SGSCloud_Rad* routines not compatible with fewer radiation levels than (standard) model levels

[climbfuji]

In physics/module_SGSCloud_RadPre.F90, clouds1 is updated from cldfra_bl looping over horizontal_loop_extent and vertical_dimension. clouds1 however has dimensions horizontal_loop_extent and adjusted_vertical_layer_dimension_for_radiation - or better to say should have, because it is wrong in physics/module_SGSCloud_RadPre.meta.

There are possibly other such cases. To do:

• check all variables for their correct vertical dimension
• pass in adjusted_vertical_layer_dimension_for_radiation in addition to vertical_dimension
• loop over the correct vertical dimension as required

We need to do this for both module_SGSCloud_RadPre and module_SGSCloud_RadPost.

[climbfuji]

@dustinswales is this still an issue?

[joeolson42]

May I ask why both dimensions are needed? The variables in this module are also used in other modules where only the "vertical_dimension" is used. Either someone is over-thinking the problem or I am missing something...

[climbfuji]

It's common to have more or less levels for radiation than for the rest of the physics. NEPTUNE for example is using extra radiation layers on top (if I recall correctly), and the deep-atmosphere UFS configuration also has different numbers of levels for radiation and "the other physics".