CMAQ Model Performance Evaluation with the updated CB4-2002

Center for Environmental Research and Technology/Air Quality Modeling

University of California at Riverside

CMAQ Model Performance Evaluation with the updated

CB4-2002

Chao-Jung Chien, Gail Tonnesen, Bo Wang Tiegang Cao, Zion Wang,

Mohammad Omary, Youjun Qin

University of California – Riverside, CE-CERT

Models-3 CMAS Workshop, Oct.27-29, 2003, RTP, NC



CB4 mechanism update

• Current regulatory version of CB4 (Gery et al. 1989) has modifications but is outdated.

• New updated CB4, CB4-2002, released by Jeffries et al. in late 2002

– Contains 100 principle reactions and same CB4 model species (excluding secondary organic aerosol precursor species and aqueous species)

– Fully re-evaluated with smog chamber data

– Better documentation; the old CB4 in which many of the changes since 1989 have not been well documented.

– Four versions of “swappable” olefin chemistry. Recommended replacement version is used.

.! KINETICS DATA SOURCES:!'97 NASA DOCUMENT:!"CHEMICAL KINETICS AND PHOTOCHEMICAL DATA FOR USE IN! STRATOSPHERIC MODELING," DEMORE, ET AL., JANUARY 15,

! 1997, JPL PUBLICATION 97-4.! '97 IUPAC DOCUMENT:! "EVALUATED KINETIC AND PHOTOCHEMICAL, AND! HETEROGENEOUS DATA FOR ATMOSPHERIC CHEMISTRY:! SUPPLEMENT V," ATKINSON, ET AL.,! J. PHYS. CHEM. REF DATA, VOL 26, NO 3 AND NO 6, 1997...<IB5F> NO3 + NO2 = N2O5 # 2.00E-30^-4.4&1.40E-12^-0.7; <IB5R> N2O5 = NO3 + NO2 # [email protected]*E<IB5F> ;! IB5F: NASA00, T2, T3; IB5R KEQ=3.0E-27@-10991/T).. !<IB6> AND<IB7> HOMOGENEOUS HYDROLYSIS ONLY; WAHNER ET AL, GRL, 25(12):2169, 1998 <IB6> N2O5 + H2O = 2.0*HNO3 # 2.5E-22;<IB7> N2O5 + H2O + H2O = 2.0*HNO3 + H2O # 1.8E-39;



Major revisions in CB4 mechanism update• Updated photolysis rates; applying 12 photolysis rates, an expansion from current 7 (6

original) photolysis rates used in CB4.

Reaction CB4 CB4_2002

O3 = O3P 0.053/<NO2_CBIV88> J[O3_to_O3P]

O3 = O1D 1.0/<O3O1D_CBIV88> J[O3_toO1D]

NO3 = 0.89*OH + 0.89*O +0.11*NO 33.9/<NO2_CBIV88> J[NO3_NO] & J[NO3_NO2]

HONO = OH+ NO 0.1975/<NO2_CBIV88> J[HONO_NO]

HONO = HO2 + NO2 J[HONO_NO2]

H2O2 = 2*OH 0.255 /<HCHOmol_CBIV88> J[H2O2_OH]

FORM = 2*HO2 + CO 1.0/<HCHOrad_CBIV88> J[HCHO_HO2]

FORM = CO 1.0/<HCHOmol_CBIV88> J[HCHO_H2]

ALD2 = XO2 +2*HO2 + CO + FORM 1.0/<ALD_CBIV88> J[CH3CHO_HCO]

OPEN = C2O3 + HO2 + CO 9.04/<HCHOrad_CBIV88> J[CHCHO_HO2]

MGLY= C2O3 + HO2 + CO 9.64/<HCHOrad_CBIV88> J[CHCHO_HO2]

ISPD = 0.333*CO + 0.067*ALD2 + 0.9*FORM + 0.832*PAR + 1.033*HO2 + 0.7*XO2 + 0.967*C2O

30.0036/<ACROLEIN> 0.001/J[ACRO_RO2]



Major revisions in CB4 mechanism update (cont’d)

• Additional homogeneous N2O5 hydrolysis reaction are added:

– N2O5 + H2O + H2O 2HNO3 + H2O (Wahner et al. GRL, 1998)

ReactionCB4

(v.0301)CB4

(v.0602)CB4

(v4.3)CB4_02

N2O5 + H2O 2HNO3

k = 1.3x10-21 k = 2.6x10-22 k = 0.0 k = 2.6x10-22

N2O5 + H2O + H2O 2HNO3 + H2O

k = 1.8x10-39



• Reaction rate of OH + NO2 HNO3, recommended by Troe (2001) is used. (12% lower than current CMAQ rate)

* from Jeffries et al. (2002)


Comparison of four recommendations for the

reaction rate of

OH + NO2 (at 300K)*

92%

138%

100%112%

J. TROE, Int. J. Chem. Kin. 33, 2001

IUPAC

NASATroe

CMAQ




• Other major revisions made in CB4_2002 from old CB4– Reaction rates and product distributions are updated to

reflect the most recent research findings; changes also include:• New kinetic rate expression for HONO formation, and the

use of new cross-sections for the HONO dissociations.• Updated PAN formation and decay kinetics.• Revised rates and product yields for Olefin chemistry:

OLE/ETH + O3 (NO3, O3P) to be consistent with current literature.

• However, aromatic chemistry left unchanged.



Additional reactions added to CB4_2002

ADDED RXNS CB4_2002, k

N2O5 + H2O + H2O = 2HNO3 + H2O

1.8E-39

NO3 + OH = NO2 + HO2

2.2E-11

NO3 + HO2 = HNO3 + O2 9.2E-13

NO3 + NO3 = 2NO2 + O2 8.5E-13 @ 2450

O3 + O(3P) = 2O2 8E-12 @ 2058

HONO = HO2 + NO2 J[HONO_NO2]

OH + HO2 = O2 + H2O 4.8E-11 @ -250

“… some reactions are in the mechanism just for those conditions that are quite different than the ones in the outdoor chamber simulation…” Jeffries H.



Implementation of CB4_2002 with aerosol and aqueous extensions in CMAQ v.4.3

CB4_02_AE3_AQ

• CB4_2002 in “Morpho” language format, needs to be converted to CMAQ format

• Followed the changes made in cb4 for new version of CMAQ– New deposition velocity surrogates – New cloud scavenging surrogates– Eliminated advection and diffusion of fast-reacting species (e.g. OH,

HO2)– Modified gas-phase Monoterpene reaction rates (to be consistent with

those in SAPRC mechanism).

• With and without gaseous reaction rate constants for N2O5 hydrolysis (Sensitivity run : CB4_02 _zeroN2O5 )

• Process new photolysis rate tables (JTables)



Evaluation of CB4_2002

• CMAQ v.4.3; smvgear option– CB4_02_ae3_aq with gaseous N2O5 reactions (CB4_02)– CB4_02_ae3_aq zero out gaseous N2O5 reactions (CB4_02_zeroN2O5)

• WRAP 1996 January and July episodes• Model difference: CB4 vs. CB4_02 vs. CB4_02_zeroN2O5

– Daily average spatial plots and domain daily average time series plots for O3, HNO3, PAN, N2O5, ANO3, ASO4, and SOA

• Comparisons against ambient data: – Normalized Mean Bias (NMB%) and Mean Fractionalized Bias (MFB%)– AQS (AIRS): July hourly O3 – IMPROVE: daily NO3, SO4, OC– CASTNet: July daily HNO3, NO3, NH4, SO2, SO4



O3, CB4 vs. CB4_02 vs. CB4_02_zeroN2O5

January• Jan. daily average concentration comparison.

• Hourly variations as high as 40 ppb.

• CB4_02_zeroN2O5 produced highest avg. O3 concentration.

CB4 __CB4-02 __CB4-02-zeroN2O5 __

1st Layer Domain Daily Avg.



O3, CB4 vs. CB4_02 vs. CB4_02_zeroN2O5

July


CB4 __CB4-02 __

CB4-02-zeroN2O5 __

• July daily average concentration comparison.

• Higher O3 in urban areas. Hourly variations more than 100 ppb in some areas.

• CB4 produces highest average O3, while CB4_02 least.



HNO3, CB4 vs. CB4_02 vs. CB4_02_zeroN2O5

January•Jan. daily average concentration comparison.

•CB4_02 produced more than 1 ppb HNO3 than CB4.

•CB4 and CB4_02_zeroN2O5 produces similar HNO3.


CB4 __CB4-02 __

CB4-02-zeroN2O5 __



HNO3, CB4 vs. CB4_02 vs. CB4_02_zeroN2O5

July

CB4 __CB4-02 __

CB4-02-zeroN2O5 __



• CB4_02 higher in urban areas. Hourly variations as high as 10 ppb.

• CB4_02 produces highest average HNO3, whereas CB4_02_zeroN2O5 least.



PAN, CB4 vs. CB4_02 vs. CB4_02_zeroN2O5

January

CB4 __CB4-02 __

CB4-02-zeroN2O5 __




PAN, CB4 vs. CB4_02 vs. CB4_02_zeroN2O5

July

CB4 __CB4-02 __

CB4-02-zeroN2O5 __




N2O5, CB4 vs. CB4_02 vs. CB4_02_zeroN2O5

January

CB4 __CB4-02 __

CB4-02-zeroN2O5 __




N2O5, CB4 vs. CB4_02 vs. CB4_02_zeroN2O5

July

CB4 __CB4-02 __

CB4-02-zeroN2O5 __




ANO3, CB4 vs. CB4_02 vs. CB4_02_zeroN2O5

January•Jan. daily average concentration comparison.

•CB4_02 produced as high as 5 ug/m3 of NO3 than CB4.

•CB4 and CB4_02_zeroN2O5 produces similar NO3.

CB4 __CB4-02 __

CB4-02-zeroN2O5 __




ANO3, CB4 vs. CB4_02 vs. CB4_02_zeroN2O5

July• July daily average concentration comparison.

• CB4_02 produced more than 2 ug/m3 of NO3 than CB4 in some areas.

• CB4_02 produced highest NO3 average; CB4 produced a little higher than CB4_02_zeroN2O5.

CB4 __CB4-02 __

CB4-02-zeroN2O5 __




ASO4, CB4 vs. CB4_02 vs. CB4_02_zeroN2O5


• CB4 produced less than 0.5 ug/m3 SO4 than CB4_02.

• CB4 produced slightly lower SO4 than CB4_02 and CB4_02_zeroN2O5.

1st Layer Domain Daily Avg.CB4 __

CB4-02 __CB4-02-zeroN2O5 __



ASO4, CB4 vs. CB4_02 vs. CB4_02_zeroN2O5

July


CB4-02 __CB4-02-zeroN2O5 __


• As high as 1 ug/m3 of difference between CB4 and CB4_02.

• Overall, CB4_02_zeroN2O5 produced similar SO4 avg. to CB4_02.



SOA, CB4 vs. CB4_02 vs. CB4_02_zeroN2O5


• No significance difference.

• All three produced similar avg. conc. of SOA.


CB4 __CB4-02 __

CB4-02-zeroN2O5 __



SOA, CB4 vs. CB4_02 vs. CB4_02_zeroN2O5

July• July daily average concentration

comparison.

• CB4 produced slightly higher than CB4_02.

• CB4_02_zeroN2O5 produced similar SOA avg. to CB4_02.


CB4-02 __CB4-02-zeroN2O5 __



AQS, CB4 vs. CB4_02 vs. CB4_02_zeroN2O5

O3_July

CB4 CB4_02 CB4_02_zeroN2O5

NMB(%) -3.6 -5.5 -3.9

MFB(%) 5.1 2.6 4.3



IMPROVE, January, Normalized Mean Bias

-10.0

0.0

10.0

20.0

30.0

40.0

50.0

60.0

No

rma

lize

d M

ea

n B

ias

(%

)


CB4 17 41 15 -5

CB4_02 21 57 23 -4

CB4_02_zeroN2O5 21 41 17 -4

SO4 NO3 NH4 OC



IMPROVE, July, Normalized Mean Bias

-60.0

-50.0

-40.0

-30.0

-20.0

-10.0

0.0

No

rma

lize

d M

ea

n B

ias

(%

)


CB4 -29 -50 -37 -30

CB4_02 -28 -44 -35 -31

CB4_02_zeroN2O5 -28 -53 -36 -31

SO4 NO3 NH4 OC



CASTNet, July , CB4 vs. CB4_02 vs. CB4_02_zeroN2O5

G_HNO3 P_NO3 T_NO3CB4 CB4_02

CB4_02(zeroN2O5)

CB4 CB4_02CB4_02

(zeroN2O5)CB4 CB4_02

CB4_02(zeroN2O5)

NMB(%) 12 19 8 -61 -52 -62 -5 3 -8

MFB(%) 4 7 0.2 -103 -96 -106 -12 -9 -16

G_SO2 P_SO4 P_NH4CB4 CB4_02

CB4_02(zeroN2O5)

CB4 CB4_02CB4_02

(zeroN2O5)CB4 CB4_02

CB4_02(zeroN2O5)

NMB(%) 48 48 48 -31 -30 -29 -14 -11 -12

MFB(%) 28 28 28 -41 -40 -39 -25 -23 -23



Summary of CB4_2002 modeling results • Major difference found for O3, HNO3, ANO3

– Higher O3 concentration in urban area but lower in rural area. (probable cause: combined effect of OH+NO2 rate, less long range transport of PAN formation)

• Comparisons with ambient data: – Improvements are mixed and relatively small

• Should gaseous N2O5 reactions be taken out?• Would it sensitive to emission changes (control runs)?Next steps…• Process analysis to identify key contributors to the changes• Implement EBI/MEBI (Modified Euler Backward Iterative)

solver for CB4_02 • Implement paraffin secondary organic aerosol into CB4_02.• Improve aromatic chemistry mechanism.



Summary of CB4_2002 modeling results

• Bottom line…– In spite of relatively small changes comparing the

ambient data, CB4_2002 has better science and should be included in future CMAQ release.

CMAQ Model Performance Evaluation with the updated CB4-2002

Documents