Abnormal PM2.5 simulation with RADM2/MADE/SORGAM for 2017 on CONUS

[pwang18]

Hi all,

I did a hindcast on CONUS for PM2.5 and ozone with RADM2/MADE/SORGAM using WRF/Chem 3.9.1.1. I compared my simulation result with CMAQ and EPA observations. My ozone simulation looks reasonable, while PM2.5 is different with both CMAQ results and EPA observation of regional distribution and order of magnitude.

May I ask for help/comments on my process?
1, I used NEI 2017 and processed it with the EPA_ANTHRO_EMIS tool. Is there any modification needed in my "EPA_anthro_emis_2017NEI.inp"?
2, Did anyone get good PM2.5 CONUS simulation results with RADM2/MADE/SORGAM?
3, I appreciate any other suggestions.

Best,
Phillip

 

[jordanschnell]

Hi Phillip,

Based on your plots, my guess is that either

1) you have dust emissions turned on and they are dominating your PM concentrations
- What PM species is high? ( p25j? )
2) are you using chemical lateral boundary conditions from another model? It looks like the high concentrations are coming in from the southern and northern edge

Jordan

 

[pwang18]

Hi Phillip,

Based on your plots, my guess is that either

1) you have dust emissions turned on and they are dominating your PM concentrations
- What PM species is high? ( p25j? )
2) are you using chemical lateral boundary conditions from another model? It looks like the high concentrations are coming in from the southern and northern edge

Jordan

Hi Jordan,
Thanks so much for your reply.
1) Yes, I turned on dust with duat_opt=4. Yes, p25j is higher. Won't the PM be even lower if this is the case that dust dominated the PM?
2) I used CAM-Chem and mozbc for lateral boundary and initial conditions. The problem is I only have a script for mozart-gocart instead of radm2/MADE/SORGAM. Do you mind sharing the script for radm2/MADE/SORGAM?
Thanks so much.

 

[pwang18]

I didn't see my attachment, so I pasted the code here.

mozart-gocart_mozbc_InitialCondition.inp

&control

do_bc = .false.
do_ic = .true.
domain = 1
dir_wrf = ''
dir_moz = ''
fn_moz = 'hb0000.nc'
moz_var_suffix = ''

spc_map = 'o3 -> O3', 'n2o -> N2O', 'no -> NO',
'no2 -> NO2', 'nh3 -> NH3', 'hno3 -> HNO3', 'hno4 -> HO2NO2',
'n2o5 -> N2O5', 'h2o2 -> H2O2',
'ch4 -> CH4', 'co -> CO', 'ch3ooh -> CH3OOH',
'hcho -> CH2O', 'ch3oh -> CH3OH', 'c2h4 -> C2H4',
'ald -> CH3CHO', 'acet -> CH3COCH3', 'mgly -> CH3COCHO',
'pan -> PAN', 'mpan -> MPAN', 'macr -> MACR',
'mvk -> MVK', 'c2h6 -> C2H6', 'c3h6 -> C3H6', 'c3h8 -> C3H8',
'c2h5oh -> C2H5OH', 'c10h16 -> MTERP',
'isopr -> ISOP','acetol -> HYAC', 'mek -> MEK',
'bigene -> BIGENE', 'bigalk -> BIGALK',
'tol -> TOLUENE+BENZENE+XYLENES',
'cres -> CRESOL', 'dms -> DMS', 'so2 -> SO2',
'BC1 -> 1.0*bc_a4;1.e9',
'BC2 -> 1.0*bc_a1;1.e9',
'OC1 -> 1.0*pom_a4;1.e9',
'OC2 -> 1.0*pom_a1+1.0*soa1_a1+1.0*soa1_a2+1.0*soa2_a1+1.0*soa2_a2+1.0*soa3_a1+1.0*soa3_a2+1.0*soa4_a1+1.0*soa4_a2+1.0*soa5_a1+1.0*soa5_a2;1.e9',
'SEAS_1 -> 1.0*ncl_a1+1.0*ncl_a2;1.e9',
'SEAS_2 -> 0.5*ncl_a3;1.e9',
'SEAS_3 -> 0.5*ncl_a3;1.e9',
'SEAS_4 -> 0.0*ncl_a3;1.e9'
'DUST_1 -> 1.0*dst_a1+1.0*dst_a2+0.02*dst_a3;1.e9',
'DUST_2 -> 0.93*dst_a3;1.e9',
'DUST_3 -> 0.05*dst_a3;1.e9',
'DUST_4 -> 0.0*dst_a3;1.e9',
'DUST_5 -> 0.0*dst_a3;1.e9',
/

 

[jordanschnell]

Hi Phillip,

Dust_opt = 4 will not work with the chem option you have chosen. What do you mean by "Won't the PM be even lower if this is the case that dust dominated the PM?"? It looks like your PM is too high, so lower would be better. If you'd like to include dust, you can try using option = 13.

As for mozbc, it doesn't look like you are using the model for lateral boundary conditions (LBCs) -- i.e., do_bc = .false. As for a mapping table, I don't have one offhand, but here looks to be an example: RACMSORG.inp for CAM-Chem - I imagine you would be able to find one. Your current namelist is not doing anything for aerosols as (e.g., BC{1-2}, OC{1-2}, etc..) are not in your mechanism. You can look at Registry/registry.chem to see those species.

 

[Stephie_MM]

Hello everyone!

I'm new using WRF-CHEM and I am really strugguling about the IC and BC.
Would you mind share with me, where I can find the file that I should put in fn_moz = 'hb0000.nc' ?
I'm trying to use T1_MOZCART input file.