feature update: instrument and satellite meta tables for arrays

[jrknezha]

This PR adjusts how satellite and instrument information is stored and related to array metrics and types. Previously, we had a single satellite meta down to the channel associated with a type, this was too restrictive. This PR splits this information into two tables sat_meta (sat id, name, short name) and instrument meta (name, num channels, scan angle) and associates them separately between the array metric type (instrument) and array metric values (specific sat).

This will extend our ability to use the satellite and instrument in flexible ways for arrays. In the future, these tables can also be associated with the experiment metric and metric types tables for single values.

Modified files for the new table structures and relationships, handler files and testing, and added new files for instrument meta handler and tests.

I also commented out the harvest innov stats tests which is a known failure and complicated testing procedures (this should be addressed in future PRs and has been captured as issue #38)

[jrknezha]

need to add the updated tables to the readme

[jrknezha]

ReadME has been updated and the PR is fully ready for review

[amschne]

Can reproduce passing unit tests. Noting here to also include in this PR a new namedtuple in src/harvest_translator.py that corresponds to the new expt_array_metrics table.

I provided a few minor (stylistic) comments to consider.

[jrknezha]

confirming at this point that all tests are running successfully with the new changes. since we don't have any harvesters set up for the array metrics, I can't write in a realistic test for the new harvest with is_array part

[KevinCounts]

All tests are passing for me, and everything else looks good. Double checked code and readme changes against the db schema changes as well. I think this is good to go.