Release v0.7.0: encapsulated calculations from Neal et al (2022)

Major changes in user interface

• Encapsulated the latitudinal boundary conditions and inversion algorithm used in Neal et al (2022, GRL) to QGFieldNHN22. The existing algorithm from Nakamura and Huang (2018, Science) is encapsulated in QGFieldNH18. Current QGField class is equivalent to QGFieldNH18 (for backward compatibility). These classes are also available in the xarray interface QGDataset (Thanks Christopher Polster!).
• Updated interfaces of F2PY modules to remove redundant arguments. Please recompile F2PY modules following instructions from README.md.
• Significant improvement in Documentation. In particular, with the use of nbsphinx, jupyter notebook examples are now available in the documentation. Great thanks to Christopher Polster for the enhancement.
• All jupyter notebook examples are fixed. Notebook examples are available for both QGFieldNH18 and QGFieldNHN22. They can be run with the sample data available in Dropbox folder (See Quick start in readme.md for details).
• Added new unit tests which enhance test coverage.

Migration guide from v0.6.x -> 0.7.0

• Now northern_hemisphere_results_only becomes an input argument for QGField. Please initialize this when creating QGField object and remove this argument from the methods (compute_reference_states and compute_lwa_and_barotropic_fluxes).
• eq_boundary_index is now an input arguemnt merely for QGFieldNHN22, but not QGFieldNH18. Remove it if you are using QGField/QGFieldNH18. You can refer to the jupyter notebook examples for the updated usage.