ORAS5¶

ECMWF Ocean Reanalysis System 5 (ORAS5)

Available

At a glance¶

Note on period: 1958–2014 is the Consolidated product, forced by reanalysis atmospheric fields (ERA-40, ERA-Interim); 2015 onward is the Operational product, forced by ECMWF operational analyses. When checked on 2026-10-11, data were available through August 2026.

Overview¶

A global ocean and sea-ice reanalysis produced with ECMWF's OCEAN5 system, which couples the NEMO ocean model with the NEMOVAR (3D-Var FGAT) assimilation system to ingest subsurface temperature and salinity, sea-level anomalies and sea-ice concentration. The same system provides the ocean initial conditions for ECMWF seasonal forecasts. The CDS publishes monthly means from one of the five ensemble members.

What changes with the successor¶

ORAS5 → ORAS6 Planned¶

ECMWF's next ensemble reanalysis system for ocean and sea ice keeps the resolution (¼°, 75 levels) but changes the model (NEMO4/SI3), the atmospheric forcing (hourly ERA5) and the data assimilation (ensemble-based 3D-Var with direct SST assimilation). It provides the ocean and sea-ice boundary conditions for ERA6. As of June 2026 the 1993-onward period was to go first to ECMWF Member States and external collaborators; no CDS release date has been announced.

Sources: ECMWF Newsletter 180 — ECMWF's next ensemble reanalysis system for ocean and sea ice (2024) · ECMWF news — New ORAS6 ocean reanalysis (2026-06-08) · Browne et al. (2026) Sea ice data assimilation in ORAS6, The Cryosphere · Zuo et al. (2019) ORAS5 system paper

Compare the two side by side

Available variables¶

Main variables from the CDS download form. Names in backticks are those used in CDS API requests; the variable names inside the downloaded NetCDF files may differ, so open a file to check.

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 (75 levels, vertical_resolution=all_levels)¶

Ocean single level (vertical_resolution=single_level)¶

Surface fluxes¶

Sea ice¶

The full variable list is in the official documentation.

How to get the data¶

• Copernicus Climate Data Store (CDS) — reanalysis-oras5 — Web download / CDS API (cdsapi)
  • Requires an ECMWF account and acceptance of the dataset licence (CC-BY).
• CDS — ORAS5 monthly time-series (ARCO) dataset — Web download / CDS API / Zarr (beta)
  • An experimental dataset released in April 2026, restructured for fast retrieval of long time series at a single point; it contains a subset of variables and can output CSV. Use the original dataset when you need full spatial fields.
• Universität Hamburg ICDC (EASY-Init ORAS5 mirror) — Web download
  • A separate mirror covering only some periods (e.g. 1979–2018). ICDC notes masking errors near coasts and the sea floor in the files interpolated to 1°, so use the native-resolution data.

Example code¶

Put the CDS API URL and your personal key in ~/.cdsapirc before running. Request keys and values are taken from the CDS download form.

# pip install --upgrade cdsapi
import cdsapi

request = {
    "product_type": ["consolidated"],   # use "operational" for 2015 onward
    "vertical_resolution": "single_level",
    "variable": ["sea_surface_temperature", "sea_surface_height"],
    "year": ["2010"],
    "month": ["01", "02", "03"],
}

client = cdsapi.Client()
client.retrieve("reanalysis-oras5", request).download()  # saved under the server-assigned file name

Practical tips¶

• 1958–2014 (Consolidated) and 2015 onward (Operational) use different atmospheric forcing and observations. For long records you must request both product_types, and allow for a possible discontinuity around 2015.
• The data are on the tripolar model grid, so they cannot be overlaid directly on a latitude–longitude grid. Velocities come both along the model axes and rotated; use the rotated variables when you need eastward and northward components.
• Post-2015 ORAS5 on the CDS uses different forcing and observations from the ORAS5 distributed by Copernicus Marine, so values can differ. State in your paper which source you used.

Notices and change history¶

References for your paper¶

Dataset citation¶

CDS monthly data — DOI 10.24381/cds.67e8eeb7

@misc{oras5_cds,
  author    = {{Copernicus Climate Change Service, Climate Data Store}},
  title     = {{ORAS5} global ocean reanalysis monthly data from 1958 to present},
  year      = {2021},
  publisher = {Copernicus Climate Change Service (C3S) Climate Data Store (CDS)},
  doi       = {10.24381/cds.67e8eeb7},
  note      = {Accessed on DD-MMM-YYYY}
}

Reference papers¶

ORAS5 system paper — DOI 10.5194/os-15-779-2019

@article{zuo2019oras5,
  author  = {Zuo, Hao and Balmaseda, Magdalena Alonso and Tietsche, Steffen and Mogensen, Kristian and Mayer, Michael},
  title   = {The {ECMWF} operational ensemble reanalysis--analysis system for ocean and sea ice: a description of the system and assessment},
  journal = {Ocean Science},
  volume  = {15},
  number  = {3},
  pages   = {779--808},
  year    = {2019},
  doi     = {10.5194/os-15-779-2019}
}

Acknowledgements¶

Acknowledge the use of C3S/ECMWF data and give the CDS dataset DOI and access date. See the "Citation and attribution" section of the CDS dataset page for the recommended wording.

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¶

• CDS ORAS5 dataset page
• ECMWF ORAS5 page
• OCEAN5 system documentation (ECMWF Technical Memorandum)
• ECMWF Forum — oras5 tag
• ECMWF Newsletter 180 — ORAS6