Hello,
I am about to estimate our future storage requirements for locally storing a subset of ERA6 data at native resolution (1H data). Specifically, I am interested in surface (sf), model (ml), and pressure (pl) level data, and would like to roughly calculate the storage requirements by parameter.
Could you provide any indications on how the file size per parameter (1 file with 24 hourly time steps) will roughly be? Alternatively, are there any prototype files available that I could check to get a better sense of the file structure?
Any help would be greatly appreciated.
Best regards,
Angelika
Dear Angelika, it will probably be between 3-4 times larger than for ERA5.
That’s for data in regular lat-lon. For ERA6 we plan to store the data in it’s native grid (which has less grid points towards the poles) so if you can handle those that would be a savings of about 60%, i.e. Increase 0.6*(3-4) = 2-2.5 larger than reg lat-lon ERA5.
Regards,
Hans
Dear Hans,
Thank you very much for your estimate of the ERA6 storage requirements. I have meanwhile looked at the ERA6 prototype test dataset 2 (published Nov 2024: ECMWF | ERA6 prototype test sample dataset 2 from Oct 2019 to Sept 2020).
Unfortunately, these prototype files are still mostly in GRIB1 rather than the final ERA6 file format, GRIB2. Will there be an updated prototype dataset that more closely reflects the final ERA6 format and metadata? This would greatly help users in preparing for a smooth migration from ERA5 to ERA6 in their workflows.
Thanks a lot and kind regards,
Angelika