CS3SMOS¶

CryoSat-2/Sentinel-3/SMOS Merged Sea Ice Thickness (CS3SMOS, v300)

Available

At a glance¶

Note on period: Based on the public distributions. Copernicus Marine (SEAICE_GLO_PHY_L4_NRT_011_014) holds v300 Northern Hemisphere data from 2023-10-18 and Southern Hemisphere data from 2025-04-18; AWI near-real-time operations began with the 2025/26 winter (first weekly field 2025-10-15 to 21). Northern Hemisphere data cover 15 October–15 April and Southern Hemisphere data 15 April–15 October only. AWI presented a 15-year record from 2010 at ESA LPS25, but the official pages do not say where that full reprocessing is published.

Overview¶

Weekly sea ice thickness from merging radar altimeter thickness from CryoSat-2 and Sentinel-3A/B, which is most reliable over thick ice, with SMOS L-band radiometer thickness, which resolves thin ice, by optimal interpolation. It is version v300, the successor to CS2SMOS (v206). The grid is refined to 12.5 km, Antarctic winter thickness is produced for the first time, and the files are simplified to the parameters needed for sea ice volume. AWI produces it under an ESA project and publishes it on the AWI FTP server; Copernicus Marine redistributes it as the near-real-time L4 product SEAICE_GLO_PHY_L4_NRT_011_014.

Available variables¶

Variables as given in the v300 file header in the Copernicus Marine Product User Manual (PUM) and in the Copernicus Marine metadata. AWI states that the v300 file content was simplified to the parameters needed for sea ice volume; the v206 background, innovation and per-sensor thickness variables are not included.

instantaneous mean accumulated maximum minimum static — instantaneous: value at the time stamp · mean/accumulated/maximum/minimum: over the indicated interval · static: does not change with time

Merged thickness¶

Auxiliary fields and flags¶

The full variable list is in the official documentation.

How to get the data¶

• AWI FTP — sea ice product root — Anonymous FTP (web browser, FTP client or command line) No login needed
  • No account needed. According to the quicklook links on the AWI wiki, the latest Northern Hemisphere files are under cryosat2_smos/v300/nh/LATEST/; check the FTP listing for the Southern Hemisphere and past-season folders.
• Copernicus Marine — SEAICE_GLO_PHY_L4_NRT_011_014 — Copernicus Marine Toolbox (Python, CLI) / web download
  • Free Copernicus Marine account required. Dataset IDs: Arctic esa_obs-si_arc_phy-sit_nrt_l4-multi_P1D-m (from 2023-10-18), Antarctic esa_obs-si_ant_phy-sit_nrt_l4-multi_P1D-m (from 2025-04-18). FMI renames the files slightly, to W_XX_ESA_SMOS_CS2_12P5KM_EASE2__o_v300_01_l4sit.nc.
• AWI CS3SMOS product page (documents and change log) — Web documentation
  • Summarises the v300 changes. The attached PDD and Read-me documents (v1.5) still describe the previous CS2SMOS.
• AWI CS2SMOS-PDS wiki (announcements) — Web documentation
  • The first place where season start and end, data gaps and quality warnings are announced. Contact cs2smos-support@awi.de.

Example code¶

Assumes a Northern Hemisphere v300 file from the AWI FTP server or Copernicus Marine. AWI originals (quicklook naming W_XX-ESA,SMOS_CS2_S3A_S3B,NH_12P5KM_EASE2_…) and Copernicus Marine files (W_XX_ESA_SMOS_CS2_NH_12P5KM_EASE2_…) are named slightly differently, so match on the end of the name.

import glob
import xarray as xr

f = sorted(glob.glob("*NH_12P5KM_EASE2_*_v300_01_l4sit.nc"))[-1]
ds = xr.open_dataset(f)

sit = ds["sea_ice_thickness"].squeeze()        # m, mean thickness over the ice-covered fraction
sic = ds["sea_ice_concentration"].squeeze()    # %
qf = ds["quality_flag"].squeeze()              # 0 altimetry+radiometry, 1 altimetry, 2 radiometry, 3 interpolated, 4 no sea ice
ocean = ds["density_of_ocean"]                 # ocean fraction per cell (0–1)

print(ds["time"].values, ds["time_bnds"].values)   # window centre and 7-day bounds

# Sea ice volume (km3) = cell area × ocean fraction × concentration × thickness (formula from the PUM)
cell_area = 12.5e3 * 12.5e3                    # m2
vol = (cell_area * ocean * sic / 100 * sit).sum() / 1e9
print(f"{float(vol):.0f} km3")

# Fraction of ice cells estimated from altimetry only, without SMOS (check RFI-affected seas)
ice = qf <= 3
print(float((qf == 1).sum() / ice.sum()))

Practical tips¶

• Differences from v206: the grid changed from 25 km EASE2 Northern Hemisphere (432 × 432) to 12.5 km EASE2 for both hemispheres (864 × 864, EPSG:6931/6932), Sentinel-3A/B altimetry was added, and the SMOS input moved from v3.3 to v3.6. The land/ocean mask now matches the ESA CCI sea ice products.
• The uncertainty model has changed. Altimetry uncertainty now follows AWI v2.6, which separates systematic and random components, and is substantially higher than in v206; SMOS uncertainty adds the effect of concentrations below 100% and has a 20 cm minimum. Larger uncertainties do not mean lower quality, so do not filter v300 with thresholds tuned for v206.
• The file format is simpler: v206 variables such as analysis_sea_ice_thickness, background_sea_ice_thickness, innovation, smos_sea_ice_thickness and cryosat_sea_ice_thickness are gone, replaced by sea_ice_thickness, status and quality flags and the ocean fraction. File names also changed (SMOS_CS2 → SMOS_CS2_S3A_S3B, 25KM → 12P5KM, v206 → v300), so existing scripts need updating.
• Files on consecutive days are 7-day means that overlap by six days. For independent samples take every seventh file, and align dates using time (window centre) or time_bnds.
• The L4_SIT_Open collection on the ESA SMOS dissemination server stops in April 2025 (v206); get v300 from the AWI FTP server or Copernicus Marine. The C3S Climate Data Store's 'Sea ice thickness gridded data for the Arctic' L4 (CryoSat-2, Sentinel-3, SMOS; Arctic October–April) is a separate distribution, and the Copernicus Marine QUID states that the reprocessed version of this product is to be published in the Climate Data Store.

Notices and change history¶

References for your paper¶

Dataset citation¶

Copernicus Marine distribution (SEAICE_GLO_PHY_L4_NRT_011_014) — DOI 10.48670/moi-00125

@misc{cmems_seaice_glo_phy_l4_nrt_011_014,
  author       = {{E.U. Copernicus Marine Service Information (CMEMS)}},
  title        = {Sea Ice Thickness derived from merging of {L-Band} radiometry and radar altimeter derived sea ice thickness ({SEAICE\_GLO\_PHY\_L4\_NRT\_011\_014})},
  publisher    = {Marine Data Store (MDS)},
  doi          = {10.48670/moi-00125},
  note         = {Accessed on DD MMM YYYY}
}

Reference papers¶

CS2SMOS/CS3SMOS merging method paper (citation requested by AWI) — DOI 10.5194/tc-11-1607-2017

@article{ricker2017cs2smos,
  author  = {Ricker, Robert and Hendricks, Stefan and Kaleschke, Lars and Tian-Kunze, Xiangshan and King, Jennifer and Haas, Christian},
  title   = {A weekly {Arctic} sea-ice thickness data record from merged {CryoSat-2} and {SMOS} satellite data},
  journal = {The Cryosphere},
  volume  = {11},
  number  = {4},
  pages   = {1607--1623},
  year    = {2017},
  doi     = {10.5194/tc-11-1607-2017}
}

SMOS Antarctic thin sea ice thickness paper (for Southern Hemisphere data; cited in the Copernicus Marine PUM) — DOI 10.5194/essd-16-3149-2024

@article{kaleschke2024smos_antarctic,
  author  = {Kaleschke, Lars and Tian-Kunze, Xiangshan and Hendricks, Stefan and Ricker, Robert},
  title   = {{SMOS}-derived {Antarctic} thin sea ice thickness: data description and validation in the {Weddell Sea}},
  journal = {Earth System Science Data},
  volume  = {16},
  number  = {7},
  pages   = {3149--3170},
  year    = {2024},
  doi     = {10.5194/essd-16-3149-2024}
}

Acknowledgements¶

Along with Ricker et al. (2017), AWI asks for the acknowledgement: "The production of the merged CryoSat-SMOS sea ice thickness data was funded by the ESA project SMOS & CryoSat-2 Sea Ice Data Product Processing and Dissemination Service, and data from DATE to DATE were obtained from AWI." Check the AWI CS2SMOS-PDS wiki for the current wording. As of October 2026 the official pages give no dataset DOI or method paper specific to v300. If you downloaded from Copernicus Marine, add "This study has been conducted using E.U. Copernicus Marine Service Information" together with the DOI.

Information on the datasets used in this study was compiled with the help of the Korean Climate Data Navigator (KCDN) (Sim, 2026; https://doi.org/10.5281/zenodo.23300923).

Links¶

• Copernicus Marine Product User Manual (PUM) — file header, variables, time convention
• Copernicus Marine Quality Information Document (QUID) — validation and input changes
• ESA LPS25 poster — 15 Years of Global Winter Sea Ice Thickness & Volume from CS3SMOS
• AWI Sea Ice Thickness from Satellite Radar Altimetry (CryoSat-2, Sentinel-3; altimetry input v2.6)
• ESA SMOS-CryoSat L4 Sea Ice Thickness catalogue (previous v206)
• This site — SMOS thin sea ice thickness (a CS3SMOS input)