Adding Integrated Canopy Effects to AQM

[drnimbusrain]

@BrianCurtis-NOAA @JianpingHuang-NOAA @KaiWang-NOAA

This PR adds the inline canopy photolysis effects to AQM/Online-CMAQ, and references issue #16 , which provides some more information on the canopy photolysis effect.

Note that this is only an option and does not change results from the base configuration. When turned on, this option also only affects grid cells that are satisfy contiguous canopy conditions in the code.

Canopy inputs are added to the aqm.rc like any other external input sources. Required variables are the forest canopy height (FCH), forest fraction (FRT), clumping index (CLU), population density (POPU), leaf area index from ECCC (LAIE), and four in-canopy level specific cumulative LAI fractions (C1R, C2R, C3R, and C4R). I have created these canopy inputs based on collaboration with ECCC. The differences are both AQM and CMAQ specific.

I have created the canopy input files for Online-CMAQ at the RRFS_CONUScompact_13km and NA13km domains here on Hera: /scratch2/NCEPDEV/naqfc/RRFS_CMAQ/canopy/

Here are the example canopy additions to the aqm.rc template that should be added and then populated based on updated sample configuration files in UFS-SRW-App/online-cmaq for EMCs testing.

canopy_yn: true

canopy_type: canopy

canopy_format: netcdf

canopy_file: {{ aqm_rc_canopy_file_fp }}

canopy_frequency: static

canopy_species::
FCH 1.00000 FCH m
FRT 1.00000 FRT 1
CLU 1.00000 CLU 1
POPU 1.00000 POPU 10000_people/10km2
LAIE 1.00000 LAIE 1
C1R 1.00000 C1R 1
C2R 1.00000 C2R 1
C3R 1.00000 C3R 1
C4R 1.00000 C4R 1
::

These effects have been tested in both the offline NAQFC ("NACC-CMAQ") and Online-CMAQ for the RRFS_CONUScompact_13km domain during August 2019 on Hera. Results show improved model performance and reduced ozone overpredictions in regions of contiguous canopies. There are negligible impacts on summer PM2.5. Please see my DSRA/EMC presentation for results.

Also, I have tested the canopy photolysis codes in the latest workflow with latest EMC/AQM codes, and ran the model for a 6-hr engineering test over the target larger NA13km domain on WCOSS2. Engineering tests of the photolysis effect over the NA domain seem OK:

@chan-hoo If/when this PR is accepted, can we rework the UFS-SRW-App to eventually remove the --canopy option and the canopy specific workflow changes/options as the integrated/default? Thank you.

[BrianCurtis-NOAA]

@drnimbusrain If there's still work ongoing, it might be helpful to change this to a draft PR until the work is ready.

[drnimbusrain]

@BrianCurtis-NOAA @JianpingHuang-NOAA With all else going on with AQMv8 we would like to move forward our first approach with integrated canopy effects for AQM now, and mark this as now ready for review. We show stable, and favorable results on ozone predictions with negligible computational impact with this approach.

Impacts on near surface AQM ozone for Aug 1-15, 2023.

Impacts on near surface meteorology for Aug 1-15, 2023 (from associated CCPP physics canopy impacts in other PRs, see below)

Reductions in near surface ozone bias against AirNow for Aug 1-15, 2023

Note that to achieve these impacts above, other PRs are being made to fv3atm/ccpp-physics to update the canopy diffusivity effects as an option in the PBL scheme used in both GFSv16 and GFSv17 (i.e., the SATMEDMF scheme). Also, there are changes here on AQM side that facilitate canopy data read (via aqm.rc) and transfer to fv3/ccpp-physics for usage there.

Please also note we will have updates to the codes in the future, which introduces an explicit sub-canopy 3 layer in AQM (as an option likely to this integrated effect), and anticipate larger impacts on ozone and more improved performance. However, we would like to move this PR with integrated approach forward now, as it would be too cumbersome to introduce all codes at once, and we need to get much of the other canopy code, data, and additional structure in to move forward first.

The updated canopy data needed as inputs to AQM for these effects are found on Hera, currently only for the 13 km and 9 km NA domain: /scratch2/NAGAPE/arl/Patrick.C.Campbell/RRFS_CMAQ/fix/canopy/AQM_NA_13km/gfs.t12z.geo.*.canopy_regrid.nc.
/scratch2/NAGAPE/arl/Patrick.C.Campbell/RRFS_CMAQ/fix/canopy/AQM_NA_9km/gfs.t12z.geo.*.canopy_regrid.nc
These canopy data are based on our recent publication here.

We also have associated PRs out for canopy data via FV3 and for integrated canopy effects on eddy diffusivity via ccpp-physics PBL schemes:
FV3: NOAA-EMC/fv3atm#928
CCPP: ufs-community/ccpp-physics#253

@iri01 @bbakernoaa

[bbakernoaa]

@drnimbusrain does PR #101 need to go in first for this PR?

[drnimbusrain]

@drnimbusrain does PR #101 need to go in first for this PR?

@bbakernoaa Yes, probably best they can merge #101 first, and then merge our canopy changes on top of that. There may be a few small conflicts to resolve, but no biggie.

Also FYI @JianpingHuang-NOAA @BrianCurtis-NOAA our initial integrated canopy approach (for associated canopy diffusivity effects) is moving forward for FV3/CCPP, e.g., ufs-community/ccpp-physics#253