Closed jmrgonza closed 1 year ago
Fixed via #9
Hi @jmrgonza , Sorry for this! I worked on #9 without actually realising it was related to this issue. Thus I also missed your point. I will look int o it.
@jmrgonza comments by mail:
I saw the code modifications, but it does not solve the problem. The grid used in the runoff-mapper maps is a gaussian one. The latitudes are not equidistant.
Now fixed with N128 grid (transposed) via #11
I think that the grid used for the definition of the
RNFA
grid is wrong: https://github.com/uwefladrich/rdy2cpl/blob/701a45aa090fac1f855fb5b3d607d6cd0e6ef204/rdy2cpl/model_spec/ecearth.py#L95 According to the coordinates in therunoff_maps.nc
file, I think that theRNFA
grid should be a full gaussian grid with 512 longitudes and 256 latitudes. It may be important that the data in this file is saved with locations in the southern hemisphere coming first, in contrast with the usual layout of data in theIFS
gaussian grids.