How to compare sensitivity of O3 to VOC between CMAQ/DDM and CAMx/DDM

CMAQ-DDM-3D
1

Hi,
I am utilizing CMAQ/DDM version 5.4.0.3 to conduct sensitivity of O3 to NOx and VOC. My intention is to compare the model results with those from CAMx/DDM version 7.0. However, I’m encountering challenges when attempting to compare the sensitivity of O3 to total VOC. This complexity arises due to disparities in the VOC species listed within the sensitivity control files, as well as differences in the chemistry mechanisms employed by the two models. The VOC species outlined in the respective models are provided below. Given this scenario, I’m contemplating the feasibility of directly comparing sensitivity outcomes between these two models. Please let me know if you have any ideas on this topic. Thanks.

26 VOC species in cb6r5_ae7_aq used in CMAQ v5.4
'ALD2 ','ALDX ','ETH ', 'ETHA ','ETOH ','FORM ','IOLE ','ISOP ', 'MEOH ','OLE ','PAR ','TERP ','TOL ', 'XYLMN ','NAPH ','ETHY ','PRPA ','ACET ', 'KET ','GLY ',‘BENZENE’,'GLYD ','MEPX ', 'APIN ','SOAALK ','MGLY '/

19 VOC species in CB6r4 used in CAMx v7.0
‘PAR’, ‘ETHA’, ‘MEOH’, ‘ETOH’, ‘ETH’, ‘OLE’, ‘IOLE’, ‘ISOP’, ‘TERP’, ‘FORM’, ‘ALD2’, ‘ALDX’, ‘TOL’, ‘XYL’, ‘PRPA’, ‘BENZ’, ‘ETHY’, ‘ACET’, ‘KET’

2

There have been other published efforts that compared CMAQ and CAMx that tried to reconcile mechanism and species lumping differences. It might be helpful to scan through the literature for that. The DDM sensitivity code does not specifically label any species as VOCs, because there is no need to do so.

Sergey

3

Sergey,

I appreciate your prompt response. I am currently in the process of reviewing papers that compare CMAQ and CAMx. I will read more to gain a more comprehensive understanding.

You’re correct, the DDM sensitivity code doesn’t explicitly categorize any species as VOCs. In our efforts to explore methods for the possibility of reducing VOC emissions and implementing VOC emission control strategies, my focus usually extends beyond individual VOC species. Our aim is to curtail overall emissions, encompassing all VOCs or the total VOC.

This rationale underscores my intention to delve into the DDM sensitivity of VOCs as a collective entity, which encapsulates all VOC species, in relation to ozone. Does this approach make sense?

Thanks.

4

That sounds like a fine approach to me.

Sergey

5

Hello, I am also using DDM of CMAQ to make VOC contribution to HCHO. I would like to know how you handled the total VOC at last

6

Hello LMZ,

A question similar to yours has been discussed in this previous thread. As discussed there, you will have to list all the emitted VOC species towards which you would like to calculate FORM sensitivity.

For example, if your emitted VOC species are AACD, ALD2, MEOH, FACD, ALDX, and ETHA (this very likely is an incomplete list, you will have to check your emission files) and you would like to calculated FORM sensitivity towards those species, you could define a “EVC” group as follows:

EVC     
 EMIS
  GRIDDED_EMIS
 SPECIES
  AACD, ALD2, MEOH, FACD, ALDX, ETHA

In your ASENS output file you should then see FORM_EVC (among other species) which would be the sensitivity of FORM towards emissions of the (VOC) species included in the EVC definition. As noted above, the full list of your emitted VOC species is likely much longer, and as also noted by Sergey in the thread linked above, the exact definition of “VOC” for your analysis is something you will have to decide on.

Also, please do not post the same question in multiple threads as it introduces unnecessary clutter to the forum.

7

First of all, thank you very much for your professional reply, and I apologize to you for sending the same question twice. Because it is the first time for me to use this platform, I am not particularly familiar with our rules, and I will pay special attention to it in the future. According to your suggestion, I have set the format as shown in the file, and it can run now. I want to extract the sensitivity identification of NWVOC emissions to formaldehyde from the grid. Which software should I use for post-processing? Do we have any ready-made post-processing programs that we can use?
sensinput.2018_12NE3.txt (2.5 KB)

8

To get started with post-processing outputs, you may find the information in Chapter 8 of the users guide helpful.

In particular, the combine program released as part of the CMAQ code repository allows you to aggregate individual variables in raw model output files into variables of interest. To do so, you will have to create custom SpecDef species definition files that define the aggregations you would like to perform. Previous discussions (here and here) on using combine to post-process CMAQ ISAM outputs may be helpful.

Based on the sensinput file you posted, the sensitivity parameter you define as EMISNVOC would track the sensitivity of all model species to FORM emissions, not the sensitivity of FORM to VOC emissions. If you want to calculate sensitivity to VOC emissions, you will have to list all of the relevant emitted species in the line after the “SPECIES” entry, as discussed in my previous post.

Finally, as discussed in the DDM-3D user guide chapter I linked to in one of my earlier responses in the other thread, the sensitivity parameter name should only be 3 characters long, not 8 as in your definition of “EMISNVOC”. The reason is that the variables in the ASENS output file will be combination of model species names and sensitivity parameter name, separated by an underscore. Since the total variable length is limited to 16 and the names of some model species can be fairly long, defining a sensitivity parameter name with more than 3 characters can cause problems.

10

Hello Professor, I would like to ask you one more question. My input emission is the MEIC list of Tsinghua University, and the chemical mechanism is CB05. However, since CMAQ5.4 only supports the CB6 mechanism, do I need to change the MEIC input emission of my CB05 mechanism to the CB6 mechanism in this case? Currently, the CMAQ5.4 model with input CB05 emissions data can also run. Is it possible to use the input emission mechanism without converting CB05 to CB6, or should I copy the chemical mechanism from CMAQW5.2 to CMAQ5.4? If I need to convert the CB05 mechanism for input emissions to CB6, what tools can you recommend?

11

I can only provide general answers to your questions since I do not have a chemistry background.

If you read through the original CB6 documentation (Yarwood, G., Whitten, G.Z., Jung, J., Heo, G., Allen, D.T., 2010. Development, evaluation and testing of version 6 of the Carbon Bond chemical mechanism (CB6), Final report to the Texas Commission on Environmental Quality, available here, with an additional conference preprint available here), you will note the following in Section 4.2 of the full report:

Emission inventory preparation for VOCs includes a step called chemical speciation where the VOC species included in the inventory are assigned to the model species included in the chemical mechanism. There are 5 new VOC model species in CB6 that should be considered in emissions processing:
…

CB6 is backward compatible with CB05 and CB4 and can be used with emission inventories … that were prepared for CB05 or CB4. However, updating model inputs to CB6 is preferable to take full advantage of mechanism improvements.

You will have to decide for your project if using the coarser level of chemical speciation applied in processing cb05 emissions is appropriate for the chemical processes you would like to study. If you have access to the underlying VOC inventories for different emissions source categories and cb6 speciation profiles for these categories, generating CMAQ-ready cb6 emissions using a tool like SMOKE would probably be worth considering. Carrying forward the CB05 mechanism from CMAQv5.2 to CMAQv5.4 may be possible but likely would not be very straightforward and is not a route I would recommend pursuing.

12

First of all, thank you very much for your professional and timely guidance. Thanks to your guidance, I can basically run well now. I am especially happy to have found our website in time, which has greatly helped and improved my learning of CMAQ5.4 and simulation. At present, I still have a question, which I would like to consult with professor. I used DDM module to conduct simulation. I want to do pollutant inversion and obtain the contribution concentration of pollutant source emission to the receptor. At the same time, I also need to compare the total simulated concentration of pollutant with the satellite concentration data. So this comparison is how many to how many levels do I have to average when I’m extracting the concentration? And how does the concentration of the pollutants extracted match up with the satellite data?Which means I have to simulate the concentration unit how does it match up with the satellite simulation unit

13

Comparing model output data to satellite products can be complex and depends on the specific satellite products you plan on using. There is no turnkey package available in the CMAQ code repository to support such analyses.

The cmaqsatproc package developed by Dr. Barron Henderson (@barronh) and available through his github repository may be useful depending on your application.

Since your most recent questions are no longer related to CMAQ-DDM3D, you might receive more responses if you start a new thread specific to your post-processing question. Forum users who might have answers to your post-processing questions may not read the current thread because of its unrelated title.

14

CTM_LOG_031.v54_intel_d01_20170129.txt (11.2 KB)
Hello, Professor, I have another question for you. When I used DDM simulation in CMAQ5.4, I had to input emission and used MEIC emission, which was written in the sensinput.2018_12NE3.dat file:

EVC

EMIS

GRIDDED_EMIS

SPECIES

FORM, ETHA, PAR, ETH, OLE, IOLE, ISOP, TERP, MEOH, ETOH, ALD2, ALDX, TOL, XYL

Currently, the error is reported in the following file.

15

Since your emission file contains the CB05 species XYL which in CB6R3 was split between XYLMN (xylene minus naphtalene) and NAPH (naphtalene), you need to map your XYL emissions to one or both of these CB6R3 model species via the DESID module before you can reference then as part of the sensinput.dat DDM-3D control file.

Specifically, you will need to edit CMAQ_Control_DESID_cb6r5_ae7_aq.nml to provide a rule to map your XYL emissions to the XYLMN and maybe NAPH model species. After doing so, you would then reference XYLNM and NAPH instead of XYL in sensinput.dat.

I cannot provide guidance on what a proper mapping would look like, as this goes back to sector-specific speciation factors that normally would be taken care of during emissions inventory processing through tools like SMOKE.

If you decided that you want to map all of your XYL emissions to XYLMN and none to NAPH, the new rule you would add to CMAQ_Control_DESID_cb6r5_ae7_aq.nml would look like this:

'EVERYWHERE', 'ALL'         ,'XYL'  ,'XYLMN'       ,'GAS'  ,1.  ,'UNIT','a', 
'EVERYWHERE', 'ALL'         ,'XYL'   ,'NAPH'        ,'GAS'  ,0.  ,'UNIT','a',

If you instead wanted to apply a 90/10 split to your XYL emissions, the new rule you would add to CMAQ_Control_DESID_cb6r5_ae7_aq.nml would look like this:

'EVERYWHERE', 'ALL'         ,'XYL'  ,'XYLMN'       ,'GAS'  ,0.9  ,'UNIT','a', 
'EVERYWHERE', 'ALL'         ,'XYL'   ,'NAPH'        ,'GAS'  ,0.1  ,'UNIT','a',