Fix issue 203a

[kujaku11]

Updating how metadata is translated from MTH5 to TF.

• updated to use only a single metadata object survey_metadata pulled from a run_ts object.
  • Added logic to get the survey ID from KernelDataset.df
  • Removed older logic to get station metadata from MTH5
• Updated tests/synthetic/test_processing.py to output a z-file directly from TF object
• TODO: Pack decimation level, number of samples, and band indicies into TF for a more accurate and complete representation of the transfer function estimates, and output to a z-file.

[kujaku11]

@kkappler I update mt_metadata which fixed some of the failed tests. The ones failing now are:

• docs test
• RMS tests

I didn't pull from FC's. So maybe once you've cleaned up your FC branch, that can be pulled into this branch and then complete the pull requests if tests pass.

[kkappler]

@kujaku11 I manually merged your tests into fourier_coefficients on fix_issue_203a_manual_merge and tests pass (except docs) on 3.8. I am abandoning this branch and will merge fix_issue_203a_manual_merge into fourier_coefficients