Deprecate auroras TTFZ - use mt_metadata.TF instead

[kkappler]

This is not trivial because many tests use the TTFZ for validation.

TTFZ is used for apparent_resistivity calculations which are cited in tests.
So we need

• An apparent resistivity calculator that interfaces with transfer function as a standalone method
• TTFZ extends aurora/transfer_function.base.TransferFunction so we would need to add the mt_metadata there as well

This is a pretty tricky task. Probably best to start by making TransferFunction base use mt_metadata.TF first.
There is a lot of duplication between auroras TransferFunction base and mt_metadata TF.

One way to do this would be to start by

• adding a to_mt_metadata() method to TransferFunction base
• store a copy of the mt_metadata object in TransferFunction
• then after the arrays are initialized, call to_mt_metadata() so that the self.mt_metadata_tf is updated
• then, update both the original arrays (to keep tests passing) and set self.mt_metadata_tf whenever the tf value is set or updated.
• Then replace self.tf with something that references the mt_metadata object
• By that point, the mt_metadata object is doing everything, and the arrays are only used for setting values
• Then try removing the setting of the original arrays and see of tests still pass

[kkappler]

• A Along the way, we should simplify the way that metadata is passed to the TF object. Will create a branch to attempt to pack the survey_metadata into the KernelDataset, during initialize_dataframe_for_processing.

[kkappler]

Apparent resistivity does not exist in mt_metadata. The aurora method should move to sandbox until we are integrated with MTPy