[main] reduce SBN/AWIPS product volume

[MatthewPyle-NOAA]

We can generate roughly 15-20 GB/day for both RRFS and REFS.

A single long deterministic cycle currently is producing about 173 GB by itself.

Have started jotting down thoughts about how this volume could be reduced here.

[BenjaminBlake-NOAA]

@MatthewPyle-NOAA Thanks for opening this issue and putting together some initial thoughts. I will add some comments to the Google Doc and can assist with updating the parm/wmo files accordingly.

[BenjaminBlake-NOAA]

With the initial changes to the isobaric output (1000, 925, 850, 700, 500, 400, 300, 250, 200), removing RH, and limiting the files to 3-hourly, a single long deterministic cycle generates roughly 38 GB. This is still too high if we can only generate 15-20 GB per day, but it gets us a lot closer to where we need to be. Will discuss further in tomorrow's meeting with the SBNRC.

[BenjaminBlake-NOAA]

The initial changes for this issue are now included with PR #1039 - they were tested for a single long deterministic cycle on Cactus.