Add new models: multi-sector, tracer, isotope

[brendan-m-murphy]

In order of complexity, we would like to extend RHIME to include:

1. multi-sector model: inversions with multiple flux sectors
2. additional tracer which is linear with respect to these models e.g. ethane+methane
   • ratio between gases is uncertain, adds uncertainty to sensitivity in model
   • e.g. eq. 5 and 6 here: https://acp.copernicus.org/articles/22/3911/2022/ (slightly more generic, because here, we could assume zero emissions for one tracer for one sector)
   • see also, Alice's thesis
3. relationship between emissions and species are non-linear e.g. isotope data
   • in this case, some of the observed data are ratios of quantities calculated via (sensitivity) x (flux)
   • Alice wrote code to do this Adding multi gas model #24, but it uses a custom MCMC algorithm. Ideally we would use PyMC
   • Regardless of how we sample, Adding multi gas model #24 contains a lot of useful code for dealing with ratios of isotopes
   • there is a non-linearity introduced since the observed "delta" values are related to the fluxes via ratios.

A potential issue will be specifying these models in a configuration file. We should allow for any number of species and any number of sectors. When using tracers and isotopes, parts of the forward model are "relative" to other parts of the forward model, and we need to be able to specify what species are measured as ratios of another species. We also need to be able to specify when the non-linearity for delta quantities need to be used.