BDMS 122: lexicon updates & other updates

[jacob-a-brown]

Why

This PR addresses the following problem / context:

• The lexicon table should contain all NM_Aquifer lookup table values so that data can be transfered

How

Implementation summary - the following was changed / added / removed:

• Added values to core/lexicon.json

Notes

Any special considerations, workarounds, or follow-up work to note?

• Used full names except for organizations since they are commonly referred to with their abbreviations, like USGS
• The transfer scripts will need to be updated to use code/meaning mappings so the transferred data will be in the lexicon term table
• Added some verbiage to the README to help with future model revisions

[codecov-commenter]

Codecov Report

✅ All modified and coverable lines are covered by tests.

Files with missing lines	Coverage Δ
tests/conftest.py	100.00% <ø> (ø)
tests/test_observation.py	96.85% <ø> (ø)

... and 3 files with indirect coverage changes

[jirhiker]

I worry about transferring the formation and lithology fields without doing some cleanup.

For instance, is there a distinction between Sand, silt and gravel and Gravel, sand and silt, or would one record suffice? Doesn't the USGS have a controlled vocabulary for lithology?

Formation records seem a mess

[jacob-a-brown]

To facilitate the transfer of data we can keep all of the lookup table values as-is, then we can work to clean up those terms. Since a lexicon_term has onupdate="CASCADE" any changes should propagate to the appropriate places. I could also remove formation and lithology from the lexicon, but I'd be concerned that would inhibit data transfers. (If we wait to record lexicon terms for formation and lithology we could just wait to transfer all formation and lithology values until it's cleaned up and mapped to the new terms.)

[ksmuczynski]

Looks good. I agree that there are several lookup tables that could benefit from a clean up. Modifying the lexicon terms later doesn't seem like a big deal, but what about cases where the cleanup is significant and lexicon terms are removed (but the removed terms are referenced in other tables? Should we set some kind of an ON DELETE constraint or flag it in some other way? Is it important to figure out now or later?

[jacob-a-brown]

That's a good point, particularly if some are duplicative and we decide to just keep one (e.g. sand, silt, gravel and gravel, sand, silt).

So remove them for now?

[jirhiker]

Lets remove for now. We need AMP to review and cleanup. This could be something Kimball could help with