Add sedimentation of rhoa, q_tot and moisture tracers to edmf

[trontrytel]

This PR adds the sedimentation terms due to precipitation to rhoa, q_tot and other moisture tracers and enthalpy. Here are the derivation notes.

TODO:

• Add the terms to h advection
• Debug snow in RICO
• Test with TRMM

Next steps

• Add more lazy expressions. For example for the w of all the moisture tracers
• Loop over tracers in a more concise way
• Add the one missing term that is currently making the simulations unstable (or understand why we don't want it)
• Understand what terms need to be added to implicit solver
• Add additional detrainment like term to move mass from updrafts to the environment to ensure that the environmental values don't go too negative from advection. As is, right now, we don't have any means to ensure positivity in the environment while solving prognostic advection on the grid mean and updraft.

DyCOMS
summary.pdf

RICO
summary.pdf

TRMM
summary.pdf

[tapios]

Basically looks good to me, modulo the minor comments I left (the most important among them may be to use upwinding also for the log(\hat rho) derivative). I also checked the derivations in your notes. They look correct.

[tapios]

Can we write the following block (and the preceding one above) as a loop of condensate species, so we do not need to repeat the same pattern so many times? (And can also add other condensate species, e.g., graupel, more easily later if we really want to.)

[trontrytel]

Yes, I'll do it after I'm done with the PR. It's easier for me to debug when everything is written out like this. But I'll turn it into a concise loop over tracers and velocities for better readability afterwards