calculate vorticity using diff_2d_vector in LLC90 grid

[ruhuihuang]

When I try to calculate the vorticity of velocity of ECCOv4 (LLC90 data) using the following python codes:
zeta0=grid_llc.diff_2d_vector({'X': (data_uv.VVELMASS * data_uv.dyC), 'Y': (data_uv.UVELMASS * data_uv.dxC)},boundary='fill')
zeta2 = (zeta0['X'] -zeta0['Y']) / (data_uv.rA)
---------------------------------------------------------------------------
NotImplementedError Traceback (most recent call last)
Cell In[18], line 1
----> 1 zeta0=grid_llc.diff_2d_vector({'X': (data_uv.VVELMASS * data_uv.dyC), \
2 'Y': (data_uv.UVELMASS * data_uv.dxC)},\
3 boundary='fill')

File /g/data/hh5/public/apps/miniconda3/envs/analysis3-23.01/lib/python3.9/site-packages/xgcm/grid.py:2422, in Grid.diff_2d_vector(self, vector, **kwargs)
2403 def diff_2d_vector(self, vector, **kwargs):
2404 """
2405 Difference a 2D vector to the intermediate grid point. This method is
2406 only necessary for complex grid topologies.
(...)
2420 are differenced vector components along each axis
2421 """
-> 2422 return self._apply_vector_function(self.diff, vector, **kwargs)

File /g/data/hh5/public/apps/miniconda3/envs/analysis3-23.01/lib/python3.9/site-packages/xgcm/grid.py:2378, in Grid._apply_vector_function(self, function, vector, **kwargs)
2376 position, coord = axis._get_position_name(component)
2377 if position == "center":
-> 2378 raise NotImplementedError(
2379 "Only vector interpolation to cell "
2380 "center is implemented, but vector "
2381 "%s component is defined at center "
2382 "(dims: %r)" % (axis_name, component.dims)
2383 )
2385 x_axis_name, y_axis_name = list(vector)
2387 # apply for each component

NotImplementedError: Only vector interpolation to cell center is implemented, but vector X component is defined at center (dims: ('k', 'face', 'j_g', 'i'))

Any help will be appreciated.
Thanks you.

[ruhuihuang]

The test data can be down load from here
uv_test.zip

[ifenty]

@ruhuihuang

I think you need to multiply the uvelmass, vvelmass by the grid cell edge lengths, dyG and dxG, not the grid cell center distances, dxC, dyC.

'''zeta0=grid_llc.diff_2d_vector({'X': (data_uv.VVELMASS * data_uv.dyC), 'Y': (data_uv.UVELMASS * data_uv.dxC)},boundary='fill')
zeta2 = (zeta0['X'] -zeta0['Y']) / (data_uv.rA)

[ruhuihuang]

@ruhuihuang

I think you need to multiply the uvelmass, vvelmass by the grid cell edge lengths, dyG and dxG, not the grid cell center distances, dxC, dyC.

'''zeta0=grid_llc.diff_2d_vector({'X': (data_uv.VVELMASS * data_uv.dyC), 'Y': (data_uv.UVELMASS * data_uv.dxC)},boundary='fill') zeta2 = (zeta0['X'] -zeta0['Y']) / (data_uv.rA)

Thank you, but not this case.