Scheduled Downtime
On Friday 21 April 2023 @ 5pm MT, this website will be down for maintenance and expected to return online the morning of 24 April 2023 at the latest

interaction of aerosols in the Thompson scheme

jpen

New member
Is there any reference or documentation on the level of aerosol interaction implemented in the Thompson aerosols microphysics scheme? What is known is that nwfa and nifa are read from the aerosol climatology file (QNWFA_QNIFA_SIGMA_MONTHLY.dat) which are used in the module_mp_thompson_aerosols that includes subroutine gt_aod to compute AOD.
  1. Do the concentrations (nwfa and nifa) vary in (or interact with) other processes?
  2. the output field 'scalar' includes aerosols (seasalt, sulfate, dust, bc, oc). Where are the emissions/ forcings read?
  3. in the new version, with the thompson_aerosols microphysics scheme, is AOD the only relevant aerosol related variable to be saved?
 
Please see my answers below:
Is there any reference or documentation on the level of aerosol interaction implemented in the Thompson aerosols microphysics scheme? What is known is that nwfa and nifa are read from the aerosol climatology file (QNWFA_QNIFA_SIGMA_MONTHLY.dat) which are used in the module_mp_thompson_aerosols that includes subroutine gt_aod to compute AOD.

This paper describes in detail the aerosol-aware Thompson microphysics scheme:
Thompson, Gregory, and Trude Eidhammer, 2014: A study of aerosol impacts on clouds and precipitation development in a large winter cyclone. J. Atmos. Sci., 71.10, 3636-3658. doi:10.1175/JAS-D-13-0305.1
  1. Do the concentrations (nwfa and nifa) vary in (or interact with) other processes?
They are also active in MYNN PBL scheme. Please see the code "bl_mynn_subroutines.F90", in which the subroutine "mynn_tendencies" handles PBL tendencies including the changes of qnwfa,qnifa,qnbca, etc.
  1. the output field 'scalar' includes aerosols (seasalt, sulfate, dust, bc, oc). Where are the emissions/ forcings read?
I am not sure of this issue.
  1. in the new version, with the thompson_aerosols microphysics scheme, is AOD the only relevant aerosol related variable to be saved?
Would you please clarify this question? My understanding is that AOD is only used in RRTMG radiation scheme . Let me know if I am wrong.
 
Specifically for your question (2), which files contain these aerosols (seasalt, sulfate, dust, bc, oc)? Thanks.

Is there any reference or documentation on the level of aerosol interaction implemented in the Thompson aerosols microphysics scheme? What is known is that nwfa and nifa are read from the aerosol climatology file (QNWFA_QNIFA_SIGMA_MONTHLY.dat) which are used in the module_mp_thompson_aerosols that includes subroutine gt_aod to compute AOD.
  1. Do the concentrations (nwfa and nifa) vary in (or interact with) other processes?
  2. the output field 'scalar' includes aerosols (seasalt, sulfate, dust, bc, oc). Where are the emissions/ forcings read?
  3. in the new version, with the thompson_aerosols microphysics scheme, is AOD the only relevant aerosol related variable to be saved?
 
Thanks for the response and reference. I ran with the 'mp_thompson_aerosols' scheme for 72 hours however, values for aerosols (seasalt, sulfate, dust, bc, oc) concentrations are all zeros in all the timesteps and at all the aerosol levels of the outputs. Hence, I want to know about the emissions. The values of AOD seems to be fine. In this scenario, the related variables of interest seems to be only the AOD. Or am I missing anything other than including the 'mp_thompson_aerosols' option?
 
Last edited:
Would you please upload your namelist.atmosphere and a single MPAS output file that contains aerosols (seasalt, sulfate, dust, bc, oc)? Thanks.
 
please find attached the namelist file. How should I attach the netcdf file? Using convert_mpas, I extracted the aerosol variable and the filesize is about 358 MB. I gzipped it and have attached.
 

Attachments

  • namelist.atmosphere_aero.txt
    2.1 KB · Views: 3
  • latlon.nc.gz
    1.2 MB · Views: 1
Last edited:
@jpen
In MPAS-v8.2.0, those aerosol variables (e.g., seasalt, sulfate, dust, bc, oc) are included in Registry. However, they are never updated in physics. This is why their values are 0. I guess these variables are prepared for MPAS-chem. I will talk to our experts and get back to you.
 
@jpen
Based on our expert, these aerosol variables were only needed by the CAM radiation code (which nobody uses anymore and is quite old). They are not used in MPAS-A. Therefore, please don't worry for the zero values.
 
Thanks for the reply. I understand. I ran a limited area simulations with aerosols but the runs are crashing after a few days. However, the runs without aerosols are going through. The namelist, and the logs are attached.
 

Attachments

  • slurm-11254273.out.txt
    3 KB · Views: 1
  • err_log.atmosphere.0000.out.txt
    3.8 MB · Views: 2
  • ss_LAS_10kms_CP_GF_AERO_2017_namelist.atmosphere.txt
    2.1 KB · Views: 1
Hi,
Please clarify that your run with the option below failed :
Code:
    config_physics_suite = 'convection_permitting'
    config_microp_scheme = 'mp_thompson_aerosols'

But if you run with the option below, MPAS worked fine.
Code:
    config_physics_suite = 'convection_permitting'

Let me know if I misunderstand something here.
 
By saying "ran a limited area simulations with aerosols ", do you mean 'with aerosols' in microphysics or in radiation?
 
Please confirm whether my message below is correct or not?

If the option "config_microp_scheme = 'mp_thompson_aerosols'" leads to MPAS failure, we will need to explore and fix possible issues.

Hi,
Please clarify that your run with the option below failed :
Code:
    config_physics_suite = 'convection_permitting'
    config_microp_scheme = 'mp_thompson_aerosols'

But if you run with the option below, MPAS worked fine.
Code:
    config_physics_suite = 'convection_permitting'

Let me know if I misunderstand something here.
 
Hi,

I apologize for getting back to you late, but this is because we were busy with WRF tutorial in the Feb 3 week. NCAR HPC was down this week and the waiting queue for job running is long.

I run a test case using MPAS-v8.2.0 and same physics options used in your case. My test is done successfully. This implies that mp_thompson_aerosols works as we expect.

For your failed case, since it crashed after more than 2 days of integration, I am suspicious that something else might be wrong that resulted in the crash. Would you please compile MPAS in debug mode, then rerun this case? By running in debug mode, we will get the information when and where the model first crash. This will be helpful for us to figure out what is wrong.
 
Thanks Chen for your response. May I know for how long did you ran the simulations with aerosols and was it for a limited area or global? I rechecked my compilation log and observed that debugging option was ON when compiling. In this case, the log files I attached earlier (on Feb 02) should include debugging mode. Can you please take a look at them?
 
Last edited:
Top