search

geoschem / geos-chem

GEOS-Chem "Science Codebase" repository. Contains GEOS-Chem science routines, run directory generation scripts, and interface code. This repository is used as a submodule within the GCClassic and GCHP wrappers, as well as in other modeling contexts (external ESMs).
http://geos-chem.org
Other
162 stars 155 forks source link

Feature request: Harmonize aerosol optics look-up tables used as inputs to photolysis and RRTMG #2214

Open yantosca opened 3 months ago

yantosca commented 3 months ago

Name and Institution (Required)

Name: Bob Yantosca Institution: Harvard + GCST

Confirm you have reviewed the following documentation

New GEOS-Chem feature or discussion

I am creating this issue as a placeholder so we don't lose sight of it. There is currently a disconnect in the optical properties uses by the photolysis schemes (FAST-JX and Cloud-J) and the radiative transfer model (RRTMG).

@lizziel wrote:

  • Cloud-J does not use the so4.dat and org.dat LUTs [look-up tables]. It uses a single file for all aerosol properties and does not have a column for humidity. I have discussed this with Michael Prather and he is open to working with someone on this.

  • Cloud-J takes concentration rather than optical depth as input. To prepare inputs to Cloud-J I had to write computation to get interpolated effective radius and extinction per grid box, and use it to compute wet concentration for aerosols undergoing hygroscopic growth. The optical properties used are not from so4.dat and org.dat. Some care would need to be taken to exten changes in PR #1902 to preparing the inputs for Cloud-J since it is not a straightforward translation.

  • I think consolidating the RRTMG LUTs (e.g. so4.dat and org.dat) with the dat file used in Cloud-J (formerly jv_spec_mie.dat in Fast-JX) would be good to do at the same time as this work. Fast-JX originally used jv_spec_mie.dat, and then someone switched the GEOS-Chem code to use the RRTMG LUTs in some places, hence the complications now. The tables are supposed to contain duplicate data but it is not obvious they do. It would be great to have an aerosols person advise on this. Consolidating the LUTs would avoid more headaches in the future.

Also it might be nice to see if we could adopt a common file format (such as YAML) for aerosol look-up tables, which would make the configuration files more clear and readable.

Tagging @randallvmartin @djxjacob @beckyalexander @theloniuspunk @wporter

pratherUCI commented 3 months ago

Agreed. It would be great to agree on a standard table format.

First let's agree on what optical properties we need and make that a standard record.

Second, we need to establish the degrees of freedom (does each aerosol type that is transported (i.e., kg per cell) have unique optical properties (unlikely) or depend on added water? or history of water? Does GEOS Chem keep track of hysteresis (is the RH going up or down?)

For Cloud-J my interest is primarily in the first standard: for the std wavelength bins Fast-J needs: m^2/kg (extinction optical size) or else m^2/kg where m^2 = integral pi*r^2 and Q = ratio of optical area to real area. SSA = pomega-0 = single scat albedo P1-P7 = terms in Legendre expansion of scattering phase function. That is all.

lizziel commented 3 months ago

Here are few links for easy reference to tables. I chose SO4 here for simplicity: