Port abstract aerosol interface + BAM implementation from CAM to CAM-SIMA

[jimmielin]

What is the feature/what would you like to discuss?

Bring in the abstract aerosol interface code from CAM into CAM-SIMA

This is code to be ported into the host model side. atmos_phys companion issue to follow.

Anything interfacing with physics will have quantities provided via the CCPP framework to the physics

List of modules to port (current CAM src/chemistry/aerosol)

• aerosol_properties_mod (provides abstract :: aerosol_properties & aero_name_len)
• aerosol_state_mod (provides abstract :: aerosol_state & ptr2d_t)
• aerosol_optics_mod (provides abstract :: aerosol_optics for sw/lw props)

After these abstract types are brought in (hopefully copy-pasted so they do not diverge from CAM) bring in bulk aerosol

Bulk aerosol modules

• bulk_aerosol_properties_mod (provides bulk_aerosol_properties)
• bulk_aerosol_state_mod (provides bulk_aerosol_state)
  aerosol state depends on pbuf in CAM (when option is N -- need to change to constituents object. I think this means there is no distinction between A and N anymore... to check)

Wire them up to the CCPP'ized BAM (bulk aerosol model)

Then CCPPize the aerosol optics module specific implementations

Aerosol optics specifics

• Prep work: rad_constituents and phys_prop (wip in atmos_phys)
• Port whatever is needed from aerosol_optics_cam -- this may or may not be on the host model side or atmos_phys side
• aer_rad_props wrapper to call aerosol_optics_cam (may not be needed)
• aerosol_optics_cam_sw / aerosol_optics_cam_lw logic needs to be extracted and put into abstract interface
• ... all this in CAM feeds back to rrtmgp_set_aer_lw/rrtmgp_set_aer_sw -- which calls aer_rad_props

call stack in CAM for rrtmgp <-> aer optics

rrtmgp_inputs_cam.F90:rrtmgp_set_aer_lw()
  └─> aer_rad_props.F90:aer_rad_props_lw(icall, state, pbuf, aer_lw_abs)
      └─> aerosol_optics_cam.F90:aerosol_optics_cam_lw(list_idx, state, pbuf, odap_aer)
          ├─> bulk_aerosol_state_mod.F90:bulk_aerosol_state(state, pbuf)
          ├─> bulk_aerosol_properties_mod.F90:bulk_aerosol_properties()
          └─> For each aerosol bin:
              ├─> aeroprops%optics_params() → rad_cnst_get_aer_props()
              │   └─> Returns: opticstype, lookup tables, density, hygro, etc.
              ├─> Create aerosol_optics object based on opticstype:
              │   ├─> hygroscopic_aerosol_optics()
              │   ├─> insoluble_aerosol_optics()
              │   └─> etc.
              ├─> aerostate%get_ambient_mmr() → rad_cnst_get_aer_mmr()
              │   └─> Returns: mmr pointer to state%q or pbuf field
              └─> aero_optics%lw_props(ncol, ilev, iwav, pabs)
                  └─> Returns: pabs (absorption coefficient)

rrtmgp_inputs_cam.F90:rrtmgp_set_aer_sw()
  └─> aer_rad_props.F90:aer_rad_props_sw(icall, state, pbuf, ...)
      └─> aerosol_optics_cam.F90:aerosol_optics_cam_sw(list_idx, state, pbuf, ...)
          ├─> [Same initialization as LW]
          └─> For each aerosol bin:
              └─> aero_optics%sw_props(ncol, ilev, iwav, pext, pabs, palb, pasm)
                  └─> Returns: extinction, absorption, SSA, asymmetry

Is there anyone in particular you want to be part of this conversation?

@peverwhee @nusbaume

Will this change (regression test) answers?

I Don't Know

Will you be implementing this enhancement yourself?

Yes, but I will need some help