radiation: fix broken year warnings

[brianhenn]

The radiation port, if run for certain years, had warning statements that inadvertently crashed it. This PR replaces them with print statements (the way the rest of the port outputs to console at runtime; should probably be replaced by logging throughout but that's a separate issue).

Coverage reports (updated automatically):

[brianhenn]

Fair question. Yes I think the intent is to get the year with same position in the solar cycle as the desired year, but we could also ask GFDL for an updated solar constant data file.

[spencerkclark]

Last month I happened to add an updated set of forcing files to gs://vcm-fv3config/data/base_forcing/v1.2. Other than the following updates, all other files are identical to their previous versions:

• The addition of CO2 for years prior to 2010 and after 2016, and the update of the 2015 and 2016 CO2 data files to reflect observations rather than projections.
• The addition of volcanic aerosol forcing data (only used prior to 1999).
• An update to the solar constant file to include values through 2033.

Indeed it looks like this code is complaining about the solar constant data, so it seems like this update should fix it!

[brianhenn]

Cool thanks @spencerkclark . Will be easy to update the GCS file that the rad port pulls its solar data from (which is just a reformatted copy of the original forcing files).