Verify that Vermillion data set has a constant depth of 8 ft

[padilla410]

Two issues were identified in resolving unit conversion inconsistencies in data parsers. Two parsers need a secondary review to ensure that they have constant depths (as defined in their parsers):

• parse_URL_Temp_Logger_2006_to_2017
• parse_Lake_Vermilion_2016

Comments from PR #274:

the "8ft" must be in the explainer for this, but it seems odd to see this magic number in here without a comment. Again, this isn't a change your PR needs to make.

_Originally posted by @jread-usgs in https://github.com/USGS-R/lake-temperature-model-prep/pull/274#discussion_r781224662_

Since I don't see a comment in the function, just verifying here that this dataset only has data at 11ft depth?

_Originally posted by @jread-usgs in https://github.com/USGS-R/lake-temperature-model-prep/pull/274#discussion_r781222341_

A fully resolved issue would involve adding a clarifying comment into the relevant parsers.

[padilla410]

Hey @limnoliver, I just wanted to ping you on this issue before I resolve it. There are two sets of data parsers where the depth field is manually set:

• In parse_vermillion_files (example) the following files have some sort of manual depth setting:
  • Set to 0 ft depth:
    • Joes_Dock_2013
    • Joes_Dock_Logger_2012
  • Set to 8 ft depth:
    • Lake_Vermilion_2016
    • Logger_Temps_2009_Joes_Dock
    • Logger_Temps_2009_Open_Water
    • Logger_Temps_2010_Open_Water
    • Logger_Temps_2011_Open_Water
    • Open_Water_Logger_2013
    • Temp_Logger_Data_2015
    • Vermilion_Logger_2014
• In parse_mndow_coop_files - parse_URL_Temp_Logger_2006_to_2007 has depth set to 11 ft

I took a look at all of the files listed above and the hardcoded depths seem reasonable (i.e., raw data didn't come with a depth field). Is there an old email floating around somewhere that documents the hardcoded depths? I would like to add a comment to the parser reflecting this, if possible.

[limnoliver]

Hi Julie-

Found a quick answer for the URL_Temp data, which is that the cooperator provided the depth in the explainer file.

Will do some more poking for Vermillion...

[limnoliver]

Looks like the Vermillion munge predates me (e.g., bulk add of parser files here). My best guess is that @jread-usgs might have the email that is referenced here. Saying that because it looks like a lot of though/investigation went into Vermillion specifically.

[padilla410]

Thanks for the quick turn around, @limnoliver -- really appreciate it! I'll follow up with @jread-usgs.

[jordansread]

Looks like those did indeed predate Sam's creation of the coop parser parts of the pipeline. The files were from a prior repo that aided earlier versions of this work: https://github.com/USGS-R/pgml_temperature_prediction/tree/master/1_all_raw_data/in

Those parsers came from here and we may need to check with David too (I am seeing if I can track down the email in question)

[padilla410]

Jordan was able to track down a 2018 email on the Vermilion data. I have added an additional explainer file documenting it.

I will close this issue when I add a PR with a few comment updates after wrapping up the Univ of MO data this morning.

Thanks all for running down these files.

[padilla410]

I am closing this issue. It looks like the depth of "8 ft" was provided in an email. As I mentioned earlier in the thread, I added an "explainer" file to the shared google drive.