I have two data:

1. 3D data with z(height) levels
2. 3D pressure data which are at the same levels

Then, I want to interpolate the first 3D data to another 3D pressure data. I have to iterate through x, y and interpolate the data by pressure levels.

This is too slow when the 3D array is large. Is there a more efficient and easier method to accomplish this?

MCVE Code Sample

```python import xarray as xr import numpy as np

x = 10 y = 10 z = 4 data = np.arange(xyz).reshape((x, y, z))

bottom_up data

bottom_up = xr.DataArray(data, coords={'x': np.arange(x), 'y': np.arange(y), 'z': np.arange(z)}, dims=['x', 'y', 'z'] )

corresponding pressure data

pressure = xr.DataArray(data+1, coords={'x': np.arange(x), 'y': np.arange(y), 'z': np.arange(z)}, dims=['x', 'y', 'z'] )

pressure levels which data are interpolated to

interp_p = xr.DataArray(data[:, :, :-2]+0.5, coords={'x': np.arange(x), 'y': np.arange(y), 'z': np.arange(z-2)}, dims=['x', 'y', 'z'])

empty DataArray where interpolated values are saved

output = interp_p.copy(data=np.full_like(interp_p, np.nan))

iterate through x and y

for x in bottom_up.x: for y in bottom_up.y: # replace bottom_up with pressure bottom_up = bottom_up.assign_coords(z=(pressure[x, y, :]).values) # interpolate data to interpolated pressure levels output[x, y, :] = bottom_up[x, y, :].interp(z=interp_p[x, y, :], kwargs={'fill_value': 'extrapolate'} ) print(output) ```