Laboratory for Satellite Altimetry / Sea Ice and Polar Dynamics Science Team
Arctic Sea Ice Roughness and Deformation from ICESat-2 Laser Altimetry
Pressure ridges, formed by convergence between ice floes, are a dominant feature of sea ice topography in the Arctic (Figure 1). Ridging increases ice thickness, and surface roughness modifies the air-ice-ocean momentum flux in polar waters, impacting air flow across the ice and ocean circulation below it. The presence of pressure ridges in the ice pack impedes travel across, through and beneath the ice. Knowledge of sea ice roughness and deformation is vital for understanding the rapid changes underway as the ice cover transitions from older, thicker ice to younger, thinner ice that is less resilient to melt-out in summer.
Figure 1. Left: Aerial view of a sea ice pressure ridge in Arctic sea ice (credit: Sinead L. Farrell). Middle: Convergence of ice blocks during pressure ridge formation in the Arctic Ocean (credit: Sinead L. Farrell). Right: Schematic showing the role that winds and ocean currents play in the formation of sea ice features including ridge sails, keels and leads through convergence and divergence, respectively. (credit: https://meereisportal.de).
The ICESat-2 mission (Markus et al., 2017), launched in 2018, carries the Advanced Topographic Laser Altimeter System (ATLAS) which can capture sea ice deformation features in high-resolution and at the pan-Arctic scale (Farrell et al., 2020). We apply the University of Maryland-Ridge Detection Algorithm (UMD-RDA), a bespoke sea ice surface retracker developed by Duncan & Farrell (2022), to high-resolution (0.7 m) surface height measurements in the level 2 ICESat-2 ATL03 Global Geolocated Photon data product (Neumann et al., 2025), as illustrated in Figure 2. This allows us to derive sea ice geophysical variables that describe ice surface topography (Table 1) including surface roughness and ridging intensity for 1 km along-track segments, as well as pressure ridge length, height and width of individual pressure ridge sails, and the distance between sequential sails (Figure 3).
Figure 2. A 5 km-long segment of ICESat-2 (IS2) surface height measurements across deformed Arctic sea ice acquired north of the Canadian Arctic Archipelago on April 22, 2019 showing the level 2 ATL03 global geolocated photon height data product (black dots), level 3 ATL07 sea ice height data product (red line) and the level 3 University of Maryland-Ridge Detection Algorithm (UMD-RDA sea ice data product) sea ice height data product (blue line). Credit: Duncan & Farrell (2022).
| Variable | Definition |
|---|---|
| Local Level Ice Surface (LIS) | Modal surface elevation (m) of sea ice plus any overlying snow, calculated within 25 km-long, non-overlapping along-track segments (calculated during data generation but not currently reported in data product) |
| Surface Roughness (surface_roughness_1km) |
Standard deviation of surface elevations (m) within 1 km-long, non-overlapping along-track segments |
| Sail Apex | Local maximum that is equal to or greater than a specified threshold (here, 0.2m or 0.6m), relative to the LIS, that satisfies the Rayleigh criterion |
| Sail Height (sail_H) |
Height (m) of a sail apex, that is equal to or above a specified threshold (here, 0.2m or 0.6m), relative to the LIS, that satisfies the Rayleigh criterion |
| Maximum Sail Height (max_sail_H_1km) |
Maximum sail height (m) of largest sail apex in each 1 km-long, non-overlapping along-track segment |
| Sail Spacing (sail_D) |
Distance (m) between consecutive sail apices (local maxima) that satisfy the Rayleigh criterion |
| Ridging Intensity (ridging_I_1km) |
Reported for each 1 km-long,
non-overlapping along-track segment. Unitless. |
| Sail Width (sail_W) |
Width (m) of sail at 50% of sail height |
| Ridge Length (ridge_L) |
Along-track distance (m) between the points of intersection of the minima on either side of a sail apex and the LIS (can encompass more than one sail apex) |
Figure 3. Profile plots of sea ice elevation (black line) obtained with the UMD-RDA in ice-covered polar waters on April 22, 2019, annotated to illustrate the deformation variables (green dashed lines) provided in the data product. Elevation is relative to the DTU21 mean sea surface (MSS) model (Andersen et al., 2023). (a) Heavily deformed sea ice (indicated by "A", top-left inset map). The local level ice surface (LIS, gray dashed line) and the 0.2 m (T20, blue dashed line) and 0.6 m (T60, red dashed line) thresholds are computed to aid in detection of pressure ridge sails (red and blue dots) and to calculate other ice deformation parameters. (b) Same as in (a) but for a region of moderately deformed sea ice (location "B", top-left inset map). Credit: Duncan & Farrell (2026).
Sea ice elevation in the UMD-RDA data product is referenced to the DTU21 MSS (Andersen et al., 2023), and atmospheric and tidal corrections have been applied. Cloud masking and data filtering follows that provided in the level 3 ICESat-2 ATL07 Sea Ice Surface Height data product (Kwok et al., 2025). Along-track deformation variables are calculated for all three ICESat-2 strong beams and used as input to create monthly data files at 3.125 km, 10 km and 25 km resolution, provided on the Equal-Area Scalable Earth version 2 (EASE2) grid (Brodzik et al., 2012) using the Northern Hemisphere Lambert Azimuthal projection. Data are bounded by the region of the Arctic Ocean where mean sea ice concentration for each respective month is ≥ 15%. Three parameters are included with each deformation variable: count, the deformation variable itself, and the interpolated deformation variable. Figure 4 shows a side-by-side comparison of uninterpolated (Figure 4a) and interpolated (Figure 4b) sea ice surface roughness in April 2025 at 10 km resolution.
Figure 4. (a) Sea ice surface roughness in April 2025, gridded at 10 km resolution. (b) Interpolated sea ice surface roughness in April 2025, gridded at 10 km resolution. Credit: Duncan & Farrell (2026).
Further Reading:
For further details about the UMD-RDA data product and the computation of sea ice deformation variables, please refer to this UMD Sea Ice Roughness and Deformation User Guide, (PDF, 19.86 MB).
Acknowledgments:
All those interested in using the UMD-RDA sea ice data product are asked to cite one of the following data products, depending on the resolution used:
Duncan, K., & Farrell, S. L (2026). Gridded 3.125 km Arctic sea ice deformation data derived from the ICESat-2 University of Maryland-Ridge Detection Algorithm [dataset]. PANGAEA, DOI: 10.1594/PANGAEA.990275
Duncan, K., & Farrell, S. L (2026). Gridded 10 km Arctic sea ice deformation data derived from the ICESat-2 University of Maryland-Ridge Detection Algorithm [dataset]. PANGAEA, DOI: 10.1594/PANGAEA.990265
Duncan, K., & Farrell, S. L (2026). Gridded 25 km Arctic sea ice deformation data derived from the ICESat-2 University of Maryland-Ridge Detection Algorithm [dataset]. PANGAEA, DOI: 10.1594/PANGAEA.990274
For methodology, users may also refer to the following publication:
Duncan, K., & Farrell, S. L. (2022). Determining variability in Arctic sea ice pressure ridge topography with ICESat-2. Geophysical Research Letters, 49, e2022GL100272. DOI: 10.1029/2022GL100272.
This work was supported by NOAA Grant NA24NESX432C0001 (Cooperative Institute for Satellite Earth System Studies – CISESS) at the University of Maryland/ESSIC and through the NASA Cryosphere Program.
Data Product and Data Access:
UMD-RDA sea ice data product -
Files are in NetCDF format and provided in monthly grids at 3.125 km, 10 km and 25 km resolution.
As of April 2026, data are available for the period October 2018 – September 2025.
The data are permanently archived at pangaea.de, as follows:
- 3.125 km: DOI: 10.1594/PANGAEA.990275
- 10 km: DOI: 10.1594/PANGAEA.990265
- 25 km: DOI: 10.1594/PANGAEA.990274
End Users:
The UMD-RDA data product provides a monthly multiyear history of sea ice surface roughness and ice deformation across the Arctic Ocean at the pan-Arctic scale. These data have a wide variety of uses from operational activities to modeling to basic environmental research. For those operating in ice-covered Arctic waters, such as mariners or those conducting underwater operations, the data provide a seasonal perspective of ice pack convergence and deformation vital for selecting safe routing of ship traffic and other marine vessels. The data will improve the characterization of sea ice dynamics in high-resolution sea ice models and will also inform the parametrization of form drag (flow) above and beneath the ice. These data also form the basis for calibration and validation for sea ice backscatter from other satellite missions such as Sentinel-3, Sentinel-6, SWOT, ASCAT, RCM and NISAR, and the roughness distribution can be integrated in the modeling of radar altimeter waveforms for improved sea ice thickness estimation.
References
Andersen, O. B., Rose, S. K., Abulaitijiang, A., Zhang, S., and Fleury, S. (2023). The DTU21 global mean sea surface and first evaluation, Earth Syst. Sci. Data, 15, 4065–4075. DOI: 10.5194/essd-15-4065-2023
Brodzik, M. J., B. Billingsley, T. Haran, B. Raup, M. H. Savoie. 2012. EASE-Grid 2.0: Incremental but Significant Improvements for Earth-Gridded Data Sets. ISPRS International Journal of Geo-Information, 1(1):32-45, DOI: 10.3390/ijgi1010032
Duncan, K., & Farrell, S. L. (2022). Determining variability in Arctic sea ice pressure ridge topography with ICESat-2. Geophysical Research Letters, 49, e2022GL100272. DOI: 10.1029/2022GL100272
Duncan, K., & Farrell, S. L. (2022). Investigating how Arctic sea ice ridge topography varies with ice age. Poster presented at American Geophysical Union Fall Meeting, Chicago, IL, USA, December 12-16, 2022. DOI: 10.22541/essoar.167214354.46100162/v1
Duncan, K., & Farrell, S. L. (2026). ICESat-2 University of Maryland-Ridge Detection Algorithm Gridded Arctic Sea Ice Deformation Data Product, User Guide, Version 1, DOI: 10.1594/PANGAEA.990265
Farrell, S. L., Duncan, K., Buckley, E. M., Richter-Menge, J., & Li, R. (2020). Mapping sea ice surface topography in high fidelity with ICESat-2. Geophysical Research Letters, 47, e2020GL090708. DOI: 10.1029/2020GL090708
Kwok, R., Petty, A. A., Cunningham, G., Markus, T., Hancock, D., Ivanoff, A., Wimert, J., Bagnardi, M., Kurtz, N. & the ICESat-2 Science Team (2025). ATLAS/ICESat-2 L3A Sea Ice Height (ATL07, Version 7). [Data Set]. NASA National Snow and Ice Data Center Distributed Active Archive Center, Boulder, Colorado USA. DOI: 10.5067/ATLAS/ATL07.007
Markus, T., Neumann, T., Martino, A., Abdalati, W., Brunt, K., Csatho, B., Farrell, S., Fricker, H., Gardner, A., Harding, D., Jasinski, M., Kwok, R., Magruder, L., Lubin, D., Luthcke, S., Morison, J., Nelson, R., Neuenschwander, A., Palm, S., Popescu, S., Shum, C. K., Schutz, B. E., Smith, B., Yang, Y., & Zwally, J. (2017). The Ice, Cloud, and land Elevation Satellite-2 (ICESat-2): Science requirements, concept, and implementation. Remote Sensing of Environment, 190, 260–273. DOI: 10.1016/j.rse.2016.12.029