Expand NetCDF test data (with groups/various levels)

[srizvi-NASA]

The current test data for converting netcdf to cog seems limited. To ensure thorough testing, I would like to request adding more netcdf files with the following variations:

1. Files with and without groups.
2. Files with different data levels (e.g., TEMPO granules from L2 or L3).

Including a wider variety of data will help identify and address potential issues during the conversion process.

[frankinspace]

Only concern here is the size of the test data files; we can't use git-lfs (technically we can but the free space we are given is shared across the entire podaac/ org and not very large, and there's too much red tape involved trying to purchase more LFS storage)

[frankinspace]

Another thought I've had is to setup a pytest fixture that could download granule CMR based on concept-id prior to test and then remove it after test