Source Temperature Capability

[ljlamarche]

Add capability to invert non-Maxwellian spectra, which requires determining the source temperature.

Methodology proposed by Jules:

1. Find E0 and Q with normal lookup method
2. Look at E0 right outside of arc
3. Temperature should remain constant as arc goes past, so find temperature outside of arc in "background drizzle" unaccelerated Maxwellian region
4. Use temperature to get new lookup tables with accelerated Maxwellian

This is probably tractable, but may be challenging with only three colors.