mars

The mars source will retrieve the data from the ECMWF MARS archive. For that, you need to have an ECMWF account and build your dataset on one of the Centre’s computers, or use the ecmwfapi Python package.

The yaml block can contain any keys that following the MARS language specification, with the exception of the date, time` and step.

The missing keys will be filled with the default values, as defined in the MARS language specification.

input:
  mars:
    levtype: sfc
    param: [2t, msl]
    grid: [0.25, 0.25]

Data from several levels types must be requested in separate requests, with the join command.

input:
  join:
  - mars:
      levtype: sfc
      param: [2t, msl]
      grid: [0.25, 0.25]
  - mars:
      levtype: pl
      param: [u, v]
      grid: [0.25, 0.25]

See Naming variables for information on how to name the variables when mixing single level and multi-levels variables in the same dataset.

cds

For users outside of ECMWF organisation, it is possible to access ERA5 data through the Copernicus Climate Data Store cdsapi instead.

The steps to setup a CDS API account are detailed here.

The only difference with the previous MARS recipes is the addition of the use_cdsapi_dataset key:

input:
    mars:
        use_cdsapi_dataset: "reanalysis-era5-complete"
        class: ea
        levtype: sfc
        param: [10u, 10v, 2t, msl]
        grid: n320

This process can take some time because of the high demand on the CDS server.