lst_source_list = ['c2_corr', 'c2', 'low_latency']

'low latency' being the new option from Martha's group.

lst is then based off that.

c2_corr is default and results in:
lst = openet.core.common.landsat_c2_sr_lst_correct(sr_image, landsat.ndvi(prep_image))

lst_source_list = ['c2_corr', 'c2', 'low_latency']

'low latency' being the new option from Martha's group.

lst is then based off that.

c2_corr is default and results in:
lst = openet.core.common.landsat_c2_sr_lst_correct(sr_image, landsat.ndvi(prep_image))

Added a few more checks on the lst_source value, added an entry in the Image class docstring, and added code to apply a default scale_factor if one is not set.

Started adding support for falling back on the input LST image if the LST source image does not exist, but ended up not implementing this (for now).  If the user is expecting to use the LST source, it seems better to fail instead of returning an estimate based on the input LST, but that is not ideal when mapping over a collection.  We might want to instead set the LST to a masked image or something so nothing is computed but an image is generated.

Adding lst_source_id to generic image properties dictionary