the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 05 Sep 2025
A Climate Data Record of Sea Ice Age Using Lagrangian Advection of a Triangular Mesh
Abstract. We present LM-SIAge, a new Climate Data Record (CDR) of Arctic sea ice age spanning the period from 1991 to 2024. The dataset is based on a novel Lagrangian advection scheme applied to a triangular mesh, which conserves sea ice age fractions and reduces numerical diffusion compared to the previous Eulerian approach. LM-SIAge is derived from satellite observations of sea ice concentration and drift, and represents fractional age classes per grid cell. The record captures the spatial and temporal evolution of first- to sixth-year ice, including uncertainty estimates that account for both sea ice concentration and drift uncertainties.
We compare LM-SIAge with existing products from NSIDC and C3S, finding consistent large-scale trends—such as the decline of older ice—but also identifying systematic differences. Trend analysis confirms a significant reduction in sea ice age and a general increase in the area of first-year ice. Validation with IABP buoys indicates good consistency, with most discrepancies occurring near the ice edge.
The LM-SIAge dataset improves the observational basis for Arctic monitoring and contributes to the Global Climate Observing System (GCOS) Essential Climate Variables. It is publicly available and suitable for climate studies, model evaluation, and data assimilation.
- Preprint
(8872 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
- RC1: 'Comment on essd-2025-477', Anonymous Referee #1, 07 Oct 2025
-
RC2: 'Comment on essd-2025-477', Anonymous Referee #2, 14 Oct 2025
This study by Korosov et al. aims to present a new climate data record of Arctic sea ice age, LM-SIAge, developing a Lagrangian advection method on a triangular grid to minimize numerical diffusion compared to a grid-based method. On a daily basis, covering a wide range of years with a new approach, this dataset would be positioned as an important contribution to the scientific community, and potential users can effectively use this dataset for Arctic weather and climate monitoring, model validation, as well as data assimilation. The manuscript is well written and organized, and the research topic is well-fitting within the journal's scope. I recommend a revision to address the specific comments below before the manuscript is accepted for publication
1. One thing that I was concerned about was a validation and uncertainty analysis of the constructed dataset. The authors used buoy position data to validate the age product. It is an acceptable approach. However, the used buoy trajectories cover only between 2002 and 2024. The product for the earlier period of 1991 – 2011 was not validated in this manuscript. The uncertainty inherent in this product depends on the quality of the passive microwave-based SICs. The uncertainty of SICs can vary with seasons. The seasonal uncertainties of the age product due to the SIC uncertainties should be addressed. In addition, the uncertainty section is difficult to read and understand. I recommend reformulating this section.
2. In methodology, the authors chose certain parameters, e.g., mesh element size thresholds, angle, and element area, which seem to be used without any justification. It would be better if the authors provide how sensitive the results are changed due to these parameters or references to support these values.
3. Are there any possibilities of ridging or deformation processes with different age categories affecting the proposed algorithm?
4. It is very fair to provide Figure 15 for the NSIDC product as well.
5. In supplementary videos, I found that somewhat permanent multiyear ice exists in the coastline of the Kara Sea and Islands between the Kara and Laptev Seas. Is this correct?Citation: https://doi.org/10.5194/essd-2025-477-RC2 -
RC3: 'Comment on essd-2025-477', Anonymous Referee #3, 15 Oct 2025
This manuscript, “A Climate Data Record of Sea Ice Age Using Lagrangian Advection of a Triangular Mesh” by Korosov et al., is well written and scientifically sound. It presents a novel and carefully implemented approach to estimate sea-ice age with improved accuracy. Sea-ice age is a key variable for understanding Arctic sea-ice dynamics and thermodynamics, and the development of an improved methodology for its derivation will be welcomed by many researchers in the cryospheric community.
The authors introduce a Lagrangian-tracking framework based on triangular meshes to reduce tracking errors in sea-ice age estimation. The approach is clearly described and appears to be robust and efficient. The results demonstrate that this method substantially improves the spatial and temporal consistency of sea-ice age products.
Although direct validation of sea-ice age is inherently challenging due to the scarcity of observational references, the authors have utilized available datasets effectively. The validation efforts are appropriate and sufficient within the limitations of current observations.
Overall, the manuscript represents a valuable contribution to the field. I recommend publication after minor revisions.
Specific CommentsSection 3.1 – Representation of Age Evolution
The computation of sea-ice age change appears to assume that all ice categories evolve similarly in time. In reality, when sea-ice concentration decreases, younger ice tends to melt more rapidly than older ice. It would be helpful if the authors could clarify whether this differential melting behavior is accounted for in their formulation, or discuss its potential implications for the resulting age distribution.
Clarity of Method Description (Figures 5 and 6)
Since the triangular-mesh approach and its associated remeshing procedure may be unfamiliar to many readers, the description of these processes could be improved for clarity. Figures 5 and 6 are central to understanding the proposed algorithm, but both could be made more legible and intuitive.
In Figure 5, particularly in the left and right examples, it is difficult to visually identify what has changed before and after the remeshing process. The green lines representing the remeshed state are also hard to distinguish.
In Figure 6, the blue and black lines in the left panel are not easily distinguishable. Enhancing the color contrast or line thickness would improve readability.
Adding brief explanatory captions to Figures 5 and 6 that summarize what each step represents (e.g., “before remeshing,” “after remeshing,” “regularized mesh”) would help non-specialist readers.Citation: https://doi.org/10.5194/essd-2025-477-RC3 -
EC1: 'Comment on essd-2025-477', Clare Eayrs, 23 Oct 2025
Dear Authors,
The reviewers agree that your dataset makes a valuable contribution and has strong potential for publication, but there are some substantive issues that must be addressed. I invite you to continue to the next round. At ESSD, this is responding to review comments without updating the manuscript. If responses are deemed reasonable, then revisions. Please take into account the reviewers' comments and take particular care with the following items:
Uncertainty quantification and presentation
- All reviewers found the section on uncertainty difficult to follow. Please reformulate this section and consider adding a flowchart or schematic that illustrates the complete sequence of uncertainty estimation. Please define exactly what is meant by “sea ice age variable” and explain how uncertainties are calculated and combined.
- If the uncertainty varies spatially and temporally, please provide the uncertainty as a separate data file that matches the spatial and temporal dimensions of your primary product, and document how users should apply the uncertainty information.
- Please clarify whether the uncertainties shown in figures are absolute or relative values, and whether the “total uncertainty” represents the sum or the propagated metric. Please also address Reviewer 1’s query regarding capping/conditioning of the uncertainty.
Physical assumptions in the algorithm
- Reviewers 1 and 2 both highlighted that your scheme assumes equal scaling of fractions of different age classes. Please discuss the limitations of the mapping approach during convergence/ridging and how the assumption of equal scaling fractions of different age classes could influence the product and comparisons with other datasets. Reviewer 3 also asked whether your formulation accounts for the fact that younger ice tends to melt more rapidly when the concentration decreases. Please clarify whether this differential melting is represented, and if not, what the likely effect is on the age distribution.
Justification for parameter choices
- Please include a justification for the numerical parameters in the mesh set-up and either add references or a brief discussion on the sensitivity of these parameters to support these values.
Figures
- Please add labels to colorbars and address the specific review comments.
Please specify the DOIs for the OSISAF datasets that the user should download in the README file in the Zenodo repository. I noticed you recommend the Sea Ice Concentration Climate Data Record Release 3 (OSI-450-a), but this has been superseded by version 3.1 (OSI-450-a1). I strongly encourage the authors to update the sea ice age product to base it on the latest sea ice concentration CDR, thereby making it more useful for downstream users. An update now would also help prevent confusion with future work based on newer data and likely increase the dataset's usage and citations. If this update is not feasible at this stage, it would be helpful to outline any plans for future updates.
Citation: https://doi.org/10.5194/essd-2025-477-EC1 - AC1: 'Response to comments on essd-2025-477', Anton Korosov, 24 Oct 2025
Data sets
Arctic Sea Ice Age Climate Data Record Version 2.1.1 Anton Korosov and Léo Edel https://doi.org/10.5281/zenodo.15773500
Model code and software
Sea ice age computation software, PySeaIceAge-2.1.1 Anton Korosov and Léo Edel https://doi.org/10.5281/zenodo.16881687
Video supplement
Animation of the Arctic Sea Ice Age Climate Data Record Version 2.1 Anton Korosov https://youtube.com/shorts/h5MTsKuT8ic
Explanatory animations for the Arctic Sea Ice Age Climate Data Record Version 2.1 Anton Korosov https://doi.org/10.5281/zenodo.16744295
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 1,069 | 58 | 27 | 1,154 | 22 | 24 |
- HTML: 1,069
- PDF: 58
- XML: 27
- Total: 1,154
- BibTeX: 22
- EndNote: 24
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Please find the review attached