issues: 860038564

Is your feature request related to a problem? Please describe. Currently .clip can either take a scalar or a identically dimensioned array. numpy allows for broadcasting, which we can do in xarray too:

```python

In [11]: da = xr.DataArray(np.arange(24).reshape(2,4,3), dims=['a','b','c']) ...: da Out[11]: <xarray.DataArray (a: 2, b: 4, c: 3)> array([[[ 0, 1, 2], [ 3, 4, 5], [ 6, 7, 8], [ 9, 10, 11]],

   [[12, 13, 14],
    [15, 16, 17],
    [18, 19, 20],
    [21, 22, 23]]])

Dimensions without coordinates: a, b, c

In [12]: da.clip(da.mean('a'))

ValueError Traceback (most recent call last) <ipython-input-12-c8b9c6d0e0a1> in <module> ----> 1 da.clip(da.mean('a'))

[...]

~/.asdf/installs/python/3.8.8/lib/python3.8/site-packages/xarray/core/computation.py in apply_variable_ufunc(func, signature, exclude_dims, dask, output_dtypes, vectorize, keep_attrs, dask_gufunc_kwargs, *args) 741 data = as_compatible_data(data) 742 if data.ndim != len(dims): --> 743 raise ValueError( 744 "applied function returned data with unexpected " 745 f"number of dimensions. Received {data.ndim} dimension(s) but "

ValueError: applied function returned data with unexpected number of dimensions. Received 3 dimension(s) but expected 2 dimensions with names: ('b', 'c') ```

Adding in a broadcast_like currently allows it to work, but we can do the equivalent internally:

```python In [20]: da.clip(da.mean('a').broadcast_like(da)) Out[20]: <xarray.DataArray (a: 2, b: 4, c: 3)> array([[[ 6., 7., 8.], [ 9., 10., 11.], [12., 13., 14.], [15., 16., 17.]],

   [[12., 13., 14.],
    [15., 16., 17.],
    [18., 19., 20.],
    [21., 22., 23.]]])

Dimensions without coordinates: a, b, c ```

Additional context Numpy broadcasts the argument: https://numpy.org/doc/stable/reference/generated/numpy.clip.html