Change J-value diagnostic to be per photolysis equation

[lizziel]

Your name

Lizzie Lundgren

Your affiliation

Harvard University

Provide a clear and concise overview of the new feature requested.

This feature request is to update the J-values diagnostics to be a more comprehensive report of J-values used in photolysis within GEOS-Chem.

The way it currently works is to sum the J-values per species that are in photolysis configuration file FJX_j2j.dat. This is a sum because in some cases there are multiple photolysis reactions for the same species in that file. These J-values may be used for other species in fullchem.eqn but that usage would not appear in the diagnostics. It is also possible for some J-values computed and reported in diagnostics to not be used in GEOS-Chem at all.

There is general consensus amongst the Chemistry Working Group Co-Chairs that the most useful form of the diagnostic would be J-values per photolysis reaction in fullchem.eqn rather than FJX_j2j.dat. It should also include the mapping to the FJX_j2j.dat which itself includes mapping to FJX_spec.dat, enabling users to easily find additional information such as cross-sections used and scaling factor applied.

Given this goal, I suggest (1) set J-value diagnostic variable name to include both species name and an equation number, e.g. MEKeqn1, MEKeqn2, etc, and (2) set long name to include both the full equation and the FJX_j2j.dat ID used in fullchem.eqn. We can further remove the Is_Photolysis indicator in the species database since it is used only for J-value diagnostics and will no longer be needed given fullchem.eqn will be parsed directly.

Analysis of total per species, or relative contribution per equation for a given species, could be done in post-processing using the variable short name. Post-processing could also include extraction of scaling and other information in FJX_j2j.dat by using the long name. Title per page in the benchmark J-value files could be the long name (full equation plus j2j ID). Because those plots are listed alphabetically all equations for the same species would be consecutive in the file.

This update may be tricky to implement but will be worth it in the long run. Currently our J-value diagnostics would not catch errors in the mapping between photolysis equations and the equations in FJX_j2j.dat. The new method would be a better way to validate photolysis since the J-values will reflect what is actually applied in the model.

I am not applying a timeline to this feature request since GCST will do it as we have time.

Tagging Chemistry WG Co-chairs: @jhaskinsPhD, @barronh, @kelvinb, @luhu0

Reference(s)

No response

Will you be implementing this feature yourself?

Yes

Additional information

No response