Meaning of field 'time' in "ERA5 hourly data on single levels from 1979 to present"


I have a question about the meaning of field 'time' when downloading data from "ERA5 hourly data on single levels from 1979 to present" (here:!/dataset/reanalysis-era5-single-levels?tab=form) for "Ensemble Mean" (forecasts).

In that case, 'time' can have values 0:00, 3:00, 6:00, 9:00, etc. But given that the model runs twice a day, at 6:00 and 18:00, should I interpret that 'time' 9:00 is the first step of the 6:00-run (6:00+3h), 'time' 12:00 is the second step, and so on?. And that 'time' 21:00 is the first step of the 18:00-run (18:00+3h)? Would that be correct?

Thanks a lot for your help,


1 Like

Hi Ricardo,

Yes the ERA5 data on the CDS are presented in such a way that the 'validity time' is shown as the value of the time coordinate/on the CDS web form. This page explains the relationship between the validity time and the analysis time/forecast step:

ERA5: data documentation

See the section "Date and time specification":

Table 0: the mapping, for forecasts, between MARS date, time and step and the CDS date and time

Hope that helps,


1 Like

Hello Kevin, Table 0 says it all, thanks a lot!!

Regards, Ricardo.