Updating CLI and Acceptance Tests for non-default scale factors.

[mo-jbeaver]

Addresses #1151

This PR implements changes to the uv_index CLI and Acceptance Test.
This provides the option to use a non-default scale-factor via the CLI.

Testing:

• Ran tests and they passed OK
• Added new tests for the new feature(s)

[mo-kbogue]

I ran the acceptance test with no issues, though looking at the unit tests in improver_tests/uv_index/test_uv_index.py, I think that it may be worth adding some additional unit tests. There is already a unit test for non-default scale factor (in which 10 is used), but it might be worth taking other possibilities into account, such as including the case where a scale factor of 0 is received.

I would also like to have a better understanding and written record of where the 0.1 factor come from. From reading the previous tickets, it seems as is this possibly ported over from SPSS, but where did the 0.1 arise from in SPSS? If the justification is not clear, it may be worth revisiting if this is going to produce the best UV index for users. I do see that the jupyter notebook in the associated PR (#530) tests this 0.1 scaling and it does appear to give a reasonable answer, but is this enough of a scientific justification for its use? If we are confident in the 0.1 scaling factor, we should make sure that the reasons for this are clear and accessible.

[mo-jbeaver]

In reference to the 0.1 scaling factor, the comment linked below adds some more context.
https://github.com/metoppv/mo-blue-team/issues/1143#issuecomment-4904360262

[mo-kbogue]

Looks good to me, thanks for including that additional warning.

[mo-kbogue]

Just one small typo, but otherwise this is great - thanks for this!

[mo-kbogue]

Suggested change

	been empirically derived, or Met Office UM UV data, and should not be
	been empirically derived, for Met Office UM UV data, and should not be