Dylan Millet, Daniel Jacob, and May Fu

Harvard University

​

Thomas Kurosu and Kelly Chance

Harvard-Smithsonian Astrophysical Observatory

​

Alex Guenther

NCAR

​

Colette Heald

UC Berkeley

​

OMI HCHO measurements to test isoprene emissions and land cover

GEOS-Chem Meeting

April 11, 2007

HCHO Columns Map Isoprene Emissions from Space

Formaldehyde (HCHO) is a major breakdown product in the oxidation of isoprene

🡺 Giving us an isoprene emission proxy that can be measured from space

Relating HCHO Columns to Isoprene Emission

VOCs

HCHO

OH, hν

k_i, Y_i

OH, hν

k_HCHO

Local Ω_HCHO-E_i Relationship

Palmer et al., JGR (2003,2006)

Millet et al., JGR (2006)

HCHO vertical columns measured by OMI

(Summer 2006)

Relating HCHO Columns to Isoprene Emission

What drives variability in column HCHO?

Measured HCHO production rate vs. column amount

Isoprene dominant source when Ω_HCHO is high

Ω_HCHO variability over N. America driven by isoprene

Other VOCs give rise to a relatively stable background Ω_HCHO

🡪 Not to variability detectable from space

INTEX-A

Millet et al.,

JGR (2006).

Measuring HCHO from Space

1995-2001

40 x 320 km footprint

global coverage: 3 days

GOME

2002-present

30 x 60 km footprint

global coverage: 6 days

SCIAMACHY

2004-present

13 x 24 km footprint

global coverage: 1 day

OMI

SBUV instruments in low Earth orbit

OMI HCHO Columns over North America

OMI (2006) vs. GOME (1996-2001)

OMI: better counting statistics, cloud filtering 🡪 less retrieval noise

10-20% lower than GOME (accounting for interannual climate differences)

June-August

GOME

OMI

Relating HCHO Columns to Isoprene Emissions

Uniform

Ω_HCHO-E_isoprene relationship

Variable

Ω_HCHO-E_isoprene relationship

Ω_HCHO = SE_isoprene+ B

[10¹⁵ molecules cm^-2]

[10¹³ atomsC cm^-2 s^-1]

OMI Isoprene Emission

Model of Emissions of Gases and Aerosols from Nature

G2006

Guenther [2006] (MODIS)

CLM

Community Land Model

(AVHRR)

[10¹³ atomsC cm^-2 s^-1]

Environmental drivers (T, hν, LAI, leaf age, …)

Vegetation-specific baseline emission factors

Drive MEGAN with 2 land cover databases

MEGAN Isoprene Emissions

w/ G2006 vegetation

MEGAN Isoprene Emissions

w/ CLM vegetation

Guenther et al., ACP (2006)

12.2 TgC

Broadleaf trees

Broadleaf trees

Spatial Patterns in Isoprene Emissions

OMI – MEGAN Isoprene Emissions

June-August, 2006

Large sensitivity to surface database used

MEGAN higher than OMI over dominant emission regions

MEGAN

MEGAN

OMI – MEGAN Isoprene Emissions

June-August, 2006

MEGAN with CLM vegetation

MEGAN with G2006 vegetation

OMI

OMI

Bottom-Up Emissions Too High in Dominant Source Regions

G2006 Broadleaf Trees

MEGAN w/ CLM Land Cover

MEGAN w/ G2006 Land Cover

OMI – MEGAN Isoprene Emissions

June-August, 2006

MEGAN emissions >70% too high over much of the Ozark Plateau, Upper South, Upper Midwest

Large regional emissions driven by oak tree cover, high temperatures

Broadleaf tree isoprene emissions overestimated

CLM-Driven Emissions Too Low in Deep South

OMI – MEGAN Isoprene Emissions

June-August, 2006

Bias in modeled emissions: >100%

Underestimate of broadleaf tree or shrub coverage

-or-

Modeled emissions from evergreen trees or crops too low

CLM

Fineleaf Evergreens

CLM

Crops

[10¹³ atomsC cm^-2 s^-1]

Constraints on Emission Factors

Possible explanation for OMI-MEGAN discrepancy:

Fast chemical loss within forest canopies?

Farmer and Cohen [2007]

Regress OMI isoprene emissions against G2006 PFTs

​

Variable EFs

Broadleaf tree emissions need to be uniformly decreased by 44%

​

Constant EFs

More consistent with OMI

Optimum broadleaf tree EF:

• 13.1 x 10¹² atomsC/cm²/s

• similar to MEGAN mean

• rejects MEGAN’s use of 3-4x higher EFs in certain locations

The End

OMI vs. Aircraft Data & GEOS-Chem

MEGAN isoprene emissions using 2 land cover databases

Guenther [2006] (MODIS) – G2006

Community Land Model (AVHRR) - CLM

G2006

CLM

EF

[%]

[10¹⁰ atomsC/cm²/s]

[10¹³ atomsC cm^-2 s^-1]

PFT Coverage and Isoprene Emission Factors (EFs)

Isoprene Emission

OMI Isoprene Emission vs. MEGAN-CLM

Uniform

Ω_HCHO-E_isoprene relationship

Variable

Ω_HCHO-E_isoprene relationship

OMI - MEGAN

OMI - MEGAN

MEGAN w/ CLM vegetation

[10¹³ atomsC cm^-2 s^-1]

OMI isoprene emission (uniform slope)

OMI isoprene emission (variable slope)

OMI Isoprene Emission vs. MEGAN-G2006

Uniform

Ω_HCHO-E_isoprene relationship

Variable

Ω_HCHO-E_isoprene relationship

OMI - MEGAN

OMI - MEGAN

MEGAN w/ G2006 vegetation

[10¹³ atomsC cm^-2 s^-1]

OMI isoprene emission (uniform slope)

OMI isoprene emission (variable slope)