Revision as of 21:23, 8 April 2020 by P.petrelli (talk | contribs)

ERA5 is a climate reanalysis dataset, covering the period 1950 to present. ERA5 is being developed through the Copernicus Climate Change Service (C3S).

You can use the ERA5 Explorer tool to get a preview of the data (only few variables currently available).

Data processing for ERA5 is carried out by ECMWF, using ECMWFS' Earth System model IFS, cycle 41r2. The name ERA refers to 'ECMWF ReAnalysis', with ERA5 being the fifth major global reanalysis produced by ECMWF (after FGGE, ERA-15, ERA-40, ERA-Interim). The next reanalysis, ERA6, is planned for around 2020.

We are providing a subset of ERA5 while waiting for a longer term solution to store and manage this big dataset. Currently data is available from 1979 to present (1981 to present for ERA5 Land). We are downloading variables on pressure levels, surface, wave model and ERA5 Land, refer to the ERA5 variable list for more details. For the moment we are not considering downloading model level data as the size of these and the download speeds make this difficult and we wouldn't be able to provide enough data to use them to run a model. For more information on ERA5 data refer to the ECMWF website.

If you need other variables for your current research then you can contact us at

NB: this is a temporary fix so we will download only variables which are useful for a variety of projects and that are needed now, we will have to prioritise giving precedence to CLEX users.

Download frequency

Currently ERA5 data is available with a 5 days delay, which means that some of the most recent data hasn't been pre-validated yet.

We will still update the dataset monthly, but up to the previous month. So the data we are providing will have from 15-45 days delay from real-time.

If we find that we have to retract data too often, after CDS has run their QC checks, we will extend this interval to minimise the risk for users of accessing faulty data.

In any case is very important to keep in mind the potential of errors when using the three most recent months of data

Requests constraints

We will only accept requests for:

- data on regular 0.25 by 0.25 lat/lon grid - global region for surface variables, wave model and ERA5 Land 

- Austral-Asian region for pressure level. The region includes Australia and the Maritime continent and the CORDEX domain and is defined as:

Latitude: 20 to -57

Longitude: 78 to 220

NB longitude for the global region is -180 to 180 - analysis data and no ensemble data (analysis includes surface forecast fields such as precipitation).

It is currently impossible to provide input data for a climate model. You can submit a request for downloading following these criteria. We will only consider requests that specify the exact code/stream etc. You can refer for more information to the Copernicus Data Store and to the ERA5" catalogue.

Be as specific as possible in your request:  it is always better to specify the parameter code rather than just the name, the same variable can be available on different streams and remember to tell us if you need 1 hr or monthly data (see Copernicus website.)


Latest updates

ERA5 Land update: monthkly averages are now also available and the data currently goes back to 1981.

ERA5 Land has been released: ERA5-Land is a reanalysis dataset providing a consistent view of the evolution of land variables over several decades at an enhanced resolution compared to ERA5, currently from 2001 to date. years back tom1979 should become available around September. For more information and a view of available variables visit the Copernicus website .

We have started downloading variables and you can request more variables now via the helpdesk. The variables will be downloaded as for the surface fields on a global grid (NB this has a higher resolution than ERA5) at 1 hr temporal resolution.

Renamed variable: form ocean waves 10m wind to ocean surface stress equivalent 10m neutral wind

The reason for the change was a parameter name clash between variables in ERA5 wind and ERA5 ocean waves. The change affected both "wind" and "dwi" variables and caused a delay in downloading these variables from 1979 to 1999 included and from October 2018 to February 2019.  

We added minimum and maximum 2m temperature, these surface variables are the first ones available for the entire period 1979 -2018 together with: mean sea level pressure, 2m temperature, convective precipitation and total precipitation.


Known Issues

Known issues from ECMWF website

Dec 7 2018: an error has been discovered affecting the values of the following variables

CDS Dataset ERA5 hourly data on pressure levels from 2000 to present]
Parameters affected U-component of wind, V-component of wind
Level All pressure levels
Time periods affected 01 Jan. 2000 - 31 Dec. 2008
01 Jun. 2018 - present

This problem was caused by the computation of U/V from Vorticity (Vo) and Divergence (D) already interpolated onto a regular grid in CDS. U/V should have been computed from Vo and D in the spherical harmonics representation (the native model representation) and then interpolated to a regular grid before saving to the climate data store.

The problem is now fixed and we have re-downloading the data, download should be completed by the 19 March 2019. From the CDS communication: "During our investigations, we found that wind values are far too low on pressure levels at the poles in CDS, which has been documented as a known issue. Please use winds at neighbouring locations at 89.75 (N/S) and 89.5 (N/S) for Reanalysis and Ensemble members winds instead of 90 (N/S)."

Issues and other feedback shared by users: you can check this page for issues and other information shared by other users of ERA5, feel free to add your own contribution.


ERA5 data is open access and free to download for all uses, including commercial use. As for ERAI and other ECMWF datasets, though, you need to register with ECMWF and accept their terms and conditions, in this particular case ERA5 is covered by the Copernicus license



data is downloaded as netcdf on a regular 0.25x0.25 longitude/latitude global grid at 1 hr resolution, as available from the Copernicus server. These are the maximum spatial and temporal resolutions possible for this dataset.


filename: <variable>_era5_global_<from-date>_<to-date>.nc

Pressure levels

data is downloaded as netcdf on a regular 0.25x0.25 longitude/latitude grid at 1 hr resolution, on an Australian region defined as in the criteria.


filename: <variable>_era5_aus_<from-date>_<to-date>.nc

NB variable names are as in the ECMWF param tables, our ERAI netcdf files were following where possible the CMIP conventions.

Monthly data will be downloaded on a global grid.

Wave surface level

data is downloaded as netcdf on a regular 0.5x0.5 longitude/latitude grid at 1 hr resolution, on a global grid.


filename: <variable>_era5_global_<from-date>_<to-date>.nc

ERA5 Land

For more specific information on ERA5 Land: copernicus documentation

data is downloaded as netcdf on a regular 0.1x0.1 longitude/latitude grid at 1 hr resolution, on a global grid.


filename: <variable>_era5land_global_<from-date>_<to-date>.nc

Requested variables

ERA5 1 hr variables list

ERA5 monthly variables list


Note on forecasts

Forecasts are accumulated values defined at 20 steps (0-19) starting from times 06.00 and 18.00.

The accumulations are over the hour ending at the forecast step which is a different behaviour from era interim. The 0 step is always zero because it would be the accumulation from timestep and timestep +0 hours (I.e. 06.00 to 06.00)

The others would be for example for precipitation: Time 06.00 Step 1: forecast of accumulated precipitation between 06.00-07.00 Time 06.00 Step 19: forecast of accumulated precipitation between 06.00-01.00+1 day