PM2.5 concentration from history files

[ssanyal3@illinois_edu]

Hi,I have outputs of separate aerosols in my history files. How do I calculate the PM2.5 concentration from the separate aerosol species? Swarnali

 

[alcide_zhao@ed_ac_uk]

Hi Swarnali,i wodner if you get this solved? since I  am also facing this questions and cannot find any documentations to help me out. Would be much appreciated Cheers, AlcideAlcide D. Zhao——PhD in Atmospheric and Environmental SciencesSchool of GeoSciences, University of EdinburghE-mail:Alcide.Zhao@ed.ac.uk

 

[Yuan Wang]

Alcide, You have to DIY for an accurate calculation of total PM2.5 in MAM3/7. Take MAM3 for example, total PM2.5 is composed of total mass in the Aitken and Accumulation modes and a part from the Coarse mode. To get the coarse part, you need to integrate its log-normal spectrum up to 2.5 micron (size range of coarse mode is typically 0.8 to 3.6). Yuan

 

[alcide_zhao@ed_ac_uk]

Hi all,

An update on Yuan's comments. CESM1 doesn't output PM2.5 directly, so it needs to be calculated from the species of the chemical mechanism. If you use the Bulk Mode aerosol scheme in CESM,
       PM2.5=NH4NO3+SO4+OA+BC+DST01+DST02+SSLT01+SSLT02,
       with OA=1.8xOC+SOA, BC=CB1+CB2, OC=OC1+OC2 (which are the hydrophobic and hydrophilic forms).
       DST01 and DST02 are dust with diameter < 2.5 um and SSLT01 and SSLT02 are seasalt with diameter < 2.5 um.
       Also check what SOA scheme when applicableIf you use CAM3/7, as commented by Yuan above, please refer Table 1 in this paper (https://doi.org/10.1175/JCLI-D-11-00650.1) for the diameter distribution of aerool particles in the aerosol modes, and sum the mass of all aerosol species up (caution for sea-salt and dust in the Coarse mode, which should be only a part of the integraion). Note that in MAM3/7, sulphate is partly neutralized by ammonium in the form of NH4HSO4 so that ammonium is effectively prescribed and NH3 aerosol is not simulated.

Please feel free to comment/perfect this!

Cheers, Alcide

 

[anushree]

Hi all,
With cesm2.1.1, PM2.5 is now a direct output.But does anyone know about its constituents.Kindly help if you have answers.
Regards,
Anushree

 

[Francesca Bruno]

I ran a simulation using CESM1.2.2 (with MAM3), but there is no direct PM2.5 concentrations in the history file. I used some methods to calculate PM2.5, but found that there is large bias between modelde PM2.5 and observed PM2.5, especially for some highly-polluted regions (such The gangetic plain, JAVA Island and North China). However, the default emission inventories in MAM3 can show the pattern (where PM2.5 in the gangetic plain, JAVA Island and North China is the largest worldwide). Therefore, I am thinking about whether my method of caculating PM2.5 is not correct. The following is my way to get PM2.5:

PM2.5=bc_a1_SRF+pom_a1_SRF+so4_a1_SRF+so4_a2_SRF+so4_a3_SRF*r+soa_a1_SRF+soa_a2_SRF+dst_a1_SRF+dst_a3_SRF*r+ncl_a1_SRF+ncl_a2_SRF+ncl_a3_SRF*r
r here is the proportion of aerosols below 2.5 micron in the Coarse mode to total aerosols in the Coarse mode.

Whether I should use other variables to calculate PM2.5? like num_a1_SRF(num_a1 in bottom layer) or BURDEN1 (Aerosol burden mode, unit:kg/m2)?

 

[nusbaume]

Hi All,

I am just letting you know that I am moving this thread to the CAM-Chem forum, as that way it will be seen by our aerosol experts (who hopefully will have the relevant answers to your questions).

Thanks, and good luck with your queries!

Jesse

 

[Hemraj]

Hi all,

An update on Yuan's comments. CESM1 doesn't output PM2.5 directly, so it needs to be calculated from the species of the chemical mechanism. If you use the Bulk Mode aerosol scheme in CESM,
PM2.5=NH4NO3+SO4+OA+BC+DST01+DST02+SSLT01+SSLT02,
with OA=1.8xOC+SOA, BC=CB1+CB2, OC=OC1+OC2 (which are the hydrophobic and hydrophilic forms).
DST01 and DST02 are dust with diameter < 2.5 um and SSLT01 and SSLT02 are seasalt with diameter < 2.5 um.
Also check what SOA scheme when applicableIf you use CAM3/7, as commented by Yuan above, please refer Table 1 in this paper (Toward a Minimal Representation of Aerosols in Climate Models: Comparative Decomposition of Aerosol Direct, Semidirect, and Indirect Radiative Forcing) for the diameter distribution of aerool particles in the aerosol modes, and sum the mass of all aerosol species up (caution for sea-salt and dust in the Coarse mode, which should be only a part of the integraion). Note that in MAM3/7, sulphate is partly neutralized by ammonium in the form of NH4HSO4 so that ammonium is effectively prescribed and NH3 aerosol is not simulated.

Please feel free to comment/perfect this!

Cheers, Alcide

Dear Alcide,
Thanks for the nice explanation. I have some confusion:
i. are OC and SOA completely isolated (or they are overlapped)?
ii. How about the possibility of NH4SO4 formation, just like NH4NO3? Can we separately add NH4NO3 as NH4 + NO3? The unit of NH4 and NO3 are in mol/mol (not kg/kg like NH4NO3) but we can change the unit since we know the molecular weight.
I am also currently looking for way to accurately calculate PM2.5 from my CESM results.
Thank you.

Best, Hemraj

 

[Hemraj]

Hi all,
With cesm2.1.1, PM2.5 is now a direct output.But does anyone know about its constituents.Kindly help if you have answers.
Regards,
Anushree

Dear Anushree,
I am using CESM2.1.3 with FMOZ compset but couldn't see PM2.5 in my model output. Do we need to change some settings to get PM2.5?
Thank you.

Best, Hemraj

 

[anushree]

Hi @Hemraj ,
PM2.5 is a direct output in cam6 if you are using the chemistry version i.e FCHIST. CESM2.a.b (where a and b are version numbers) is giving PM2.5 as direct output which was not the case in CESM1.2.2.My question was actually regarding CESM2.2 with FCHIST compset for which i have not found the answer yet.I will let you know if i get one.
Thank you.

 

[rrbuchholz]

This post is related to a similar post in the CAM forum: PM2.5 as output in cam6

 

[Francesca Bruno]

Alcide, You have to DIY for an accurate calculation of total PM2.5 in MAM3/7. Take MAM3 for example, total PM2.5 is composed of total mass in the Aitken and Accumulation modes and a part from the Coarse mode. To get the coarse part, you need to integrate its log-normal spectrum up to 2.5 micron (size range of coarse mode is typically 0.8 to 3.6). Yuan

Dear Wang,
I used the modal aerosol scheme (MAM3) in CAM5, and I think we can use the total mass in the Aitken and Accumulation modes and a part from the Coarse mode the to represent the PM2.5. But the point is how to get the PM2.5 fraction in coarse mode. According to the inputdata file of MAM3 (atm/cam/physprops/mam3_mode3_rrtmg_c110318.nc; also be attached), the geometric standard deviation, the geometric mean diamater and the size range seem to be prescribed and constant. If so, the lognormal distribution of aerosol is prescribed and does not vary with time and location. Then the PM2.5 fraction of coarse mode will be constant. However, the aerosol size changes over time, which is the deliquescence process. I wonder if MAM3 sets the aerosol log-normal distribution as a constant for a particular mode, and if not, how to get the PM2.5 fraction for the coarse mode? Many thanks! I've been puzzled by this question for a long time.

It seems that I can not upload the NC file here. So I show the relative variables in the following picture: