Comparison of OH in GEOS-Chem and GMI/Aura4

Jennifer Logan, Jingqiu Mao, Junhua Liu, Inna Megretskaia

HarvardBryan Duncan

NASA

GEOS-Chem V8-01-01 and GMI Combo/aura4 runBoth simulations used:• GEOS-4 met fields for 2005• Similar anthro. emissions• Same biomass burning emissions• Same MEGAN base rates – I think• Different lightning NOx scheme• Similar ozone columns (often within 5%)But:• Mean 0H is ~25% different, with GMI lower

OH dependence on precursors:

OH = ( 2k2[O1D][H2O] + [HO2] {k6[O3] + k7[NO]} + 2J9 [H2O2] )----------------------------------------------------------------------------------

( k3[CO] + k4[CH4] )

GMI(=JPL2006):O1D+N2: 2.15e-11exp(110/T) the sameO1D+O2: 3.3e-11exp(55/T)O1D+H2O: 1.63e-11exp(60/T)

GEOS-CHEM(v8-01-01):O1D+N2: 2.14e-11exp(110/T) the sameO1D+O2: 3.2e-11exp(70/T)O1D+H2O: 1.45e-11exp(89/T)

GMI: FAST-Jx

GEOS-Chem (v8-01-01):FAST-J from 1999.

GEOS-Chem (v8-02-01):FAST-JX 2008 input≈GMI input for J valuefor O(1D)

Difference in O(1D) + O2 is 2-3.5%, but O(1D) + N2 is the same so <0.7% diff. in O(1D) + N2, O2. At most 2% difference in O(1D) + H2O.

This doesn’t account for the differences we see in OH.

220 240 260 280 3003.95

4

4.05

4.1

4.15

4.2

4.25

4.3

4.35

4.4x 10-11

Temperature(K)

220 240 260 280 3001.95

2

2.05

2.1

2.15

2.2

2.25x 10-11

Temperature(K)

GMI(O1D+O2)GC(O1D+O2)

GMI(O1D+H2O)GC(O1D+H2O)

Rates for O(1D) quenching.

OH zonal mean, July 2005

1.2

1.4 1.6

Ratio: G-Chem/GMIMean = 1.24

GEOS-Chem GMI/Aura4

V8-01-01

Mean OH = 1.26 x 106

(1000-300 hPa)Mean OH = 1.02 x 106

(1000-300 hPa)

J value (O3 + hv -> O(1D) + O2) July 2005

Column within 5%

Ozone column used for J-value calc.

1.21.1

1.0

GEOS-Chem GMI/Aura4

Updated input to Fast-J so ≈the same as in GMI

Column within 5%

Ozone column used for J-value calc.

1.21.1

1.0

Input updated to Fast-JX 2008

Why differences? Treatment of clouds? Aerosols? Test with clear sky calc.

Makes the problem worse

NOx July 2005GEOS-Chem GMI/Aura4

1.1

1.51.8 2.5 GEOS-Chem/GMI

LNOx = 0.76 Tg(0.33 Tg, 24N-32S) LNOx = 0.49 Tg

(0.31 Tg, 24N-32S)

OH ratio has a strong dependence on NOx

1.2

1.4 1.6

1.21.1

1.0

1.1

1.51.8 2.5

OH ratio

CO ratioJ value ratio

NOx ratio

0.90.95 0.85

July 2005

Ratio plots for Oct. 2005 (not much LNOx in extratropics)

Column within 5%

OH NOx

J value

Merged ozone data set used for GEOS-Chem

New versions of GEOS-Chem

Mean OH in benchmark runs:

V7-04-13: 12.1 x 105

V8-01-04: 11.1 x 105

V8-02-01: 11.8 x 105

JPL-2006, new FAST-J input,new emissions

Old rates and FAST-J input

GEOS-5.1.0 GEOS-5

GEOS-4

Last benchmark on GEOS-4,old emissions

GEOS-5 runs used latest emissions, MEGAN rates, difference is chemistry update

Summary, GMI vs. GEOS-Chem

• Need to resolve difference in J value for O(1D) production, as both models use the same method – FAST-J

• Compare clear sky calc. to see if cloud treatment is the cause

• Compare with zero aerosols• Do both within each model• Clearly, lightning NOx also drives much of

the difference