Add Modern Earth benchmark tutorial

[jaden-lopez]

Tutorial that Nick (@Nicholaswogan) and I have been working on that computes modern Earth spectra in reflected light and thermal emission that also recreates Robinson & Salvador 2023 comparing to rfast calculations. Also links to a Zenodo repository (not yet linked in notebook, awaiting edits, but currently here: https://github.com/jajadl/PICASO_spectra_comparison/tree/main) that can calculate reflected spectra for Venus and thermal emission spectra for Mars. Not yet ready to merge but wanted to start the process, any feedback appreciated.

[Nicholaswogan]

@jaden-lopez, we should remove the absolute paths in the notebook. So, for example, we do:

# We should do this
filename_db = os.path.join(jdi.__refdata__, 'opacities', 'opacities_photochem_0.1_250.0_R15000.db')

# Lets not do this
filename_db = '/some/absolute/path/opacities_photochem_0.1_250.0_R15000.db'

For the ICRCCM file "tyrobinson_earth_PT_and_mixing.atm", we should just embed those as simple numpy arrays in the notebook (i.e. no file).

The same goes for the files "modernearth_rfast.pkl" and "modernearth_rfast_thermal.pkl'". I think it is best to embed that information as arrays in the notebook.

[Nicholaswogan]

@jaden-lopez, the new code seems to have an undefined variable df_atmo_earth? Also, when the rfast spectra are interpolated to the picaso grid (or whatever is happening there), nans are produced.

Can you address both of these?

[Nicholaswogan]

@natashabatalha, The current notebook does a cloud-free reflected light spectrum, then compares that with a cloud-free spectrum computed with rfast (spectrum embeded in notebook).

Would it be preferable to just directly reproduce figures in Ty's paper? In other words, we would reproduce Figure 4 (right) and Figure 5 in the rfast paper. For Figure 4, right, this would require adopting Ty's special cloud treatment for Earth (some mixture of ice and water). I have worked out how to do this can can share with Jaden.