Re: [ferret_users] ERA5 only shows up as a time-series

[Ansley Manke - NOAA Affiliate <ansley.b.manke@xxxxxxxx>]

Hi Alexander,

I believe this is because of the variable namse in the file.  "T" is a reserved name in Ferret/PyFerret, referring to coordiantes in the T direction.  This means thats SHOW GRID T is interpreted as asking about the "abstract" T axis, a pre-defined general axis in the T direction. 

When there are conflicts like this ,you can refer to the file variable T by enclosing it in single quotes, 'T'

Here's the documentation abou this:

https://ferret.pmel.noaa.gov/Ferret/documentation/users-guide/variables-xpressions/VARIABLES#_VPINDEXENTRY_filevars

One thing you might do is to rename the variable in the Ferret session:

https://ferret.pmel.noaa.gov/Ferret/documentation/users-guide/commands-reference/SET#_VPINDEXENTRY_1759

So you might do something like this, with a name that makes sense for the data

yes? use era_T_W_A1.nc

yes? set variable/name=temperature  'T'

Then within the session, refer to temperature  instead of T.

Ansley

On Sun, Dec 29, 2019 at 8:02 AM Alexander Audet <alexander.c.audet@xxxxxxxxx> wrote:

Hello,

          I recently tried downloading and using portions of ERA5. However, it appears that ferret only recognizes ERA5 as a time series instead of as 4D data.

          I have experimented a little bit with show grid and even setting a grid. This has revealed that while Showing the data suggests that there is data in 4 dimensions, showing the grid for one of the data variables shows that the grid is abstract and entirely in the T dimension. 

         Setting a grid then use(ing) the dataset seems to give it a grid structure, but when I try to list a subset of the dataset (doesn't allow me to list the whole dataset) the Lat-Long coordinate spaces on the grid don't seem to have data (the data points listed seem to only be listed with time).

          I both tried downloading directly from the Copernicus Climate Data Store and importing the data into my Windows 10 bash Ubuntu which created a file with a .nc extension, and downloading the data using cdsapi (Climate Data Store Application Program Interface) to download it directly into my bash environment. The second method gave me product with a .netcdf extension. Except for some issues getting all the desired variables using the second method, both products acted identically in terms of the problem outlined above.

Thank you!

Alexander Audet