Updating AQM to CMAQv5.4

[drnimbusrain]

PR Checklist

• This PR has been tested on an RDHPCS machine and/or WCOSS2. Please select below:

  • [X ] RDHPCS.
  • [ X] WCOSS2. (Issues with output total PM2.5 fractions and calculation, equal to zero only on WCOSS2) Issue has been made on UWM by Jianping: The updated AQM with CMAQ5.4 in the UFS Weather Model predicts zero PM2.5 concentrations on WCOSS2 [AQMv8] ufs-community/ufs-weather-model#2378

• [X ] This PR has been tested with the ufs-srweather-app workflow [aqm_dev] and [develop] branch.

• [X ] New or updated input data is required by this PR.
  To test this PR with CMAQv5.4 in AQMv7, you will need the following:

• Updated aqm.rc template file that has updated mechanism (CB6r5AE7) and some other updates.

• Updated diag_table_aqm.FV3_GFS_v6 and field_table_aqm.FV3_GFS_v16 to accommodate CMAQv5.4 changes.

• [ X] Baselines are expected to change.

Description

This PR provides all necessary updates to AQM develop branch for ARL's work to update CMAQv5.4 submodule (from current CMAQv5.2.1) and all necessary AQM infrastructure to accommodate the CMAQv5.4 changes compared. Since our CMAQv5.4 changes were built upon our fork of the EMC develop branch, then it also contains some initial developments of our in-canopy photolysis infrastructure and code developments. The canopy option is turned off by default, and at this time, we recommend that the canopy option remains turned OFF (in aqm.rc) until further refined developments of the canopy codes in AQM (and later in UFS-SRW-App, UWM, FV3/CCPP, etc. via separate corresponding PRs to their authoritative repos to enable full functionality).

Issue(s) addressed

#100

Dependencies

Updates to UFS-SRW-App at non-default branches (e.g., production/AQM.v7) are needed to accommodate the CMAQv5.4 updates in AQM (see new or updated data section above). The question remains as to where and how to make these PRs. Our changes for CMAQv5.4 were built upon the previous [aqm_dev] branch of UFS-SRW-App, and do not correspond to the production/AQM.v7 branch. Please advise and we can update and make the correct PR to the correct UFS-SRW-App branch to accommodate testing the CMAQv5.4 updates to AQM.

@lwcugb @bbakernoaa @ytangnoaa

[ytangnoaa]

This supersedes #98 Could you change line 23 of CMakeLists.txt to: set(CMAKE_Fortran_FLAGS_DEBUG "-O0 -g -check all -check noarg_temp_created -fpe0 -fno-alias -ftrapuv -traceback")

Updated the compiler flag.

[BrianCurtis-NOAA]

If you could test a build it would save me some time later today, but there might be some tweaks to some of the code to conform to those options.

[JianpingHuang-NOAA]

@BrianCurtis-NOAA @bbakernoaa @drnimbusrain Is this PR ready to be merged ?

[bbakernoaa]

@JianpingHuang-NOAA @BrianCurtis-NOAA it looks like we still need to add a test for this. Could you help with this @JianpingHuang-NOAA?

[BrianCurtis-NOAA]

@JianpingHuang-NOAA @BrianCurtis-NOAA it looks like we still need to add a test for this. Could you help with this @JianpingHuang-NOAA?

We need a test, yes. It also needs to be one that can run quickly. C96 preferrably, but 192 is acceptable if it makes things easier. The test needs to run well within a 30 minute wall clock. It needs to test as many features as it can reasonably. If there are other features that need testing that can't be added to one test, then two can be acceptable if necessary.

[BrianCurtis-NOAA]

@JianpingHuang-NOAA @BrianCurtis-NOAA it looks like we still need to add a test for this. Could you help with this @JianpingHuang-NOAA?

We need a test, yes. It also needs to be one that can run quickly. C96 preferrably, but 192 is acceptable if it makes things easier. The test needs to run well within a 30 minute wall clock. It needs to test as many features as it can reasonably. If there are other features that need testing that can't be added to one test, then two can be acceptable if necessary.

I hope that eventually the workflow would be able to generate a run directory suited for regression testing so that future adjustments to the UFSWM regression tests are much easier.

[JianpingHuang-NOAA]

@bbakernoaa, https://github.com/noaa-oar-arl/AQM/tree/feature/cmaq54 is 3-month old. Where is the updated package?

[bbakernoaa]

this is the latest that we have. We don't have anything newer at this juncture

[JianpingHuang-NOAA]

I heard that ARL has a fix to the soil type mix-up. Is that not for AQM?

[bbakernoaa]

This is something that needs to be changed not just on the AQM side but also the FV3ATM side as well. We just finished some preliminary testing and such today. We will need to do PRs back to the UFS weather model for both changes (ie AQM and FV3ATM repositories)

[bbakernoaa]

note that when we update to the GFS v17 physics we should also pass the fractional land use over and use this instead of the dominant value. Currently we are basically hard coding the fractional land use to be 1 for each dominate land use type in the grid cell and all other values are 0.

[drnimbusrain]

I agree with Barry. We should handle this AQM/CMAQ54 update here, and leave other updates including physics, bug fixes, etc in other separate PRs as suitable.

…

On Fri, Jan 3, 2025, 2:04 PM Barry Baker ***@***.***> wrote: note that when we update to the GFS v17 physics we should also pass the fractional land use over and use this instead of the dominant value. Currently we are basically hard coding the fractional land use to be 1 for each dominate land use type in the grid cell and all other values are 0. — Reply to this email directly, view it on GitHub <#101 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AGLFYNR5BF7MX6PJQOLMMDL2I3NMRAVCNFSM6AAAAABLA6CBLWVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDKNRZGY4DOMRZGE> . You are receiving this because you were mentioned.Message ID: ***@***.***>

[drnimbusrain]

@BrianCurtis-NOAA How's it going moving this CMAQ54 PR forward, and is there a status update on a regression test? Do you need help from us to identify inputs, run directory, etc. to help get this setup and moved forward?
@JianpingHuang-NOAA

[BrianCurtis-NOAA]

@drnimbusrain We will need C96 support for AQM to get a proper regression test (C192 is acceptable if C96 complicates things)

I don't need a workflow to get this merged, but a run directory with a working UFSWM will suffice so I can update the RT system to use it. As long as we have that, we can get this part merged in. Is this something that you might be able to put together?

[bbakernoaa]

@BrianCurtis-NOAA could we not leverage: https://github.com/ufs-community/ufs-weather-model/blob/develop/tests/tests/regional_atmaq

We would just need to update the additional CMAQ related inputs and such for the updates no?

[BrianCurtis-NOAA]

@BrianCurtis-NOAA could we not leverage: https://github.com/ufs-community/ufs-weather-model/blob/develop/tests/tests/regional_atmaq

We would just need to update the additional CMAQ related inputs and such for the updates no?

Yes, that can be used as a template.

[BrianCurtis-NOAA]

I'm happy to put the work in taking a run directory and turning it into a proper RT if that will save you from too much work. The work on AQM-Workflow will eventually know how to generate such a run directory, but that will be a while.

[drnimbusrain]

Thanks Brian. That link to regional_atmaq RT seems pretty old, yes? So, to confirm you need a C96 run directory with AQM CMAQ54 updates that you will use as a template for RT? We mainly have been running C793.

…

On Fri, Jan 17, 2025, 4:01 PM Brian Curtis ***@***.***> wrote: I'm happy to put the work in taking a run directory and turning it into a proper RT if that will save you from too much work. The work on AQM-Workflow will eventually know how to generate such a run directory, but that will be a while. — Reply to this email directly, view it on GitHub <#101 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AGLFYNSO2MYVIQ6VIYAA3CL2LFVSPAVCNFSM6AAAAABLA6CBLWVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDKOJZGIYDCOBUHE> . You are receiving this because you were mentioned.Message ID: ***@***.***>

[bbakernoaa]

@drnimbusrain I think all we need to do is provide a run directory. The truth is most of the input files as far as emissions and such can be used for either 5.2 or 5.4. We just need to make sure we have an example to show the differences in the ancillary CMAQ files needed. Other than that the input files may need updated on the ATM side to accommodate the updates of the newer ATM hash.

I can take a quick stab at it on Tuesday morning. At least to get something running. Point me to a run directory that you have for whatever resolution. Then @BrianCurtis-NOAA you could then integrate the updated files that RT needs into run RT workflow.

[drnimbusrain]

Great news, thanks Barry. I'll send a run directory we have soon.

…

On Fri, Jan 17, 2025, 9:43 PM Barry Baker ***@***.***> wrote: @drnimbusrain <https://github.com/drnimbusrain> I think all we need to do is provide a run directory. The truth is most of the input files as far as emissions and such can be used for either 5.2 or 5.4. We just need to make sure we have an example to show the differences in the ancillary CMAQ files needed. Other than that the input files may need updated on the ATM side to accommodate the updates of the newer ATM hash. I can take a quick stab at it on Tuesday morning. At least to get something running. Point me to a run directory that you have for whatever resolution. Then @BrianCurtis-NOAA <https://github.com/BrianCurtis-NOAA> you could then integrate the updated files that RT needs into run RT workflow. — Reply to this email directly, view it on GitHub <#101 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AGLFYNSMACKRC4MCPTKEC532LG5TJAVCNFSM6AAAAABLA6CBLWVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDKOJZGQ4DGNZVG4> . You are receiving this because you were mentioned.Message ID: ***@***.***>

[bbakernoaa]

@BrianCurtis-NOAA @drnimbusrain please take a look here: /scratch1/NCEPDEV/stmp2/Barry.Baker/FV3_RT/rt_994030/regional_atmaq_debug_intel

I was able to get the RT for the regional_atmaq_debug going on hera with the updated feature/cmaq54 branch of the AQM. Note that this was off the latest develop branch of the UFSWM

Files Changed

AE_cb6r3_ae6_aq.nml -> AE_cb6r5_ae7_aq.nml
CSQY_DATA_cb6r3_ae6_aq -> CSQY_DATA_cb6r5_ae7_aq
GC_cb6r3_ae6_aq.nml -> GC_cb6r5_ae7_aq.nml
NR_cb6r3_ae6_aq.nml -> NR_cb6r5_ae7_aq.nml
PHOT_OPTICS.dat
Species_Table_TR_0.nml
field_table
diag_table
aqm.rc 
CMAQ_Control_DESID.nml
CMAQ_Control_DESID_cb6r5_ae7_aq.nml
CMAQ_Control_Misc.nml

[BrianCurtis-NOAA]

are you using CCPP_SUITES=FV3_GFS_v16 ?

[bbakernoaa]

the RT was using FV3_GFS_v15p2

I'll try to quickly update it to FV3_GFS_v16

[bbakernoaa]

@BrianCurtis-NOAA I updated it to use the FV3_GFS_v16 physics

[JianpingHuang-NOAA]

@BrianCurtis-NOAA , what is the status of the PR merging ?

[BrianCurtis-NOAA]

@JianpingHuang-NOAA It will be a little bit. I'll get to looking at Barry's run dir soon.