Re: [WRF-Chem anthro_emiss] Significantly low PM10 concentration + shifted simulated ozone peak

[Alice Burton]

Dear Mary,

Thanks for your advice. As you suggested I removed :

 auxinput8_inname = 'wrfchemi_gocart_bg_d<domain>',

 auxinput8_interval_m  = 86400,

I attached my new namelist. But once running real.exe I received an error (attached). I previously faced this error, and when I added auxinput8_inname = 'wrfchemi_gocart_bg_d<domain>' to my namelist, the error disappeared.

So what is your suggestion please?

Did you check my emission and mozbc script please? Are they correct? I think the reason for getting low concentration of PM10 might be related to not providing BC/IC for it. What is your idea?

On Tue, Mar 19, 2024 at 6:19 PM Mary Barth <bar...@ucar.edu> wrote:

Hi Alice,

My colleague and I noticed two things in your namelist.input file. It appears that the wrfchemi files may not be getting read because the line is commented out in your namelist. Of course you can confirm this by looking at your rsl.out.0000 file to see if wrfchemi files are being read.

You are using option 202 but you have GOCART background fields being read. I don't know what this does, but I suggest not reading GOCART fields as you are using MOSAIC aerosol scheme.

Regarding ozone, whenever I see a time shift I check to make sure the model output and observations are on the same time zone. Often observations are in local time but WRF is using UTC time. It could also be caused by the lack of anthropogenic emissions.

- Mary

^--^--^--^--^--^--^--^--^--^--^--^--^--^--^-^--^--^--^-^--^

Mary Barth 

Senior Scientist   

phone: 303-497-8186    email: bar...@ucar.edu

NSF - National Center for Atmospheric Research

P.O. Box 3000

Boulder, CO 80307

Staff web page

SIMA Web Site

MUSICA Web Site

DC3 Web Site 

^--^--^--^--^--^--^--^--^--^--^--^--^--^--^-^--^--^--^-^--^

On Tue, Mar 19, 2024 at 5:21 AM Alice Burton <alicebu...@gmail.com> wrote:

Hi every one,

I am running WRF-chem version 4.5.2 with chem option 202. I used dust option 13 which is compatible for this option, but I noticed the simulated PM10 concentration is significantly low and unreasonable. So I need your help and advise to remove this issue.

To generate anthropogenic emission files I used edgard 2015, used mozbc for IC/BC using WACCM data.

I also noticed simulated ozone peak happened before observed peak, and I wonder why this happened (check the attached). And how can I fix it please?

I attached my namelist.input anthropogenic and mozbc script. Would you please check them. I guess there are some mistakes in equations specially PM10.  Should I change any of equations please? If so, could you please inform me.

Look forward to hearing from you.

Thanks 

[Alice Burton]

Dear Mary, 

I modified the namelist, and did not receive that error message. wrfchemi files are getting read (check the attached file).

I strongly need advice on how to better capture PM10 concentration. 

Thanks

[Alice Burton]

Dear Mary,

So many thanks for your advice. I used your mozbc script, and for the first hours of simulation the concentrations of PM10 increased in comparison with previous run using my own mozbc, but again the PM10 concentration decreased (attached). 

I checked the wrfchemi files and noticed that PM10 emissions are considerably low (attached). Because as observations show PM10 is

 currently a huge concern in our area because of their high concentration. 

So I need to increase PM10 emission in wrfchemi files.

In anthro_emission script there is a line that seems to be related to the emission of PM10 ( if not please inform me) :

'PM_10(a)->PM10 + -1.0*PM2.5',

Do you think it is ok to increase the PM10 emission by changing the coefficient of PM10? Say :

'PM_10(a)-> 70* PM10 + -1.0*PM2.5',  

Please guide me in this regard. Maybe I should use another way to increase PM10 emission in wrfchemi, if this is the case please let me know.

Thank you so much .

On Tue, Mar 19, 2024 at 10:11 PM Mary Barth <bar...@ucar.edu> wrote:

Hi Alice,

I do not see anything obviously wrong in the anthro_emis.inp . However, I recommend viewing the wrfchemi file data and comparing it to the EDGAR emissions inventory to see if they look the same.

In the mozbc input file, there is an error. Option 202 uses only 4 aerosol size bins, yet your input file is trying to assign WACCM output to 8 aerosol size bins. You need to combine bins 1 and 2, bins 3 and 4, bins 5 and 6, and bins 7 and 8 like I have in my mozbc input file (attached).

Regards,

[Mary Barth]

Hi Alice,

What emissions source do you expect to contribute to PM10 in the area you are studying? Could you let us know what region your simulation is for? 

What science question are you trying to answer? I ask because you are starting to do science rather than technical fixes. If it is an anthropogenic source, then it makes sense to address the anthro_emiss question. If it is natural dust emissions, then it makes sense to address the dust emissions parameterization and make adjustments. Same with sea salt. All of these take you doing some homework to determine what makes sense for adjusting emissions for your region. 

It would be good for you to look at chem_driver and emissions_driver to see how the code incorporates the PM10 emissions into the MOSAIC aerosol scheme. Remember that PM10 is the summation of aerosol components for a certain size range. So, when the latest simulation has higher values, what component dominates the PM10 concentration? 

I'm not sure what else I can advise at this point, but will continue to try to help with technical issues.

Regards,

Mary

^--^--^--^--^--^--^--^--^--^--^--^--^--^--^-^--^--^--^-^--^

Mary Barth 

Senior Scientist   

phone: 303-497-8186    email: bar...@ucar.edu

NSF - National Center for Atmospheric Research

P.O. Box 3000

Boulder, CO 80307

Staff web page

SIMA Web Site

MUSICA Web Site

DC3 Web Site 

^--^--^--^--^--^--^--^--^--^--^--^--^--^--^-^--^--^--^-^--^

[Alice Burton]

Dear Mary,

Thanks so much for your advice. My study area is over Tehran,Iran. As far as I know, dust, road dust and some human activity like mining ... have contributions to the PM10 concentration over the study area. So as you pointed it is important to know which source has the major contribution and fix it. 

The science question I want to answer is to investigate the role of wet deposition in scavenging aerosols and gases from the atmosphere over the study area. So I first need to have the correct simulation of both gases and aerosol concentrations.

So what is your suggestion please?

Thank you.

[Alice Burton]

Dear Mary,

I increased the PM10 emission in wrfchemi files ('PM_10(a)->500*PM10 + -1.0*PM2.5',), but I noticed the simulated PM10 concentration slightly decreased, and this is very weird to me!!!! Why should this happen?

I increased the emission of PM10 but  the concentration decreased a bit.

I also previously noticed by adding the below lines in mozbc script, PM10 concentration increased.

'num_a01->0.9996*num_a2+0.7135*num_a1+0.0000*num_a3;1.0',
 'num_a02->0.0004*num_a2+0.2847*num_a1+0.0239*num_a3;1.0',
 'num_a03->0.0000*num_a2+0.0016*num_a1+0.6258*num_a3;1.0',
 'num_a04->0.0000*num_a2+0.0000*num_a1+0.3501*num_a3;1.0'

So I think the PM10 concentrations might be affected by these variables in the mozbc script.

What is your opinion please?

Thank you.

[Mary Barth]

Hello Alice,

It would be best for you to go into the code and learn how PM10 is calculated. Start in the chem_driver.F code. Near the end of the chem_driver module is the calculation of PM2.5 and PM10. Take a look at that subroutine which may be in another file. This is the best way to understand the results you are getting. 

Mary

^--^--^--^--^--^--^--^--^--^--^--^--^--^--^-^--^--^--^-^--^

Mary Barth 

Senior Scientist   

phone: 303-497-8186    email: bar...@ucar.edu

NSF - National Center for Atmospheric Research

P.O. Box 3000

Boulder, CO 80307

Staff web page

SIMA Web Site

MUSICA Web Site

DC3 Web Site 

^--^--^--^--^--^--^--^--^--^--^--^--^--^--^-^--^--^--^-^--^

[Alice Burton]

Dear Mary,

Thanks for your advice. But I could not find the codes. Could you please provide more details?

Kind regards,

[Mary Barth]

Alice,

The codes are in the chem/ directory. 

Also, if you haven't had a chance to view the tutorial presentations, they are located at these links. 

WRF     https://www2.mmm.ucar.edu/wrf/users/tutorial/tutorial.html

WRF-Chem     https://ruc.noaa.gov/wrf/wrf-chem/Tutorial.html

- Mary

^--^--^--^--^--^--^--^--^--^--^--^--^--^--^-^--^--^--^-^--^

Mary Barth 

Senior Scientist   

phone: 303-497-8186    email: bar...@ucar.edu

NSF - National Center for Atmospheric Research

P.O. Box 3000

Boulder, CO 80307

Staff web page

SIMA Web Site

MUSICA Web Site

DC3 Web Site 

^--^--^--^--^--^--^--^--^--^--^--^--^--^--^-^--^--^--^-^--^

[Alice Burton]

Dear Mary,

Thanks for your email. In a previous email you mentioned: "Start in the chem_driver.F code. Near the end of the chem_driver module is the calculation of PM2.5 and PM10." So I followed your suggestion and opened the "chem_driver.F", and looked for the pm10 calculation, but could not find it. So I need your help to find the code related to the pm10 calculation please.

Kind regards,

[Rajesh Kumar]

Dear Alice, 

For chem_opt = 202, PM2.5 and PM10 mass concentrations are calculated in aerosol_driver.F by the subroutine sum_pm_mosaic_vbs4. Also, the emissions are translated into concentrations in module_mosaic_addemiss.F. Anthropogenic PM10 emissions are added to OIN. So, you should see the effect of changing PM10 emissions in OIN, which also represents natural dust in the model. 

Hope this helps. 

Rajesh

--

Rajesh Kumar

Interim Lab Deputy Director

Research Applications Laboratory 

NSF National Center for Atmospheric Research (NSF NCAR)

E-mail: rku...@ucar.edu

Phone: +1-303-497-1413

Web: https://staff.ucar.edu/users/rkumar

My working day may not be your working day. Please do not feel obliged to reply to this email outside of your normal working hours.