ECCO V4¶

Estimating the Circulation and Climate of the Ocean (ECCO) Version 4 Central Estimate — current main public release V4r4

Available

At a glance¶

Note on period: V4r4 covers 1992-01-01 to 2017-12-31 (PO.DAAC metadata end 2018-01-01; the last monthly-mean file when checked on 2026-10-12 was 2017-12). V4r5 covers 1992-01 to 2019-12 but is only partly available, and V4r6 (1992–2025) has yet to be released.

Overview¶

ECCO V4 is a "state estimate": the initial conditions, model parameters and atmospheric forcing (ERA-Interim plus estimated adjustments) of the MITgcm ocean–sea-ice model are adjusted so that a free-running solution fits, in a least-squares sense, the observations available since 1992 — satellite altimetry, SST, sea-ice concentration, GRACE bottom pressure, Argo and other in-situ data. Unlike a conventional reanalysis, it does not correct the model state at each observation time, so the result satisfies the model equations exactly: heat, salt, volume and momentum are conserved and budgets close. This makes it especially useful for work on why the ocean changed, such as ocean heat and freshwater budgets and attribution of sea-level change.

What changes with the successor¶

ECCO V4r4 → ECCO V4r5¶

Transition: 2025-05

A release on the same LLC90 grid and MITgcm framework that extends the optimization period to the end of 2019, uses the adjoint of the thermodynamic sea-ice model, and adds Antarctic ice shelves and tidewater glaciers. As of October 2026, however, it is not in PO.DAAC's regular downloads (Earthdata Search); only some native-grid monthly-mean variables can be obtained from an AWS requester-pays bucket.

Sources: Synopsis of the ECCO V4r5 Global Ocean and Sea-Ice State Estimate (Zenodo, 2024) · ecco_access documentation — Accessing ECCOv4 release 5 output in the AWS Cloud

ECCO V4r4 / V4r5 → ECCO V4r6 Planned¶

In replies on the ECCO support mailing list in April–May 2026, the ECCO team (Ou Wang, JPL) said V4r6, covering 1 January 1992 to 31 December 2025, is being prepared, will be officially released through PO.DAAC, and will supersede V4r5. Monthly means are to be posted on ECCO Drive first, but no release date has been set.

Sources: ecco-support mailing list — V4r6 announcement (2026-05-14) · ecco-support mailing list — ECCO data after end of 2017 (2026-04)

Available variables¶

Main variables in the PO.DAAC V4r4 datasets. Names are the variable names inside the netCDF files. Variables are grouped into thematic datasets (ShortName), and the same variable appears separately on llc90 (…LLC0090GRID…) and 0.5° (…05DEG…) grids and as daily (DAILY) and monthly (MONTHLY) means.

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

3-D ocean (50 levels)¶

Sea surface height and mixed layer¶

Surface fluxes¶

Sea ice and snow¶

Snapshots (instantaneous, llc90)¶

The full variable list is in the official documentation.

How to get the data¶

• NASA PO.DAAC — ECCO datasets — Earthdata Search / HTTPS download / earthaccess and ecco_access (Python) / direct access in AWS us-west-2
  • Requires a free NASA Earthdata account. V4r4 is split into dozens of datasets by theme, grid and time interval (e.g. ShortName ECCO_L4_TEMP_SALINITY_05DEG_MONTHLY_V4R4).
• ECCO project — V4r4 product page — Information page (citations, errata, documentation)
  • The page still said "The ECCO Drive is temporarily offline" when checked in October 2026, but the ECCO team announced on 2026-08-28 that ECCO Drive is accessible again.
• ECCO Drive — HTTPS (WebDAV) — log in with an Earthdata account
  • Hosts llc90 native-grid files (V4r4 monthly means, forcing input files, etc.) and is where V4r6 monthly means are to appear.
• ecco_access (ECCOv4-py) — V4r5 cloud access guide — Python (inside the AWS cloud only)
  • V4r5 monthly means are in the requester-pays bucket s3://ecco-model-granules/.

Example code¶

Run after pip install earthaccess xarray netcdf4. On the first login(), enter your Earthdata username and password (or store them in ~/.netrc). The ShortName is the PO.DAAC dataset name.

import earthaccess
import xarray as xr

earthaccess.login()

# V4r4 temperature and salinity, monthly means on the 0.5° lat-lon grid
results = earthaccess.search_data(
    short_name="ECCO_L4_TEMP_SALINITY_05DEG_MONTHLY_V4R4",
    temporal=("2010-01-01", "2010-12-31"),
)
files = earthaccess.download(results, "ecco_v4r4_ts")
ds = xr.open_mfdataset(files)
print(ds["THETA"])   # (time, Z, latitude, longitude)

Practical tips¶

• The 0.5° interpolated fields are for plotting and comparison. To close heat, salt or volume budgets, use the llc90 native-grid data with the grid geometry dataset (ECCO_L4_GEOMETRY_LLC0090GRID_V4R4).
• llc90 is stored as 13 tiles (tile, j, i), so it cannot be subset like a lat-lon array. The ecco_v4_py package provides tools for plotting tiles, interpolation and budget calculations.
• UVEL and VVEL are along the model grid axes, which are not east–west and north–south in the Arctic cap and elsewhere. For eastward and northward velocity, use EVEL and NVEL from the 0.5° data or rotate using the grid angles.
• June 2021 errata (V4r4b): the 2014 atmospheric pressure forcing was wrong (April and December swapped, May–November shifted by one day), affecting daily sea level and bottom pressure in 2013–2015. Use the V4R4B datasets if you need daily data for this period.
• V4r4 atmospheric forcing is ERA-Interim plus control adjustments estimated by the ECCO optimization (wind speed and air pressure are not adjusted). Do not treat the atmospheric surface state datasets as identical to the original ERA-Interim.
• ECCO V5 Alpha, ECCO2 and ECCO-Darwin are different products. State the exact release name (e.g. V4r4) in your paper.

Notices and change history¶

References for your paper¶

Dataset citation¶

PO.DAAC dataset DOI (example — temperature and salinity, monthly mean 0.5°) — DOI 10.5067/ECG5M-OTS44

@misc{ecco_v4r4_ts_05deg_monthly,
  author    = {{ECCO Consortium} and Fukumori, Ichiro and Wang, Ou and Fenty, Ian and Forget, Ga{\"e}l and Heimbach, Patrick and Ponte, Rui M.},
  title     = {{ECCO} Ocean Temperature and Salinity - Monthly Mean 0.5 Degree (Version 4 Release 4)},
  year      = {2020},
  publisher = {NASA Physical Oceanography Distributed Active Archive Center (PO.DAAC)},
  version   = {V4r4},
  doi       = {10.5067/ECG5M-OTS44}
}

PO.DAAC dataset DOI (example — temperature and salinity, monthly mean llc90 native grid) — DOI 10.5067/ECL5M-OTS44

V4r4 synopsis — DOI 10.5281/zenodo.4533349

@techreport{ecco_v4r4_synopsis,
  author      = {{ECCO Consortium} and Fukumori, Ichiro and Wang, Ou and Fenty, Ian and Forget, Ga{\"e}l and Heimbach, Patrick and Ponte, Rui M.},
  title       = {Synopsis of the {ECCO} Central Production Global Ocean and Sea-Ice State Estimate (Version 4 Release 4)},
  institution = {Zenodo},
  year        = {2021},
  doi         = {10.5281/zenodo.4533349}
}

Reference papers¶

ECCO Version 4 framework paper — DOI 10.5194/gmd-8-3071-2015

@article{forget2015ecco,
  author  = {Forget, Ga{\"e}l and Campin, Jean-Michel and Heimbach, Patrick and Hill, Christopher N. and Ponte, Rui M. and Wunsch, Carl},
  title   = {{ECCO} version 4: an integrated framework for non-linear inverse modeling and global ocean state estimation},
  journal = {Geoscientific Model Development},
  volume  = {8},
  number  = {10},
  pages   = {3071--3104},
  year    = {2015},
  doi     = {10.5194/gmd-8-3071-2015}
}

V4r5 synopsis (if you used V4r5) — DOI 10.5281/zenodo.13926715

Acknowledgements¶

The ECCO V4r4 product page states that "as a condition of using this product, you must use the following citations": (1) the dataset, (2) the V4r4 synopsis (10.5281/zenodo.4533349) and (3) Forget et al. (2015). Dataset DOIs differ for each PO.DAAC dataset (ShortName), so cite the DOI of the dataset you actually downloaded.

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¶

• ECCO project home page
• ECCO V4r4 product page (citations and errata)
• ECCO V4r4 user guide (PDF)
• ECCO dataset gallery (list of PO.DAAC datasets)
• ECCO v4 Python tutorial
• ecco_access — querying ECCO variables
• ecco-support mailing list archive