1
Improving Chemical Prof. Christopher Cappa, UC Davis Prof. Michael Kleeman, UC Davis Mechanisms for Ozone and Prof. Tony Wexler, UC Davis
Secondary Organic Carbon Prof. John Seinfeld, Caltech
California Air Resources Board 31 August 2018
Project: 12-312
Image: Wikipedia
Overview
Part 1: Improving the representation of secondary organic aerosol in air quality models
Part 2: Updating modeling scenarios for reactivity assessment
Part 3: Evaluate organic nitrate and N2O5 chemical mechanisms and their impact on secondary aerosol
UCDAVIS CIVIL AND ENVIRONMENTAL ENGINEERING & AQRC 2
Part 1: Improving SOA models
When we proposed this work:
• Air quality models simulated SOA formation using simple n-product models
𝑉𝑂𝐶 + 𝑂𝐻 → 𝛼1 ∙ 𝑃1 + 𝛼2 ∙ 𝑃2
• Predictions from n-product SOA models generally underpredicted SOA relative to observations
• There was need to represent the dynamic, multi-generational nature of photooxidation reactions
Measured Modeled
Me
asure
d/M
od
eled
[Volkamer et al., GRL, 2007]
c-r40 E C>30 :::t ......
<( 20 0 en
10
0
100
"8 E
~ ~ 10
1
0
0.1
TORCH 2003
UK PBL I <:> (I) <:> (I)
MCMA 2003 polluted urban
200
NEAQS 2002 US PBL
■
t . ACE-Asia 2001
10
1
100
10
1 10 100
1
1000 Photochemical Age [hh]
UCDAYIS CIVIL AND ENVIRONMENTAL ENGINEERING & AQRC 3
Part 1: Improving SOA models
1a. The Statistical Oxidation Model
1b. Characterizing a previously unaccounted for experimental artifact
1c. Implementation into the UCD-CIT Air Quality Model
6
~ Js C
~4
~3
)2 z
1
0
+-
t }
C 0
!! SOA precursor compound
12 11 10 9 8 7 6 5 4 3 2 1 Number of Carbon Atoms
• • • .......... ...,,. ..... -
2 4 6 8 10 12x103
Initial Seed Surface Area ~ llTI2 cm"3)
2
1.5
0.5
0
Ratio (high/no)
,........=----,=--=------=•- 5
4
3
2
1
~----~~•~o
~r 14
11
8
5
2
UCDAYIS CIVIL AND ENVIRONMENTAL ENGINEERING & AQRC 4
Part 1a: The Statistical Oxidation Model
The SOM is a chemically-based, paramaterizable chemical mechanism that accounts for multi-generation functionalization, fragmentation, and gas-particle partitioning of organic species as they are oxidized in the atmosphere
• Reaction rates of product species constrained by structure-reactivity relationships
• Reaction product distribution governed by physical constraints: adjustable
• Volatility of SOM species linked to # oxygen and carbon atoms: adjustable
• Probability of functionalization versus fragmentation depends on extent of oxygenation
[See Cappa and Wilson, ACP, 2012]
7
6
1
0 • SOA recursor cor poupd I
12 11 10 9 8 7 6 5 4 Number of Carbon Atoms
3 2 1
UCDAYIS CIVIL AND ENVIRONMENTAL ENGINEERING & AQRC 5
The Statistical Oxidation Model
The parameters governing the chemical evolution for a given SOA precursor are determined by fitting to laboratory chamber observations
Five Tunable Parameters
1. Probability of functionalization versus fragmentation
2. Change in volatility upon functionalization (i.e. oxygen addition)
3-5. Probability of adding 1, 2, 3 or 4 oxygen atoms upon functionalization
[See Cappa and Wilson, ACP (2012); Cappa et al. ACP (2013)]
7
6
1
0 • SOA recursor cor poupd I
12 11 10 9 8 7 6 5 4 Number of Carbon Atoms
3 2 1
UCDAYIS CIVIL AND ENVIRONMENTAL ENGINEERING & AQRC 6
Example Reaction for One SOM Species
Functionalization Products
𝐶10𝑂2 + 𝑂𝐻 𝑝1𝐶10𝑂3 + 𝑝2𝐶10𝑂4 + 𝑝3𝐶10𝑂5 + 𝑝4𝐶10𝑂6
𝑘𝑂𝐻,𝐶10𝑂2 , 𝑝𝑓𝑢𝑛𝑐
𝑖=1−4
𝑝𝑖 = 1
𝑓𝑓𝑟𝑎𝑔 𝐶9𝑂3, 𝐶1𝑂1 + 𝑓𝑓𝑟𝑎𝑔 𝐶9𝑂2, 𝐶1𝑂1 + 𝑓𝑓𝑟𝑎𝑔 𝐶9𝑂1, 𝐶1𝑂1 +𝐶10𝑂2 + 𝑂𝐻 𝑘𝑂𝐻,𝐶10𝑂2, (1 − 𝑝𝑓𝑢𝑛𝑐)
𝐶8𝑂3, 𝐶2𝑂2 𝐶8𝑂2, 𝐶2𝑂2 + ⋯… + 𝑓𝑓𝑟𝑎𝑔 + 𝑓𝑓𝑟𝑎𝑔
𝐶5𝑂3, 𝐶5𝑂1 𝐶5𝑂2, 𝐶5𝑂2 𝐶5𝑂1, 𝐶5𝑂3… + 𝑓𝑓𝑟𝑎𝑔 + 𝑓𝑓𝑟𝑎𝑔 + 𝑓𝑓𝑟𝑎𝑔
Fragmentation Products 1
=𝑓𝑓𝑟𝑎𝑔 𝑁𝑓𝑟𝑎𝑔𝑚𝑒𝑛𝑡𝑠
I
(
(
(
I
) (
) (
) (
'
) ( )
)
) ( )
UCDAYIS CIVIL AND ENVIRONMENTAL ENGINEERING & AQRC 7
Parameterizing SOM
• Use observations from the Caltech Environmental Chamber
• Consider a variety of anthropogenic and biogenic VOCs
• a-pinene, isoprene, benzene, toluene, m-xylene, naphthalene, dodecane isomers
• Low- & High NOx conditions
• Fit SOM to observed time-dependent SOA concentrations
• Additionally consider O:C observations as qualitative constraint
[From Zhang et al., PNAS, 2014]
Reaction Time (h)
SOA
Co
nce
ntr
atio
n (m
g m
-3 )
a-pinene + OH
Toluene + OH
60
40
20
o~ _ ___.__ _ __. __ ........_ _ __._ _ ______,...___...,____.
80---------.---....--------.............
60
40
20
0 0 2 4 6 8 10 12
UCDAVI CIVIL AND ENVIRONMENTAL ENGINEERING & AQRC 8
Part 1b: Accounting for Chamber Artifacts: Vapor Wall Losses
• Walls of environmental chambers are Teflon
• Lower volatility species can absorptively partition into chamber walls [Matsunaga & Ziemann (2010)]
• The impact of vapor-wall partitioning on SOA formation had not been previously accounted for!
Photo: www.mpic.de
UCDAYIS CIVIL AND ENVIRONMENTAL ENGINEERING & AQRC 9
Accounting for Chamber Artifacts: Vapor Wall Losses
• Perform experiments where seed High Intermediate Low surface area is varied Volatility Volatility Volatility
• Higher seed SA enhance SOA Particle Particle Particle formation relative to wall loss
• Account for dynamic partitioning in Vapor Vapor Vapor
SOM modeling
particle ↔ vapor ↔ wall Walls Walls Walls
• Fit observations with/without vapor wall loss accounted for
[Zhang et al., PNAS, 2014]
l ! t l
l 1
l 1
t l
l !
UCDAYIS CIVIL AND ENVIRONMENTAL ENGINEERING & AQRC 10
A low-NOx B high-NOX
80 60 60 40
40
20 20
0 0
80 60 60
40 40 - 20 20 ns
C") Q) I 0 0 E i...
<( 0) 80 60 Q) :::l ..._... 60 0
40 ns <5 40 ·1:
0 20 20 ::::s ti)
0 0 i::, Q)
80 60 Q)
60 ti)
40 40
20 20
0 0
80 60 60
40 40 20 20
0 0 0 5 10 15 0 5 10 15 20
Reaction Time (hours) Reaction Time (hours)
Vapor Wall Losses
• As seed SA ↑, SOA formed ↑
• Impact on both low-and high-NOx
• SOA formation in chambers suppressed relative to actual atmosphere
• Developed new SOM parameterizations that account for vapor wall losses
11
Part 1c: Improving 3D models of SOA formation
• Merge SOM with SAPRC Gas-Phase Mechanism
• SAPRC determines oxidant concentrations
• SOM driven by SAPRC [OH]
• Reactions of non-precursor SOM species do not impact oxidant concentrations
• SOM mechanism generator for FORTRAN code generation and integration into UCD-CIT
• Run low-NOx (high yield) and high-NOx (low yield) simulations separately
• Compare with “base” SOA model and a “cascading” oxidation model
[Jathar et al., GMD, 2015]
• Consider with/without vapor wall losses
6 '-a, 5 .c E ::J 4 z C: a, 3 C) >->< 2 0
1
0
1 2
functional ization
OH
fragmentation
functional ization
3 4 5 6 7 8 9 10 11 12 13
Carbon Number
UCDAYIS CIVIL AND ENVIRONMENTAL ENGINEERING & AQRC 12
The “Base” and “Cascading” Oxidation Models BASE
• SOA formation as in CMAQ 4.7
• Primarily two-product approach
𝑉𝑂𝐶 + 𝑂𝐻 → 𝛼1 ∙ 𝑃1 + 𝛼2 ∙ 𝑃2
• Lacks multi-generational aging
• Products partition between gas and particle phases
• Includes irreversible particle-phase oligomerization (tolig ~ 1 day)
[see Pankow (1996); Odum et al. (1999); Carlton et al. (2010)]
COM
• Reactive two-product model
• Products react (“age”) to form low/non-volatile species
𝑉𝑂𝐶 + 𝑂𝐻 → 𝛼1 ∙ 𝑃1 + 𝛼2 ∙ 𝑃2
𝑃1 + 𝑂𝐻 → 𝑃2
𝑃2 + 𝑂𝐻 → 𝑃𝑁𝑉
• Aging reactions are unconstrained
• Greatly enhanced SOA yields, compared to 2P model
[see Lane et al. (2008)]
[Jathar et al., GMD (2015); ACP (2016)]
UCDAYIS CIVIL AND ENVIRONMENTAL ENGINEERING & AQRC 13
SOM* versus Base
Generally similar predicted [SOA], with larger differences away from sources
*No accounting for vapor wall losses [Jathar et al., ACP (2016)]
(a) BaseM
(d) 5
4
3
2
1
0
2
1.5
1
0.5
0
(b)
(e)
SOM 2
1.5
1
0.5
0 5
4
3
2
1
0
(c) SOM/BaseM 1.4
1.2
1
0.8
0.6 1.4
1.2
1
0.8
0.6
UCDAYIS CIVIL AND ENVIRONMENTAL ENGINEERING & AQRC 14
SOM* versus Base
• Some D in predicted precursor contributions • Substantial difference in predicted SOA volatility Base SOA more • Notable biogenic contribution sensitive to dilution or temperature changes than SOM SOA
(a) Los Angeles 1,---:-..:.._____,_-~~ I
M° 0.8 E rn 0.6 3 <( 0.4 0 CJ)
8
..--.. 6 C')
E 0)
3 4 <(
0 CJ) 2
(b) Riverside 1~....:..........-~~
0.8
0.6
0.4
6
4
■ Alkane SOA ■ Aromatic SOA ■ lsoprene SOA
Terpene SOA Sesquiterpene SOA
C 0
:.::; u co I...
LL V, V, co ~
0.8 - Base --- SOM
0.6
0.4
0.2
o.o L~-----~~~_2~~c...~ 0~--~2~
-4 -3
log C* (µgm )
UCDAYIS CIVIL AND ENVIRONMENTAL ENGINEERING & AQRC
*No accounting for vapor wall losses [Jathar et al., ACP (2016)] 15
SOM* versus COM
• COM-type models predict substantially higher SOA
• Results from forcing all reaction products into particle phase
• COM “double counts” SOA formation potential
• Use of COM-type models may improve model/measurement agreement…but likely for the wrong reasons
*No accounting for vapor wall losses [Jathar et al., ACP (2016)]
(c) SOM
(f) SOM
5 4 3 2 1 0
16 14 12 10 8 6 4 2 0
(b) COM
(e) COM
5 4 3 2 1 0
16 14 12 10 8 6 4 2 0
UCDAYIS CIVIL AND ENVIRONMENTAL ENGINEERING & AQRC 16
Accounting for Vapor Wall Losses
Consider two estimates of the impact of vapor wall loss: Low and High loss
• Substantially increased SOA everywhere when VWL considered
• Greater increase outside of major source regions
(a) ~
2
1.5
1
0.5
0 5
4
3
2
1
0
Ratio (low/no) 9
7
5
3
1 9
7
5
3
1
Ratio (high/no)
UCDAYIS
14
11
8
5
2 14
11
8
5
2
CIVIL AND ENVIRONMENTAL ENGINEERING & AQRC
[Cappa et al., ACP (2016)]
17
Accounting for Vapor Wall Losses
• Largest increases observed where [SOA] is smallest
• Leads to reduction in urban-rural SOA difference
Ratio between [SOA] with/without accounting for VWL
low-VWL / no-VWL high-VWL / no-VWL
low-VWL / no-VWL high-VWL / no-VWL
SoCAB
2
10 ·@ ■
~ ~ .. ~~~e. : : . ~ 4 • Ei:I ..
3 jj'f!li:p. : : . 2
1 (a) .1
0.0 0.2 0.4 0.6 0.8 1.0 1.2
10
j : 5
4
3
2
1 (b)
0 1
Eastern US
2 3 4 5
UCDAYIS CIVIL AND ENVIRONMENTAL ENGINEERING & AQRC
[SOA] (mg m-3)
18
Accounting for Vapor Wall Losses
Fraction of total OA that is SOA No VWL Low VWL High VWL
• Change from POA dominated to SOA dominated
• Substantially improves comparison with observational estimates
1 1 0.8 0.8 0.6 0.6 0.4 0.4
(a) I:, 0.2 (b) I:>
0.2 0 0 1 1 0.9 0.9 0.8 0.8 0.7 0.7 0.6 0.6 0.5 0.5 0.4 0.4 0.3 0.3 0.2 0.2
1 0.8 0.6 0.4 0.2 0 1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2
UCDAYIS CIVIL AND ENVIRONMENTAL ENGINEERING & AQRC 19
Accounting for Vapor Wall Losses
• Much improved model/ measurement agreement in predicted OA concentration and diurnal behavior when VWL considered
• Substantial improvement in predicted O:C of the OA – related to improved SOA/OA
-.... I 160 E 0. 0.
7 120 E 0)
3 80 ,........, 0 ~ 40 -,........, <( 0 0 ........
0 0.6 :;::; co
0::: (.)
.E 0.4 0 -<x:: ()
0 0.2
0
SOM-no
(a) 160
Measured [CO¥ - 0.105ppm 120 - 0.125ppm - - · 0.085 ppm
80
- 40
0 5 10 15 20
0.6
0.4
0.2
(a)
SOM-low
(b) 160
(c)
120
80
- 40
0 0 5 10 15 20 0 5 10 15 20 -
0.6
0.4
0.2
(b) (c) 0.0 ------------ 0.0 ------------ 0.0 --------------
0 5 10 15 20 0 5 10 15 20 0 5 10 15 20 Hour of day (local) Hour of day (local) Hour of day (local)
UCDAYIS CIVIL AND ENVIRONMENTAL ENGINEERING & AQRC 20
Part 1: Improving SOA models: Summary
• Including multi-generational oxidation in a realistic way does improve simulation of SOA properties, but by itself has little impact on [SOA]
• Use of unconstrained COM-type models yield higher [SOA], but likely for incorrect reasons
• Losses of semi- and low-volatility vapors to walls of environmental chambers can bias SOA formation low
• Accounting for the influence of vapor wall loss during development of SOA parameterizations leads to substantial increases in predicted [SOA], and much improved model/measurement agreement
SOM-high 160 (c)
120
80
0 ......___.___...____...__....____, 0 5 10 15 20
Hour of day (local)
UCDAYIS CIVIL AND ENVIRONMENTAL ENGINEERING & AQRC 21
Part 1: Where Next?
• Account for partitioning of semi-volatile POA components, and oxidation of evaporated vapors, to improve spatial distributions
• Establish impact of recently recognized SOA precursors that are not traditionally considered, especially low volatility consumer products
• Understand how predicted SOA is likely to respond to future changes in the photochemical environment (e.g. NOx
versus HO2)
• Determine origin of wintertime SOA in San Joaquin Valley
160 (c)
120
80
o......_ ____________ ....._ ___ ___. 0 5 10 15 20
Hour of day (local)
UCDAYIS CIVIL AND ENVIRONMENTAL ENGINEERING & AQRC 22
Part 2: Updating modeling scenarios for reactivity assessment
• The impact of adding 1 kg of a given VOC to a representative urban atmosphere on O3 (kg-O3 per kg-VOC) is termed incremental reactivity
Δ𝑂3𝐼𝑅 =
Δ𝑉𝑂𝐶𝑖
• The maximum incremental reactivity (MIR) is the IR under higher NOx (urban) conditions
• Understanding MIR values provides guidance on control strategies where VOC control is a productive strategy to reduce ozone
• The representative urban atmosphere in which MIR values are determined for VOCs had not been updated since 1988
• There have been substantial changes (improvements) in urban air composition over the last 30 years!
UCDAYIS CIVIL AND ENVIRONMENTAL ENGINEERING & AQRC 23
UCDAYIS CIVIL AND ENVIRONMENTAL ENGINEERING & AQRC
Part 2: Updating modeling scenarios for reactivity assessment
• We have determined how updating the representative urban atmosphere to 2010 conditions impacts ozone reactivity, with updates to:
• Meteorology
• Emission rates
• Concentration of initial conditions
• Concentration of background species
• Composition of VOC profiles
• Box model scenarios are explored for 39 cities across the US
• Determined MIR scale for 1,233 individual compounds and compound-mixtures Work done by
• Developed city-specific, modernized VOC compositions, aka Melissa Venecek
“surrogate” profiles
24
US Cities Considered
• Initial selection based on 1988 IR scenarios
• Screened according to 1-h daily max O3 for 2010 for each city or region
• Added Fresno & Bakersfield, removed Chicago and Tampa
Tulsa •
San Diego
Dallas •
UCDAYIS CIVIL AND ENVIRONMENTAL ENGINEERING & AQRC 25
Updating Meteorological Conditions
• Used WRF (Weather Research and Forecasting) model to determine meteorological conditions during O3 event periods for each city
• Decrease in median temperature, planetary boundary layer height, and absolute humidity from 1988 to 2010
• Changes primarily driven by shift in season of maximum O3 from summer to spring (April) or fall (Oct.) for many cities
- 310 ::ic::
QI :; 300 ... 11) ... ~ 290 E (:!. 280
270
_4000 E -... ~3000 11) ..... ~ ~2000 C ::, 0
1!1000 ... 11) ... QI
~ 0 C.
;:;, E
25
tiD 20 E --~ 15 "C ·e f 10 (II ... :J o 5 "' ~ <t
c=:::J2010 Box and Wh isker -1988 Median Value
8 9 10 11 12 13 14 15 16 17
8 9 10 11 12 13 14 15 16 17
8 9 10 11 12 13 14 15 16 17 Hour
""'' v ,._,,..,,u ._,. v •n-,JMENTAL ENGINEERING & AQRC 26
Updating Emissions
• Decrease in per-capita emissions of non-methane organic compounds (NMOC), NOx and CO
• Shift in timing of emissions, reflects strong decrease in mobile source emissions
• Updates in biogenic emissions (e.g. isoprene) from improved mechanistic knowledge
3..00E-07 3.00IE-06
2.50E-07 2.SOIE-06 C: .2 6
C "' ..... 0 ,IQ ....
0 2-00E-07 2.00IE-06 e ~ ... I:'.! ~ ., LI.J L... E 0.. 1..SOE-07 1..SOIE-06 u~
II.I L o· u.:t= ::E 'e 0~ 1..00E-07 1..00IE-06 z-=a :E.._
~ ~ ~~ 5..00E-08 5.00IE--07 -E :;3E O..OOE+Ol O.OOIE+OO
8 :9 10 11 12 13 14 115 16 17
C: 6.00E-07
0 5.00E-07 I;; ~
~ :!{_ 4.00E-07 "'L E f 3.00E-07
~ S !. 2.00E-07
~ ~ ~ 2 I 1..00E-07
E O.OOE+OO
8 9 10 11 12 13 14 15 15 17
~ 3..SOE-05
0 3.00E-05 ~
§ x_ 250[-05
~ ~ 200[-05
'5 i 1..SOE-05
8 ~ 1..00E-05
E 5.00E-07 a E O.OOE-+00
s 9 10 11 12 13 14 15 16 17
c: 3.50[-07
.2 ~ 3.00E-07
6 □ ~ X. 2.50[-07
r ~ .S l= 2.00E-07
l 1 ! 1: 1..50[-07
l l l 1 ~ ~ 1..00E-07
g E 5.00E-08 -e O.OOE+OO
8 9 10 11 12 13 14 15 16 17
5 Hour
CIVIL AND ENVIRONMENTAL ENGINEERING & AQRC 27
Updating Background and Ground VOC Composition
• Calculate concentrations of each compound using UCD-CIT air quality model
• Group by functional group to illustrate major shifts
• Smaller alkane, alkene, aromatic contribution in 2010
• Larger alcohol and ketone contributions
• Reflects shift towards increased biogenic contribution
Ground VO Composition Profile
0% 0%
■ Alkane ■ Alkene ■ Acetylene ■ Aldehyde
■ K ton ■ Ar ■ Al hol ■ Ot h r
1988Aloft voe Comp sitio Profil
4% 0% 0% 0%
■ Alkane ■ Alkene ■ Acetylene ■ Aldehyde
■ Ketone ■ Aromatic ■ Alcohol ■ Ot er
2010 Ground voe Composition
5%
1%
■ Al ane ■ Alkene ■ Acetyle • Aide de
■ K o ■ Ar m ti ■ Alo I ■ 0th r
2010 Aloft VO Cor position Profile
6%
1%
■ Al ane ■ Alkene ■ Ace lene ■ Aide de
■ Ke one Aromatic ■ Alco ol ■ Other
UCDAYIS CIVIL AND ENVIRONMENTAL ENGINEERING & AQRC 28
Updated MIR Values
• Overall decrease in MIR values from 1988 to 2010 by 20%
• Larger changes overall to compounds with smaller MIR values
2 I ~ 1.8 .... -~ 1.6 "O
~ 1.4
---Q) 1.2 ~ a:: ~ ~ 1
a:: 2 0.8 0 .... 0.6 0 N C: 0.4 .!!:! "O Q) 0.2 2
0 0
•
• •
♦ MIR Ratio
- 1:lline
•
5 W ~ W Median 1988 MIR value (g voe/ g 03)
25
30
-m 0 t),() 25 ........ u 0
~ 20 Q) :::,
~ 15 ex:: ~
S 10 0 N C: ro -0 Q)
~
5
0
0
♦ MIR Value
-- 1:lLINE
♦
5
♦ ♦ ♦
♦
10 15 20
y = 0. 7967x + 0.3977 R2 = 0.9783
25 30
Median 1988 MIR value (g voe/ g 03 )
UCDAYIS CIVIL AND ENVIRONMENTAL ENGINEERING & AQRC 29
Understanding the Change in MIR Values
• Largest change results from updated meteorology
• Shift in seasons and lower PBL
• Notable impact of decrease of NOx relative to VOC
• Small influence of change in background VOC profile, and background reactivity of the atmosphere
Meteorology
Emissions
Aloft VOC Profile
30 C1J :::J
o £: 25 8 er:
~ 2c --;,20 .., ro 0 -~ 'i:i 0.0
.Q C1J ........ 15 ... 2 u ~ 0 C1J > > tX ~ ..!:el0 00 0 00 ... en o ..... ./:l 5
C1J 2
0
30 ..!:e
o ~ 25 ...., ro
~ >
~ ~ --;,20 ?i: C Q o ro 0.0 ·;:: 'i:i ......._ 15 ro C1J U ~20 tX.,.,>10 00 C 00 0 en ·.;; ..... V) 5
E LU
u uO o> > ..!:e
0
30
25 0 C1J
8 2 20 N £; £ er: ~ -~ ~ o"\5 0 C 0.0 -~ ro -.....
~ ] 10 tX 2 gg -9:! 5 en .;:::: ..... 0 ...
a.. 0
0
0
0
--1:1 LINE
5
--1:1 LINE
5
--1:1 LINE
5
y = 0.7838x - 0.1765 R2 = 0.9944
10 15 20 1988 MIR value (g voe/ g 0 3 )
♦ ♦ ♦
10 15 20 1988 MIR value (g voe/ g 0 3 )
10 15 20 1988 MIR va lue (g voe/ g 0 3 )
25
y = 0.8567x + 0.754 R2 = 0.956
25
y = 0.942x + 0 .1551 R2 = 0.9969
25
30
30
30
30
Ranking Compounds by MIR
• Limited changes in overall Highest MIR
ranking of compounds by MIR
• Only 15 compounds changed by >30%
• Compounds that had highest O3
formation potential in 1988 continue to have highest potential in 2010
Lowest MIR
1200
1000
~ C 800 C'O
0:::: u 0 600 > 0 T""f 0
400 • N
200
0
0 200
•
400 600 800
1988 voe Rank
•
y = 0.9951x + 3.0116 R2 = 0.9903
1000 1200
UCDAYIS CIVIL AND ENVIRONMENTAL ENGINEERING & AQRC 31
Part 2: Updated Model Scenarios Summary
• Updated emissions, meteorology, concentrations of reaction scenarios for assessment of VOC incremental reactivity
• MIR values declined by 20% from 1988 values due to changes in meteorology, emissions (NOx/VOC) and background VOC composition
• Compounds that had highest MIR values in 1988 continue to have highest values in 2010
-;;; 0
~ 25 u 0
~ 20 Cl/
-= ro > 15
0:::
2 ~ 10 0 N C: .!!! 5 "U Cl/
2
0 0
♦ MIRValue
- 1:lLINE
5
♦ ♦
♦
y = 0. 7967x + 0.3977 R2 =0.9783
10 15 20 25
Median 1988 MIR value (g voe/ g 0 3 )
30
UCDAYIS CIVIL AND ENVIRONMENTAL ENGINEERING & AQRC 32
Part 3: Evaluate organic nitrate and N2O5 chemical mechanisms and assess their impact on secondary aerosol formation
• Explicit reactions between biogenic VOCs and NOX were added to the SAPRC photochemical mechanism to evaluate SOA and organic nitrates
• 271 reactions were modified or added, following Pye et al. [ES&T, 2015]
• Includes: gas-phase production, gas-particle partitioning, hydrolysis
• Evaluate for two periods/seasons/locations:
• SoCAB during CalNEX (June/July 2010)
• SJV during DISCOVER-AQ (Jan/Feb 2013)
From Pye et al. [2015]
BVOC+ OH, 0 3
(+ INO)
BVOC+ IN03
gas
organic nitrate
deposition
organic nitrate
T=3 h i organic + H N03
UCDAYIS CIVIL AND ENVIRONMENTAL ENGINEERING & AQRC 33
Part 3: Impact of Organic Nitrate Formation
• Larger influence of expanded mechanism predicted for summer than winter
• Monoterpene nitrates and glyoxal/methyl glyoxal species make largest contributions
• PM2.5 SOA concentrations of these species approached 1 µg/m3 during summer, but were < 0.1 µg/m3 during winter
Potential Influence of NOx Control
• Monoterpene nitrates expected to decrease directly proportional to NOx
• Glyoxal/methyl glyoxal SOA expected to decrease slower than NOx
• For summertime, a 25% ↓ in NOx could produce a 0.13 mg/m3 ↓ in PM2.5 SOA
UCDAYIS CIVIL AND ENVIRONMENTAL ENGINEERING & AQRC 34
Acknowledgements
Prof. Shantanu Jathar (CSU-Fort Collins, formerly UCD)
Melissa Venecek (ARB, formerly UCD)
Dr. Xuan Zhang (NCAR, formerly Caltech)
Dr. Renee McVay (EDF, formerly Caltech)
Dr. Joseph Ensberg (UTC Aerospace Systems, formerly Caltech)
Prof. Bill Carter (UC Riverside)
Prof. Jose Jimenez (CU Boulder)
Dr. Nehzat Motallebi (CARB)
This study was funded by the California Air Resources Board, contract 12-312. The statements and conclusions in this report are those of the PI’s and not necessarily those of the California Air Resources Board.
UCDAYIS CIVIL AND ENVIRONMENTAL ENGINEERING & AQRC 35
Products Influence of vapor wall loss in laboratory chambers on yields of secondary organic aerosol. Xuan Zhang,
Christopher D. Cappa, Shantanu H. Jathar, Renee C. McVay, Joseph J. Ensberg, Michael J. Kleeman, and John
H. Seinfeld, PNAS April 7, 2014
Simulating secondary organic aerosol in a regional air quality model using the statistical oxidation model – Part 2:
Assessing the influence of vapor wall losses
Christopher D. Cappa, Shantanu H. Jathar, Michael J. Kleeman, Kenneth S. Docherty, Jose L. Jimenez, John H.
Seinfeld, and Anthony S. Wexler, Atmos. Chem. Phys., 16, 3041-3059, 2016
Simulating secondary organic aerosol in a regional air quality model using the statistical oxidation model – Part 1:
Assessing the influence of constrained multi-generational ageing. S. H. Jathar, C. D. Cappa, A. S. Wexler, J. H.
Seinfeld, and M. J. Kleeman, Atmos. Chem. Phys., 16, 2309-2322, 2016
Multi-generational oxidation model to simulate secondary organic aerosol in a 3-D air quality model. Jathar, S. H.
and Cappa, C. D. and Wexler, A. S. and Seinfeld, J. H. and Kleeman, M. J. Multi-generational oxidation model to
simulate secondary organic aerosol in a 3-D air quality model. Geoscientific Model Development, 8 (8). pp. 2553-
2567, 2015.
Updating the SAPRC Maximum Incremental Reactivity (MIR) scale for the United
States from 1988 to 2010. M.A. Venecek, W.P.L. Carter, and M.J. Kleeman. Journal of the Air & Waste
Management Association, in press, 2018.
UCDAVIS CIVIL AND ENVIRONMENTAL ENGINEERING & AQRC 36