ERA6¶

ECMWF Reanalysis v6 (ERA6)

Upcoming

At a glance¶

Note on period: ECMWF aims to extend ERA6 back to at least 1950 and up to the present (Copernicus says "over 75 years"; the C3S General Assembly talk says "75 (likely 80) years"). Ten-year streams are run in parallel, each with a one-year spin-up. According to the August 2026 ECMWF Newsletter, the 2007 and 2017 streams are running and the 1987 and 1997 streams are due to start next. As checked on 2026-10-12, no ERA6 data have been released (apart from prototype test data).

Overview¶

The successor to ERA5, produced by ECMWF for the Copernicus Climate Change Service (C3S). Built on IFS Cycle 49r2, it raises the horizontal resolution to about 14 km (ERA5: about 31 km) and is the first ECMWF flagship reanalysis to include an ocean model, coupling NEMO-4 with the SI3 sea-ice model (ocean starting conditions from ORAS6). Rescued and reprocessed historical observations and a better representation of the upper stratosphere aim to address known ERA5 issues. All data will be provided in GRIB2 with WMO units.

Available variables¶

The variable list is not final before release. New outputs announced by ECMWF include 3D and 2D ocean parameters, 11 height levels from 15 to 500 m, multi-layer snow, 3D clear-air turbulence, new surface parameters such as 2 m relative humidity, and hourly data together with daily data and monthly statistics. Variable names and units will be added at release.

How to get the data¶

• Copernicus announcement — ERA6 reanalysis production starts (release schedule and distribution plans) — Information page
  • The distribution channel and dataset names for released data have not been announced. Copernicus mentions fast retrieval of time series and profiles through Polytope and ARCO-ready versions. This entry will be updated at release.
• ECMWF — ERA6 prototype 1 sample data for testing only (June to August 2022) — IFS Experiments portal / MARS-style requests (GRIB or NetCDF)
  • A prototype experiment, not an ERA6 production; ECMWF states it "should not be used for any application or publication". Use it to test workflows (e.g. the move to GRIB2) and to send feedback. DOI 10.21957/gmf0-rw02
• ECMWF — ERA6 prototype test sample dataset 2 (October 2019 to September 2020) — IFS Experiments portal / MARS-style requests (GRIB or NetCDF)
  • Also for testing only (known missing components). DOI 10.21957/1qyd-ep25

Example code¶

The MARS-style request shown on the ECMWF prototype 1 page (hourly 2 m temperature analyses, 0.25° NetCDF), copied as given. The page does not say which tool or account to submit it with. This is test data; do not use it for research or publication.

retrieve,
class=rd,
stream=lwda,
expver=ibpb,
type=an,
levtype=sfc,
param=2t,
date=2022-08-01,
time=00/TO/23/BY/1,
grid=0.25/0.25,format=netcdf,
target='output2.nc'

Practical tips¶

• The first release covers 2007 onward only. For long-term work back to 1940, stay with ERA5 for now, and do not splice ERA5 and ERA6 into one time series.
• All data move to GRIB2 (WMO units). Test GRIB1-only tools and scripts that assume ERA5 parameter names and units on the prototype data ahead of time.
• The date ERA6 replaces ERA5T and the ERA6T latency are not yet set. If your workflow pulls ERA5 up to the present, watch for the transition plan ECMWF has promised.
• The ocean is a coupled model started from ORAS6 initial conditions; ERA6 does not assimilate ocean observations directly. Keep the differences from the ocean reanalysis (ORAS6) in mind when using ocean variables.

Notices and change history¶

References for your paper¶

Reference papers¶

ERA6 production overview (ECMWF Newsletter 188)

@article{hersbach2026era6news,
  author  = {Hersbach, Hans and Bell, Bill and Herbert, Christoph and Kaandorp, Mikael and Poli, Paul and Radu, Raluca and Schepers, Dinand and Soci, Cornel and Steele, Liam and Strube, Cornelia and Browne, Philip and Cobb, Alison and Osinski, Robert and Ribas, Roberto and Simmons, Adrian and Stockdale, Tim and Villaume, Sebastien and Wilhelmsson, Tomas and Buontempo, Carlo and McNally, Tony and Suttie, Martin},
  title   = {{ERA6} reanalysis production},
  journal = {ECMWF Newsletter},
  number  = {188},
  year    = {2026},
  url     = {https://www.ecmwf.int/en/newsletter/188/news/era6-reanalysis-production}
}

ERA6 user requirements (C3S report) — DOI 10.24381/4s0n-jm11

@techreport{c3s2024era6needs,
  author      = {{Copernicus Climate Change Service (C3S)}},
  title       = {Position Paper - {ERA6} User Community Needs},
  institution = {Copernicus Climate Change Service (C3S)},
  year        = {2024},
  doi         = {10.24381/4s0n-jm11}
}

Acknowledgements¶

ERA6 has no dataset DOI yet. For now, cite the items above only when describing the plans; after release, follow the DOI and recommended wording on the distributor's dataset page. Do not use prototype data in publications.

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¶

• ECMWF Newsletter 188 — ERA6 reanalysis production
• Copernicus — Copernicus Climate's ERA6 reanalysis production starts
• EMS 2026 abstract — From ERA5 to ERA6 (Hersbach et al.)
• C3S General Assembly (June 2026) ERA6 talk slides (Hersbach)
• ECMWF IFS Experiments — ERA6 prototype 1