Add FC/Cross Power Selection Methods

[kujaku11]

Could add some selection methods for removing noisy data.

• Coherency Thresholds based on
  • predicted electric fields from measured electric fields
  • measured and remote reference electric fields
  • predicted vertical magnetic fields from measured magnetic fields
• Polarization of measured fields

One option is to adapt tools in SigMT

[hlarniergeo]

Just a FYI regarding this topic, I started writing the BIRRP branch which will have coherency thresholding tools with min and max bounds. I found that max bounds are useful in sites with highly coherent noise. Not sure about my ETA for this, I want to get the estimator running first before starting the preprocessing stage, so if anyone has time to tackle it on the EMTF side, we can have a look at the performance of both methods and go from there.

[kkappler]

I have implemented a crude weighting scheme based on partial coherencies in aurora/transfer_function/weights/coherence_weights.py. This method loosely follows Jones & Jödicke, 1984.

It hasn't been widely tested yet though, and is not yet available as an option in the processing config. We could however use it for comparison and testing on a branch.

Reference: Jones, Alan G., and Hartmut Jödicke. "Magnetotelluric transfer function estimation improvement by a coherence-based rejection technique." SEG Technical Program Expanded Abstracts 1984. Society of Exploration Geophysicists, 1984. 51-55.