination number, and connectivity of the ice matrix. As a result, it remains unclear which specific aspects of the initial microstructure are primarily responsible for the observed differences in the subsequent evolution of grain size and grain shape descriptors.
I do not consider this a limitation of the experimental dataset itself. Rather, I believe that the manuscript would benefit from a more explicit discussion of this issue. In particular, it would be valuable to clarify to what extent the observed effects can be attributed to grain morphology versus other snow physical characteristics such as density or ice-volume fraction. A discussion of these potentially confounding factors would help readers better interpret the conclusion that "initial snow microstructure strongly controls grain-shape evolution" and would provide useful directions for future research.

Minor comments
1. The specific genus is presented as one of the analyzed descriptors, but its implications are only briefly discussed compared with the other metrics. Since this quantity was measured throughout the experiments, I believe it deserves further consideration in the Discussion section. Even if the authors conclude that specific genus is less useful for describing dry-snow metamorphism under the investigated conditions, it would be valuable to explain why this is the case and what information this descriptor may or may not provide.
2. The main conclusions of the manuscript rely on the evolution of SSA, chord-length distributions, and curvature-based descriptors, all of which may be sensitive to image resolution and smoothing procedures. I suggest that the authors provide a brief discussion regarding the suitability of the selected voxel size and image-processing workflow. Furthermore, it would be helpful to clarify how robust the reported results are to changes in image resolution and whether similar trends would be expected at finer or coarser resolutions. Such a discussion would help readers better assess the generality of the conclusions.
3. The authors suggest that future snowpack models may benefit from predicting full chord-length and mean-curvature distributions. I strongly agree with this perspective. If such an approach becomes feasible, it would represent a major step forward in snowpack modeling. Could the authors comment on how the evolution of these distributions could be represented in practical snowpack models?

Very minor comment
I suggest using a common mean-curvature range for the blue, red, and green distributions in Figure 8. The current scaling makes it difficult to directly compare.

Reply

Citation: https://doi.org/10.5194/egusphere-2026-4897-RC2

Oscar Dick, Neige Calonne, and Pascal Hagenmuller

Viewed

Total article views: 525 (including HTML, PDF, and XML)

HTML	PDF	XML	Total	BibTeX	EndNote
278	197	50	525	61	55

HTML: 278
PDF: 197
XML: 50
Total: 525
BibTeX: 61
EndNote: 55

Views and downloads (calculated since 31 Aug 2026)

Month	HTML	PDF	XML	Total
Aug 2026	34	9	3	46
Sep 2026	150	101	38	289
Oct 2026	94	87	9	190

Cumulative views and downloads (calculated since 31 Aug 2026)

Month	HTML	PDF	XML	Total
Aug 2026	34	9	3	46
Sep 2026	150	101	38	289
Oct 2026	94	87	9	190

Viewed (geographical distribution)

Total article views: 525 (including HTML, PDF, and XML) Thereof 525 with geography defined and 0 with unknown origin.

Country	#	Views	%

Latest update: 01 Nov 2026

Short summary

This work aims to pave the way towards an objective description of snow microstructure in snowpack models. We investigate the temporal evolution of several grain-size and grain-shape descriptors under a large range of dry snow metamorphism scenarios. Our results reveal that the evolution of several geometrical descriptors is more diverse than generally assumed, and that future models could benefit from representing multiple grain-size and shape descriptors.


Total:	0
HTML:	0
PDF:	0
XML:	0