Set up AR6 Constrained Sample Monte Carlo

[FrankErrickson]

Add an option to run a Monte Carlo that samples the constrained AR6 parameter samples for FAIR v1.6.2.

The parameters can be downloaded here as a JSON file: AR6 Constrained Parameters

This will likely require adding in new scaling parameters to the individual radiative forcing components (depending on which uncertainty AR6 samples).

[FrankErrickson]

This was mostly completed by 76e6e3eeb32f09191351a514974de223be72114c. Keeping this issue open to track some details with the Monte Carlo as the emerge.

AR6 Monte Carlo Details

• A number of the parameters included in the AR6 Constrained Parameters file did not vary across samples (i.e. they were in the sample data set, but were all the same). These included (1) Volcanic radiative forcing time series F_volcanic, (2) coefficient on forcing from stratospheric water vapor due to methane oxidation stwf_from_ch4, (3) reference emission value for black carbon E_ref_BC, (4) natural emissions time series for methane and nitrous oxide natural, (5) pre-industrial emission values E_pi, (6) reference forcing value for black carbon F_ref_BC, and (7) the coefficient for landuse forcing aCO2land.
• Radiative forcing from contrails was set to 0 in all periods. This is the default setting in the Python version of FAIR (i.e. I did not see a specific setting for this in the AR6 code).

AR6 Monte Carlo Remaining Questions, Tasks, & Issues

• Compare Julia Monte Carlo output to Python AR6 Monte Carlo output (might need to carry out all of the Python runs in AR6 repository to get this data)