Full moist physics

[sjboeing]

Following the same framework as saturation-adjustment MONC, i.e. separate parcel attributes for liquid water and water vapour, a (reference) pressure profile, and a saturation mixing ratio which depends on pressure and temperature, rather than height.

[sjboeing]

I think the saturation adjustment can be done more efficiently than in MONC. There is some trade-off between efficiency and readability though, and maybe the cost is relatively small compared to the ellipsoid operations.

[sjboeing]

• Choice of variables: theta or theta-theta_00 (could be 273.15 K), q_l, q_v for maximum flexibility (for 2 moment-scheme: also N_l). Buoyancy still as buoyancy.
• Need to determine reference pressure profiles, either prescribed or calculated initially
• Need to "port" test cases for new thermodynamics
• Only implement for 3D for now?
• Saturation adjustment. First check if unsaturated. If not: Ridder's method with current theta as first guess (can we do this?) and all-saturated and all-unsaturated as bounds.