...
- Sea surface temperature and sea-ice cover (see Table 2 above) are available at the usual times, eg hourly for the HRES, but their content is only updated once daily.
- Mean rates and accumulations at step=0 have values of zero because the length of the processing period is zero.
Expand title Click here to expand...ERA5: mixing CDS and MARS data In the ECMWF data archive (MARS), ERA5 data is archived on various native grids. For the CDS disks, ERA5 data have been interpolated and are stored on regular latitude/longitude grids. For more information, see Spatialgrid.
Storing the data on these different grids can cause incompatibilities, particularly when comparing native spherical harmonic, pressure level, MARS data with CDS disk data on a third, coarse grid.
Native spherical harmonic, pressure level parameters are comprised of: Geopotential, Temperature, U component of wind, V component of wind, Vertical velocity, Vorticity, Divergence and Relative humidity. When these parameters are retrieved from MARS and a coarse output grid is specified, the default behaviour is that the spherical harmonics are truncated to prevent aliasing on the output grid. The coarser the output grid, the more severe the truncation. This truncation removes the higher wavenumbers, making the data smoother. However, the CDS disk data has been simply interpolated to the third grid, without smoothing.
This incompatibility is particularly relevant when comparing ERA5.1 data (which are only available from MARS - see DataorganisationandhowtodownloadERA5 - and only for 2000-2006) with ERA5 data on the CDS disks.
The simplest means of minimising such incompatibilities is to retrieve the MARS data on the same grid as that used to store the ERA5 CDS disk data.
Known issues
Currently, we are aware of these issues with ERA5:
...