MEGAN NOx output units

MEGAN
1

Hello,

I am creating an output file for MEGAN that is similar to the B3GTS file for BEIS. I am doing this instead of the DESID method because I need separate outputs for agricultural and non-agricultural areas. I am adapting code from tmpbeis.F for this purpose, but I am struggling to ensure I use the proper units. The following comment in MEGAN_DEFN.F seems to be used for the DESID emissions, however, I would expect the following line to make the units into moles/s, not moles/s/m2. Additionally, does this moles/s emission rate need to be multiplied by 3600 to get an hourly amount of emissions? Does it account for the length of the sync timestep?

! Units for use by EMIS_DEFN.F via vdemis_me: [moles/s/m2]
        do i=1,NMGNSPC
          vdemis_me(i,:,:) = out_to_cmaq(i,:,:)*cell_area/msfx2
        end do

If I want this file to output emissions as the sum of the emissions over the period of an output time step (one hour), how would I do this? This seems to be what the documentation for BEIS claims it outputs, but in tmpbeis.F, there are two transformations I don’t fully understand.

  1. Emissions are multiplied by a factor that includes a fraction of the sync timestep to the output timestep. What does this do, and should I include it for my MEGAN output code?
! Unit conversion for the emission and fraction of the sync time step to the output time step
                  MSFC = MSFAC( S,K ) * FLOAT( TIME2SEC( TSTEP( 2 ) ) ) / FLOAT( TIME2SEC( TSTEP( 1 ) ) )
! Add the emission for the model output time step
                  EMISS( :,:,S ) = EMISS( :,:,S ) + EMPOL( :,:,K ) * MSFC
  1. Emissions are converted from per hour to per second before writing. Does each hourly output for B3GTS thus correspond to an average emissions rate over the previous hour, i.e., multiplying by 3600 would give you the mass emitted in that hour?
         IF ( LASTTIC ) THEN
            EMISS = EMISS * HR2SEC   ! convert from gm/h to gm/s

Apologies for a long post with many questions, but any help is appreciated.

2

I can comment on the MEGAN side of things. You’re right that the vdemis_me units are moles/s after that line. The comment before it should be clearer.

The emission rate is instantaneous for the time step. You would either multiply by 3600 to get an estimate of the hourly emissions, or accumulate each time step’s emissions and sum them up every hour for perfect accuracy. I’m not sure how BEIS does it but my hunch is that the estimate will be extremely close to the actual value.

Are you using BDSNP or YL for soil NO?

2 Likes
3

Thank you for the swift response. To clarify, for the approach accumulating each time step’s emissions, would you multiply each sync time step’s emissions by the length of that timestep to get a total? (multiply by TIME2SEC( TSTEP( 2 ) ) and then add)

We’re currently using YL, although we plan on testing with both. I know BDSNP simulates much higher emissions than YL, and some bug fixes are coming in CMAQ v5.5, but we might want to stick with 5.4 for now since it’s already set up with my changes. An important thing for us is accurately capturing temporal and spatial profiles of NO emissions in one of these routines. We plan on using an hourly profile of NOx emissions to partition daily magnitude estimates from our modified FEST-C to get finer-scale impacts on chemistry.