Update data to use metric units and update the wind field to include scalar variables

[ian-noaa]

We'll need to create and run a back-filling script for two related issues. It would make sense to do these at the same time.

• We need to standardize the database and ingest to consistently utilize metric units throughout.
• We also need to make sure the wind fields store scalar (speed/direction) & vector (u/v) components in obs & model documents.

Potentially, we could update the Wind field to something like below. That way we could get the Wind object and be able to consistently ask for scalar & vector variables. Whether this is a good idea heavily depends on if it's a good idea in Couchbase.

"Wind": {
  "Vector": {
    "U": 1.234,
    "V": 5.678
  }
  "Scalar": {
    "speed": 1.23,
    "direction": 360
  }
}

[randytpierce]

I re-ran the backfill of METAR obs over the weekend to re-calculate the RH using DegF instead of DegC. The obs should all be good now, to my estimation. The backfill is single threaded and it took about 6 hours. That means that if we want to backfill the model data that we have - like to change the units or to calculate WindU and WindV this single threaded technique would take 2(models) * 48 (fcstLens) * 6(hours) => 576 hours which I think is probably a no go. So the model backfill will need to be either multi-threaded or broken up into a bunch of different containers that each process a subset of the data. I lean to the later myself. That would be easy enough. We just need to figure out where to run it. Probably on rancher.

[mollybsmith-noaa]

After thinking about this over the break, I would really prefer not to store the wind in a different structure than the rest of the variables, as it will make our templating way more difficult.

[ian-noaa]

#281 is a possible duplicate of this one.

[github-actions]

This issue is stale because it has been open 90 days with no activity.

[github-actions]

This issue is stale because it has been open 90 days with no activity.