Looking at
|
# Set the coordinates for the dataset (needed to be done explicitly for curvilinear grids) |
|
if "lon" in ds: |
|
ds.set_coords("lon") |
|
if "lat" in ds: |
|
ds.set_coords("lat") |
|
|
|
if len(set(grid.axes) & {"X", "Y"}) > 0: # Only if spatial grid is >0D (see #2054 for further development) |
|
assert_valid_lat_lon(ds["lat"], ds["lon"], grid.axes) |
I think there is some degree to which we expect the F points to be named "lat" and "lon" in the input data. We should remove this assumption, instead referencing node_coordinates in the SGRID metadata.
Looking at
Parcels/src/parcels/_core/xgrid.py
Lines 176 to 183 in 08a0a26
I think there is some degree to which we expect the F points to be named "lat" and "lon" in the input data. We should remove this assumption, instead referencing
node_coordinatesin the SGRID metadata.