Double counting NH3 emissions when using NH3 Bidi flux in CMAQ

I am turning the ‘ammonia bi-directional flux’ on within the CMAQv5.3.2 runscript environment variable setenv CTM_ABFLUX Y and feeding in FEST-C input files:

setenv E2C_SOIL ${INPDIR}/toCMAQ_festc1.4_epic/us1_2016_cmaq12km_soil.nc
setenv E2C_CHEM ${INPDIR}/toCMAQ_festc1.4_epic/us1_2016_cmaq12km_time${YYYYMMDD}.nc
setenv E2C_LU ${INPDIR}/beld4_camq12km_2011_4CMAQioapi.ncf

There is also another option in the runscript that I switched on to subtract fertilizer NH3 from emissions because it will be handled by the BiDi calculation: setenv CTM_BIDI_FERT_NH3 Y.

From what I understand, in the SMOKE emission files, the ‘ag’ sector is one sector that includes NH3 emissions from fertilizers and livestock; there may be other industrial sources of NH3 as well. Does not doing setenv CTM_BIDI_FERT_NH3 Y in CMAQ runscript nullify all NH3 emissions coming from the merged 2D SMOKE emissions (using mrggrd to merge all 2D emission sector files) fed to CMAQ? From what I understand, I should only make the SMOKE fertilizer NH3 emissions zero if I am separately feeding FEST-C daily files, but not sure how I can do that. I am just trying to make sure I am not double counting NH3 emissions in CMAQ!

Fertilizer emissions should not be double counted as long as your emissions file contains the variable NH3_FERT (emissions from fertilizers) and NH3 (the total NH3 emissions from all sectors). When bidirectional NH3 is selected and setenv CTM_BIDI_FERT_NH3 Y the NH3_FERT variable from the emissions file is subtracted from the total NH3 emissions. If you do not have the NH3_FERT variable, then you should get an error when these flags are selected.

Thanks @jbash. When using NH3-bidi in CMAQv5.3.2, is it OK to have NaN values for the variables ‘NH3’ and ‘NH3_Emis’ but not ‘NH3_Flux’ in the CCTM_DRYDEP_* output file? I have more info with plots at the post NaN values of NH3_Emis variable in CCTM_DRYDEP_* file regarding this issue.

Hi @skunwar. It is not okay to get a nan in the NH3_Emis or NH3 variable. Is the model simulation stopping at the NaN? If not, I suspect you are running the m3dry deposition option as the STAGE deposition option uses the NH3_Emis variable internally while it is just a diagnostic output for m3dry. You can confirm the deposition option by searching for m3dry.F in the build directory. If the file exists, then you are running m3dry dry deposition.

If you want to run the STAGE deposition option it is only a flag to set in the bldit_cctm.csh script.

set DepMod = m3dry #> m3dry or stage

set DepMod = stage

Take care,
Jesse

Thanks a lot for the information.

I searched for the terms DepMod, stage and m3dry in my CMAQv5.3.2 (cb6r3_ae7) runscript, and did not find any of them. Then I check my CMAQ BLD* directory when I saw m3dry.F and m3dry.o files, which confirm that I am indeed running the m3dry module.

Actually, my CMAQ run has not aborted at these NaN values of NH3 and NH3_Emis in the CCTM_DEPV_* CMAQ output files, or the Infinity values of standard deviation (screenshot shown in the post NaN values of NH3_Emis variable in CCTM_DRYDEP_* file) of all FEST model output files (input to CMAQ bidi).

My questions here:

  1. Is it a possibly a problem starting with the FEST input files that I downloaded from the repository?
  2. Is there any preference between the two modules m3dry and stage for NH3-bidi process? Would using stage have avoided getting NaN values for NH3 and NH3_Emis variable in CCTM_DEPV_* files?