Add mechanism to control image interpolation

[nfahlgren]

I think in older versions of matplotlib the default method for interpolation (which occurs when images are downsized for display in Jupyter or when plots like pcv.visualize.pseudocolor are created) was None or nearest neighbor but now is "antialiased" https://matplotlib.org/stable/gallery/images_contours_and_fields/interpolation_methods.html

When downscaling causes visualization problems, it would be useful for a user to control the interpolation method and/or for us to set a default other than antialiased potentially. I think there are three potential solutions we could consider.

Option 1: Document setting the method using matplotlib

import matplotlib
matplotlib.rcParams["image.interpolation"] = "nearest"  # or None, etc.

Option 2: Add a configuration option interpolation to the Params class

Option 3: Use **kwargs in pcv.plot_image and pcv.print_image for matplotlib plots in addition to xarray DataArray plots