TOPAZ4b¶

Copernicus Marine Arctic Ocean Physics Reanalysis — TOPAZ4b (ARCTIC_MULTIYEAR_PHY_002_003)

Available

At a glance¶

Note on period: 1990 is used for spin-up, and data start in 1991. As checked on 2026-10-12, the Copernicus Marine catalogue ends at 2026-06-30 for daily data and 2026-06 for monthly data. Per the PUM, the previous year is updated each July (near-real-time inputs) and again each November (reprocessed inputs), so the most recent period may later be regenerated with different inputs.

Overview¶

A coupled ocean and sea-ice reanalysis of the Arctic Ocean and North Atlantic produced by NERSC for Copernicus Marine. It couples the HYCOM ocean model with the CICE sea-ice model and assimilates, in a weekly cycle with a 100-member Deterministic Ensemble Kalman Filter (DEnKF), satellite sea level anomalies, SST, sea ice concentration, sea ice drift and sea ice thickness (from 2010), plus in situ temperature and salinity profiles. From 1991 it provides daily, monthly and yearly temperature, salinity, currents, and sea ice concentration, thickness and drift on a 12.5 km grid, and is used to study Arctic sea-ice change, Atlantic Water inflow and freshwater storage.

Available variables¶

Variable names as given in the Copernicus Marine Product User Manual (PUM) and the dataset metadata. The daily, monthly and yearly datasets (cmems_mod_arc_phy_my_topaz4_P1D-m, _P1M, _P1Y) contain the same variables. On the original (polar stereographic) grid, vector components follow the grid x and y axes and must be rotated when converting to latitude-longitude.

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

3D ocean (40 depths)¶

2D ocean¶

Sea ice¶

Surface fluxes (separate datasets)¶

Static fields¶

The full variable list is in the official documentation.

How to get the data¶

• Copernicus Marine Data Store — ARCTIC_MULTIYEAR_PHY_002_003 — Web download (Subset, File Browser) / Copernicus Marine Toolbox (copernicusmarine)
  • Requires a free Copernicus Marine account. Dataset IDs — daily means cmems_mod_arc_phy_my_topaz4_P1D-m, monthly cmems_mod_arc_phy_my_topaz4_P1M, yearly cmems_mod_arc_phy_my_topaz4_P1Y (current version 202506); fluxes cmems_mod_arc_phy_my_mflux_P1D-m and _P1M-m, cmems_mod_arc_phy_my_hflux_P1D-m and _P1M-m. Each dataset is offered both interpolated to 0.125° latitude-longitude (default) and on the original 12.5 km polar stereographic grid (originalGrid).
• Copernicus Marine Toolbox documentation — Python / command line (CLI)
  • Version 2.x is the supported release.

Example code¶

Install with pip install copernicusmarine, then run login() once to store your credentials.

import copernicusmarine

# Once only: copernicusmarine.login()

# Daily Arctic sea ice concentration and thickness (incl. AI-corrected) — default 0.125° lat-lon grid
copernicusmarine.subset(
    dataset_id="cmems_mod_arc_phy_my_topaz4_P1D-m",
    variables=["siconc", "sithick", "sithick_corr", "vxsi", "vysi"],
    minimum_longitude=-180, maximum_longitude=180,
    minimum_latitude=65, maximum_latitude=90,
    start_datetime="2005-03-01", end_datetime="2005-03-31",
    output_filename="topaz4b_seaice_200503.nc",
)
# For the original 12.5 km polar stereographic grid, add dataset_part="originalGrid"
# and give the area as minimum_x/maximum_x, minimum_y/maximum_y (units of 100 km).

Practical tips¶

• sithick jumps around 2010, when CryoSat-2 and SMOS thickness assimilation begins. For sea ice thickness or volume trends from 1991, use sithick_corr and state in your paper which variable you used.
• On the original grid, vectors (currents, ice velocity, wind stress) are aligned with the polar stereographic grid, and the PUM says they must be rotated when reprojecting to latitude-longitude. When loading original-grid files into GIS, convert the x and y coordinates from 100 km to metres and use the projection's sphere (radius 6378273 m).
• The default 0.125° interpolated product is slightly smoother than the original grid and has a slightly larger land mask. For straits and coastal transports, prefer the original grid.
• At about 12 km spacing the model cannot resolve Arctic mesoscale eddies (Rossby radius 1–2 km). Compare eddy-scale features with higher-resolution model output.
• Vertical interpolation (cubic splines) can overshoot where temperature and salinity gradients are strong, and a few cells near the seabed may be intermittently masked.
• GLORYS12 (global 1/12°) also covers the Arctic, but TOPAZ4b is an Arctic-specific reanalysis that assimilates sea ice drift and thickness directly. Comparing the two gives a sense of the uncertainty in sea ice analyses.

Notices and change history¶

References for your paper¶

Dataset citation¶

Product DOI — DOI 10.48670/moi-00007

@misc{cmems_topaz4b,
  author    = {{E.U. Copernicus Marine Service Information (CMEMS)}},
  title     = {Arctic Ocean Physics Reanalysis ({ARCTIC\_MULTIYEAR\_PHY\_002\_003})},
  year      = {2020},
  publisher = {Mercator Ocean International, Marine Data Store (MDS)},
  doi       = {10.48670/moi-00007},
  note      = {Accessed on DD-MMM-YYYY}
}

Reference papers¶

TOPAZ4 reanalysis quality assessment — DOI 10.5194/os-13-123-2017

@article{xie2017topaz4,
  author  = {Xie, Jiping and Bertino, Laurent and Counillon, Fran{\c{c}}ois and Lis{\ae}ter, Knut A. and Sakov, Pavel},
  title   = {Quality assessment of the {TOPAZ4} reanalysis in the {Arctic} over the period 1991--2013},
  journal = {Ocean Science},
  volume  = {13},
  pages   = {123--144},
  year    = {2017},
  doi     = {10.5194/os-13-123-2017}
}

AI-corrected sea ice thickness (sithick_corr) — DOI 10.5194/tc-19-731-2025

@article{edel2025sit,
  author  = {Edel, L{\'e}o and Xie, Jiping and Korosov, Anton and Brajard, Julien and Bertino, Laurent},
  title   = {Reconstruction of {Arctic} sea ice thickness (1992--2010) based on a hybrid machine learning and data assimilation approach},
  journal = {The Cryosphere},
  volume  = {19},
  pages   = {731--752},
  year    = {2025},
  doi     = {10.5194/tc-19-731-2025}
}

SMOS sea surface salinity assimilation (cited in the PUM) — DOI 10.5194/os-19-269-2023

Acknowledgements¶

Copernicus Marine recommends the acknowledgement "This study has been conducted using E.U. Copernicus Marine Service Information;" followed by the product DOI (10.48670/moi-00007). The PUM names no specific reference paper, so cite Xie et al. (2017) for the system and, if you used sithick_corr, Edel et al. (2025).

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 page (ARCTIC_MULTIYEAR_PHY_002_003)
• Product User Manual (PUM)
• Quality Information Document (QUID)
• Arctic analysis and forecast product (TOPAZ5, ARCTIC_ANALYSISFORECAST_PHY_002_001)
• How to cite or reference Copernicus Marine products and services