NetCDF items change when downloading more variables


Today (19th of April) I downloaded a NetCDF from “ERA5 hourly data on pressure levels from 1940 to present”. I noticed that the values for a variable change when I download more variables. E.g., the mean temperature of my NetCDF was 219 K when I only downloaded the temperature variable. When I downloaded the same NetCDF, but now also with geopotential, the mean became 217 K. Things got even worse when I added more variables (mean temp = 5.8e-7 K).

I am positive that I only took the mean of the temperature, inspecting the individual values of temperature confirmed that they changed too.

Could the NetCDF data still be corrupted?

Best regards,

please refer to the announcement posted here: You may need to re-download data from the CDS

Apologies for the inconvenience caused.