I would like to make sure the retrieved data 2-3 months later is expver 1 as this information seems to be lost when requesting netcdf verses a request for grib which does have this metadata. This keyword seems to be ignored as a request using expver=1 for near real time data returned expver 5 instead of the expected 'no data available'.
Hi Andrew,
No, the expver keyword is not currently recognised by the CDS API; it just returns all available data for the specified time period,
Thanks,
Kevin
I am curious then, is there any way to know if data downloaded from CDS API is ERA5-land or ERA5-land-T, whether in the netcdf or in the request?
If not, the only way to be sure to only ever get the "final" data is to run a request with 3 months delay?
Thanks,
Hi Elodie,
If there is only 1 'experiment version' in the period requested, then there will be no 'expver' dimension in the netcdf file, so you cannot tell.
You could run requests with 3 months delay (although you should still check the location of the transition from expver=1 (ERA5-Land) to expver=5 (ERA5-LandT).
In the vast majority of cases, the ERA5-LandT data are identical to the ERA5-Land data for a given timestep.
Hope that helps,
Kevin