Modifying min(dataarray) modifies dataarray

What happened:
Getting the minimum of a data array returns a reference to that minimum data point such that when modifying the minimum, the original data array is changed as well (same for max).

What you expected to happen:
Not sure if this is intended or a bug, but I expected an independent data array (the same as using dataarray.min()). This behaviour also differs from numpy (see below).

Minimal Complete Verifiable Example:

import xarray as xr
import numpy as np
x_arr = xr.DataArray([1,2,3])
min_x = min(x_arr)
min_x += 10
x_arr
# [11,2,3]

n_arr = np.array([1,2,3])
min_n = min(n_arr)
min_n += 10
n_arr
# [1,2,3]

Environment:

Start with the minimal Python example comparing built-in min(dataarray), DataArray.min(), and NumPy min behavior. Trace the entry point used by min(dataarray), then add a regression test showing that modifying the returned minimum or maximum does not modify the original DataArray; done means the original values remain unchanged.