1 of 16

Global simulation of H₂ and HD

with GEOS-CHEM

Heather Price¹, Lyatt Jaeglé¹, Paul Quay²,

Andrew Rice², and Richard Gammon²

University of Washington, Seattle

Departments of ¹Atmospheric Sciences and ²Oceanography

2^nd GEOS-CHEM Users Meeting 6 Apr 2005

2 of 16

Sinks (Tg/yr) MOZART Novelli GEOS-CHEM

OH^c 15 19 17

Soils^c 55 56 59

Total 70 75 77

Sources (Tg/yr) MOZART^a Novelli^c GEOS-CHEM^d

Hauglustaine

Fossil Fuel 16 15±10 20

Biomass Burning 13 16±5 10

Biofuel 5^b 4.4

Photochemical 31 40 41

Methane Oxidation 26 ± 9 27

BVOC Oxidation 14 ± 7 14

Ocean 5 3 ± 2 ~

N fixation 5 3 ± 1 ~

Total 70 77 76

^aHauglustaine et al., 2002; Photochemical production includes Methane(27.5Tg) and nonmethane hydrocarbons (14.2Tg): Isoprene, Acetone, Monoterpenes, and Methanol.

^bAndreae & Merlet, 2001: bf H₂/CO = 0.32 per molecule

^cNovelli, 1999: bb H₂/CO = 0.29, for fossil fuels Novelli uses global CO source of 500Tg/yr from Logan et al., 1981, Pacnya & Graedel, 1995 and WMO, 1995

Lifetime, years 1.9 2-3 2.1

Annual Global Budget of Molecular Hydrogen in the Troposphere

3 of 16

H₂and HD in the GEOS-CHEM Model

Based on the GEOS-CHEM offline CO simulation

v5.05.04

Sinks

OH^d H₂ + OH → H₂O + H k = 1.5x10 ^-13 e^-2000/T

Soils Uniform Deposition Velocity over land = 0.042 cm/s

Sources H₂/CO (per molecule)

Fossil Fuels 0.59^a

Biomass Burning 0.30^c

Biofuels 0.32^b

Photochemical yield relative to CO

Methane Oxidation 0.50

BVOC Oxidation 0.50

^aOliver et al., 1996 CO emission inventory EDGAR

H₂/CO (per molecule) = 0.588 or 0.042Tg H₂/CO

^bAndreae & Merlet, 2001: bf H₂/CO = 0.32 or 0.023Tg H₂/CO

^cNovelli, 1999; bb H₂/CO= 0.30 or 0.022Tg H₂/CO

^dJPL reported average of nine studies detailed in Ravishankara et al., 1981 and in excellent agreement with measurements by Talukdar et al., 1996.

^k

4 of 16

H₂ ppbv

GEOS-CHEM Simulation of H₂

Surface (JJA)

Surface (DJF)

5 of 16

Validating the GEOS-CHEM H₂ simulation against CMDL H₂ Observations

CMDL sites

Surface (JJA)

CMDL sites

H₂ ppbv

Surface (DJF)

(Novelli, 1999)

Climate Monitoring and Diagnostics Laboratory: ftp://140.172.192.211/ccg/h2/flask/

6 of 16

Fall % Bias: -0.86

R: 0.71

Summer % Bias: 0.71

R: 0.80

Winter % Bias: 1.25

R: 0.67

Spring % Bias: 0.70

R: 0.56

Latitude

H₂ ppbv

H₂ Interhemispheric Gradient

~40 ppbv

gradient

GEOS-CHEM H₂ ppbv

GEOS-CHEM H₂ simulation

vs. CMDL observations

GEOS-CHEM model

NOAA CMDL observations (1989-2003)

CMDL H₂ ppbv

-90 -50 0 50 90

400 450 500 550 600

600

550

500

450

400

600

550

500

450

400

Spring

Summer

Autumn

Winter

Correlation (r=0.76)

model-obs

obs

Bias: x100 = 0.45%

7 of 16

H₂Seasonal Cycle

Barrow (89-03) Bermuda(91-03) Mauna Loa(89-03)

40.7 S, 144.7 E

Model

CMDL observations

Ascension (89-03) Cape Grim(91-03) Palmer Station(94-03)

Northern

Hemisphere

Southern

Hemisphere

H₂ ppbv

Month

2 4 6 8 10 12

Month

2 4 6 8 10 12

7.9 S, 14.4 W

Month

2 4 6 8 10 12

Month

2 4 6 8 10 12

Month

2 4 6 8 10 12

Month

2 4 6 8 10 12

H₂ ppbv

650

600

550

500

450

400

650

600

550

500

450

400

71.3 N,156.6 W

32.4 N, 64.7 W

19.5 N, 155.6 W

64.9 S, 64.0 W

Here’s the seasonal cycle for H2 at a number of sites in the N and S hemispheres. The model is in red and CMDL obs representing all years in black. The red points represent the max and min for surrounding grid boxes in the model to give a sense of the variability of the model.

Overall the seasonal cycle seems to be controlled by photochemical source, which is slightly stronger in the northern hemisphere compared to the SH. At Barrow the seasonal cycle is also affected by the soil sink. Though it looks like for the model, the Northern hemisphere soils sink is too strong in the spring & summer. We’re not including soil moisture, which accd to field & lab studies reduces the soil deposition rate, and this may account for this dip we see in spring/summer. This just reinforces that the soil sink really is a difficult part of the budget to constrain.

Overall this is a reasonable simulation so our next step was to add deuterium to the model.

8 of 16

H₂Vertical Profiles Nov 2002-Aug 2004

Park Falls, Wisc.

45.93N,-90.27W

H₂ (ppbv)

400 500 600

4

2

0

km

Poker Flat, Alaska

65.07N, -147.29W

400 500 600

H₂ (ppbv)

Sept

Oct

Nov

March

April

May

Cook Islands

-21.25S, –159.83W

400 500 600

H₂ (ppbv)

km

4

2

0

km

Soil

Model

Observations

Here’s vertical profiles for H2 at a few sites in the N and S hemispheres. The model is in red and CMDL obs representing all years in black. The red points represent the max and min for surrounding grid boxes in the model to give a sense of the variability of the model.

The altitude gradient is due to uptake of H2 by soils at the surface. At Barrow the model seems to underestimate H2 compared with obs, with greater difference at the surface, probably due to an underestimate in soil uptake during autumn.

Overall the seasonal cycle seems to be controlled by photochemical source, which is slightly stronger in the northern hemisphere compared to the SH. At Barrow the seasonal cycle is also affected by the soil sink. Though it looks like for the model, the Northern hemisphere soils sink is too strong in the winter and spring months and slightly too weak during summer. This just reinforces that the soil sink really is a difficult part of the budget to constrain.

Overall this is a reasonable simulation so our next step was to add deuterium to the model.

9 of 16

Adding hydrogen isotope (HD)

to the GEOS-CHEM model

Model development based on measured ratios of HD/H₂ for various sources, sinks, and reservoirs

Will give additional constraint to the H₂ budget sources and sinks

Determine the contributions of sources and sinks

to atmospheric δD and interhemispheric gradient (Gerst & Quay, 2000, 2001)

10 of 16

Deuterium Source & Sink Signatures

Soil, fossil fuel, and biomass burning fractionation: Gerst & Quay, 2001

OH fractionation: Ehhalt et al., 1989

δD of the global Troposphere = 130 %o

Term H₂ Tg/yr δD%o α

Fossil Fuels 20 -196

Biomass Burning 10 -293

Biofuels 4.4 -293

Methane Oxidation 28 156

BVOC Oxidation 14 156

OH Sink 17 0.601

Soil Sink 60 0.943

11 of 16

JJA

δD (%0)_SMOW

H₂ppbv

Annual δD

Surface H₂ and δD

12 of 16

δD (%0 vs SMOW)

1998,2002,2004

Ocean Cruise

Observations

Barrow

Cheeka

Peak

DJF δD Model, Surface & Cruise Observations

Biofuels

& Fossil Fuels

13 of 16

δD vs. Latitude

α_sinks

δD (atmos)

~40 %0

gradient

δD Observational Data from Rice & Quay, 2004 and Gerst, & Quay, 2001.

Additional enrichment

from Stratosphere?

14 of 16

GEOS-CHEM captures well the H₂ and δD latitudinal gradient (H₂~40ppbv, δD~40%o) and seasonality.

Soil Sink uncertainty: incorporate soil moisture, precipitation, to better constrain soil deposition

Next, help explain the δD observations of stratospheric enrichment (Röckmann et al., 2003; Rahn et al., 2003)

Could δD measurements

be used to constrain

Asian biofuel emissions?

Summary

Biofuel

+ Fossil Fuel

Biomass

Burning

Fossil

Fuels

DJF δD

1 of 16

2 of 16

3 of 16

4 of 16

5 of 16

6 of 16

7 of 16

8 of 16

9 of 16

10 of 16

11 of 16

12 of 16

13 of 16

14 of 16

15 of 16

16 of 16