Hi,
I tried to calculate my SMOKE output total VOC emissions of one city based on information in below figure to compare with VOC_INV from 2017gb_hapcap_county_monthly_report_CAPs_PEC_POC_09apr2021
However, mine is just about one tenth of EPA’s. I guess I may misunderstand VOC_INV.
Could anyone give me any idea about that inconsistent?
Thank you in advance!
The VOC_INV is VOC from the inventory file that goes into SMOKE processing. It allows us to track the VOC through the process because different species have different molar weights and are not directly additive once speciated. You should see that the sum of the VOC_INV in the output matches the VOC in your input inventory file. Does that hold true for your case?
What sector were you processing where you saw the inconsistency? And for what year?
I see, thank you for your explaination. Sorry, I should check them carefully that their units are mol/s not g/s.
I processed all sectors except fire and biog emis using NEI2017.
The VOC_INC from smoke output matched with the VOC_INC from 2017gb_hapcap_county_monthly_report_CAPs_PEC_POC_09apr2021
Hi Alison,
Is there an available file that includes a summary of specific VOC species (e.g. IOLE, OLE, PAR, ISOP, APIN, XYL…) emissions for different counties for NEI2017 or NEI2019? or even monthly total emissions? I want to QA those specific emissions in my SMOKE output files.
Thank you in advance!
Does anyone know which species are included in the VOC_INV?
Thanks in advance.
The list of VOC species included in VOC is in our recent technical support documents under
https://www.epa.gov/air-emissions-modeling/emissions-modeling-platforms
See the speciation section of Chapter 3 of the TSD of interest.
Here is the list from the 2016v3 TSD Table 3-3 for the model species mapped to VOC:
AACD Acetic acid
ACET Acetone
ALD2 Acetaldehyde
ALDX Propionaldehyde and higher aldehydes
APIN Alpha pinene
BENZ Benzene (not part of CB05)
CH4 Methane
ETH Ethene
ETHA Ethane
ETHY Ethyne
ETOH Ethanol
FACD Formic acid
FORM Formaldehyde
IOLE Internal olefin carbon bond (R-C=C-R)
ISOP Isoprene
IVOC Intermediate volatility organic compounds
KET Ketone Groups
MEOH Methanol
NAPH Naphthalene
NVOL Non-volatile compounds
OLE Terminal olefin carbon bond (R-C=C)
PAR Paraffin carbon bond
PRPA Propane
SESQ Sesquiterpenes (from biogenics only)
SOAALK Secondary Organic Aerosol (SOA) tracer
TERP Terpenes (from biogenics only)
TOL Toluene and other monoalkyl aromatics
UNR Unreactive
XYLMN Xylene and other polyalkyl aromatics, minus naphthalene
I see, thank you for this information!
Hi,
I tried to calculate the total VOC based on the molecular weight of these species you provided, the molecular weight collected from
However, my total VOC is almost two times VOC_INV. Do you have any idea about that?
Thank you in advance.
The sum of the component species don’t sum to VOC_INV because some of the species which are derived from VOC are actually TOGs but not technically VOCs. The TOG will include more mass.
I also checked XYLMN further in 1 or 2 grids which cover one city in July using NEI2017 model_real/merged2D emission, convert unit from mol/s to lb/month, then compare it with estimated xylenes(mixed isomer) -1330207 in July 2017 (collected annual one from 2017 National Emissions Inventory (NEI) Data | US EPA), there is a large difference. Merged2D one is larger than the collected one.
We note that the NEI HAPs are generated via HAP-augmentation in EIS whereas the platform is generated through VOC/TOG speciation (since it is not NBAFM) via the GSREF. We suspect that if you were to look at BENZ rather than XYLMN, the difference should be smaller. Thus the differences will vary by pollutant and how comparable the HAP augmentation profiles are as compared to the speciation profiles. In 2020 we’ve worked to make the two approaches more similar.
Thanks for your explanation!
I am using WRF-Chem, but there is a significant overprediction for OC over my interested domain in the summer months. So I need to make sure there is no problem with the VOCs emissions I generated using SMOKE.
Which one could generate more accurate emissions for VOC species, EIS or GSREF? If the first one, I may consider adjusting my emissions.
You are welcome. Which year are you running?
2019, current results are based on NEI2017, will update it using NEI2019.
We are unsure why you are assuming HAPs/HAP-speciation issues would be the reason for OC overprediction. PM emissions and speciation are likely closer to the issue. Have you reviewed those aspects of the inventory and determined which sectors are contributing the most to OC?
Also, if you review the 2017 state sector summary here, it may give hints about which sectors contribute the most OC.
https://gaftp.epa.gov/Air/emismod/2017/reports/2017gb_state_sector_totals_CDC.xlsx
Thank you for this information!
I did a quick comparison between the 2017 state sector summary and my report summary for Aug and Jul using 2017gb_17j. There is almost 50% more TOL and XYL for nonroad sectors in my summary report than in the summary you provided. I need to check if I might run SMOKE4.7 using NEI2017 incorrectly.
Thank you.
Hi, we note that TOL and XYL are not explicit species in our modeling platforms. If you are comparing TOL and XYL to toluene and xylenes in the 2017 NEI, that could be another avenue creating substantial differences
Hi,
I compared TOL and XYLMN from both files. I just compared NO2 and POC from two summary files, they are also different. Is that normal? 
Hi,
Sorry, I found I should not compare the annual total with that in Aug report. The total yearly emission in my Aug state report needs to be divided into 12, then compared with monthly emissions. The annual total emissions in different month reports are different.
