Get the correct output for the CMIP data using TRACK

[sjavis]

In #62 we were able to run TRACK on the CMIP data but the output tracks are not correct. We need to find what parameters / inputs need to be fixed to give correct output.

[sjavis]

Possible issues are:

• Are we using both hemispheres correctly? Do we need to use a separate run on one hemisphere with a negative threshold.
• I don't think the method is filtering based on a warm core as suggested in the papers. Check if this is needed or if it is included in the TRACK software. Perhaps email Kevin Hodges.
• The threshold is using the value from the TRACK / pyTRACK indat files but I don't think this is the same as in the papers.