Buoyancy Fix & WRFInput Rebalance

[AMLattanzi]

This PR completes two things.

1. It adds a flag to the WRFInput initialization that allows users to rebalance HSE given Theta_d & q_v. This uses the Newton iters function to obtain rho_d rather than taking from the wrfinput file.
2. It corrects issues in the make buoyancy routine

The mods to make buoyancy are based upon the following derivation (this is for using \theta_d)

A similar derivation may be obtained when using T directly and that essentially differs by the 1/\gamma on the P'/P term. In all buoyancy routines we have neglected the P'/P term so that was continued here.