NASA Goddard Institute for Space Studies Improving GCM ...€¦ · Figure 1. Qualitative agreement for the overall seasonal mean in Summer (JJA) aerosol optical depth at 0.55 mm complied

Improving GCM aerosol climatology using satellite and ground-based measurementsLi Liua,b, Andrew A. Lacisb, Barbara E. Carlsonb, Michael I. Mishchenkob, Brian Cairnsb

aColumbia University & bNASA Goddard Institute for Space Studies, New York, NY 10025

NASA Goddard Institute for Space StudiesNew York, NY 10025

Abstract

Results and Analyses

Figure 1. Qualitative agreement for the overall seasonal mean in Summer (JJA) aerosol optical depth at0.55 µm complied from different datasets. Numbers at top right corner represent the area weighted globalmeans.

Figure 2. GCM underestimate of the overall seasonal mean in Summer (JJA) Ångström exponent compiledfrom different datasets. Numbers at top right corner represent the area weighted global means.

Figure 3. Fractional distribution of principle species (per cent) of GISS annual - mean aerosol climatology.

J F M A M J J A S O N D 0

0.1

0.2

0.3

0.4

τ ove

r lan

d (4

5o S−45

o N)


0.05

0.1

0.15

0.2

GCMTERRAAQUAAERONETAVHRRPOLDER1POLDER2MISR

τ ove

r oce

an (4

5o S−45

o N)

Land Ocean

Figure 4. Seasonal dependence of area weighted overall monthly mean aerosol optical depthfrom different data sources. Data over land (left panel) and over ocean (right panel) have beenaveraged over 45 degree South and North latitudinal band and over available data of the res-pective instrument.

2.42.5

2.6

2.7

10.0 10.2 10.4 10.6 10.8 11.00.00.51.01.52.02.53.0


0.25

0.5

0.75

1

1.25

1.5

GCMTERRAAQUAAERONET

AVHRRPOLDER1POLDER2

Angs

trom

exp

onen

t ove

r lan

d (4

5o S−45

o N)


0.2

0.4

0.6

0.8

1

Angs

trom

exp

onen

t ove

r oce

an (4

5o S−45

o N)

Land

Ocean

Figure 5. Same as Fig. 4, but the averaged data are for the Angstrom Exponent.

Figure 6. Regions selected for comparisons of GCM aerosol climatology with satelliteretrievals.

90oS

60oS

30oS

EQ

30oN

60oN

90oN

DL 120oW 60oW CM 60oE 120oE DL

South Ocean

S Africa

Sahel

S America

PersianSaharan

EC Pacific

N Pacific

E USA

N Atlantic

W Europe

NW Africa

WC Pacific

N Pacific

E Asia

0

0.03

0.06

0.09 NW Africa

τ(0.

55 µ

m)

0

0.2

0.4

0.6Persian

0

0.1

0.2

0.3S Africa

0

0.3

0.6

0.9Sahel

0

0.05

0.1

0.15

0.2East Asia

τ(0.

55 µ

m)

0

0.05

0.1

0.15E USA

0

0.05

0.1

0.15

0.2W Europe

0

0.05

0.1

0.15

0.2N Atlantic

0

0.05

0.1

0.15

0.2N Pacific

τ(0.

55 µ

m)

0

0.01

0.02

0.03

0.04

0.05C Pacific

0

0.05

0.1

0.15South Ocean

0

0.05

0.1

0.15South Ocean

0

0.1

0.2

0.3Sahara

τ(0.

55µm

)

0

0.1

0.2

0.3

0.4Persian

0

0.05

0.1

0.15

0.2S Africa

0

0.2

0.4

0.6

0.8Sahel

0

0.1

0.2

0.3S America

τ(0.

55µm

)


0.05

0.1

0.15

0.2E USA


0.05

0.1

0.15

0.2W Europe


0.1

0.2

0.3East Asia


0.03

0.06

0.09South Ocean

τ(0.

55µm

) J F M A M J J A S O N D

0

0.1

0.2

0.3South Ocean

BCSea SaltNitrateSulfateOCDust

0

0.25

0.5

0.75

1NW Africa

A

0

0.3

0.6

0.9

1.2Persian

0

0.5

1

1.5S Africa

0

0.3

0.6

0.9

1.2Sahel

0

0.5

1

1.5East Asia

A

0

0.5

1

1.5

2E USA

0

0.5

1

1.5W Europe

0

0.5

1

1.5N Atlantic

0

0.5

1

1.5N Pacific

A

0

0.25

0.5

0.75

1C Pacific

0

0.25

0.5

0.75

1South Ocean

0

0.25

0.5

0.75

1South Ocean

0

1

2

3Sahara

A

0

1

2

3Persian

0

1

2

3S Africa

0

1

2

3Sahel

0

1

2

3S America

A


1

2

3E USA


1

2

3W Europe


1

2

3East Asia


1

2

3South Ocean

A


1

2

3South Ocean

GCMTERRAAQUAPOLDERAVHRR

NOTE: Top three rows present the Angstrom Exponentsaveraged over the water surface while the bottom threeones are the A averages over the land area for eachregion shown in Fig.6.

Figure 7. Regional analysis of overall monthly mean aerosol optical depth averaged over the various aerosol regimes shown in Fig.6. Averages are computed over water surfaces (top three rows) and land areas (bottom three rows) if the designated area contains both land and water masses.

Figure 8. Relative contributions of each principle aerosol component considered in theGCM to the total aerosol optical depth depicted by the thick black curves in Fig.7.

Figure 9. As in Fig. 7, but for seasonal dependence of overall monthly mean Angstrom Exponent at different places shown in Fig.6.

a: Aerosol optical depth and Angstrom Exponent

b: Aerosol single scattering albedo

Figure 11. January and July monthly-mean single scattering albedo for the GISS ModelE aerosol climato-logy for 1990 (left panels). TOMS Aerosol Index (AI) (right panels) has been re-scaled as (1 - 0.1 x AI)to roughly resemble the GCM single scattering albedo. Aerosol single scattering albedo measured locally at AERONET network sites is shown in the center panels. Numbers appearing in the upper right cornersare area weighted global mean values.

Figure 12. Differences in the GISS GCM andAERONET annual mean single scattering al-bedo. The single scattering albedo is reportedat 0.55 µm for the GCM, while the selectedAERONET wavelength is 0.44 µm. The num-ber in the upper-right corner represents areaweighted global mean.


0.05

0.1

0.15

0.2

0.25

Total AOD at 550 nm

GCM

TerraAERONET

Aqua

GISS GCM Aerosol at Midway Island


0.02

0.04

0.06

0.08

0.1

Sulfate

Sea Salt

NitrateBlack Carbon

Organic Carbon

Dust

Absolute value


0.3

0.6

0.9

1.2

Terra

GCM

Aqua

AERONET


10

20

30

40

50

Sulfate

Sea Salt

NitrateBlack Carbon

Organic Carbon

Dust

Percentage


0.1

0.2

0.3

0.4

Total AOD at 550 nm

GCM

Terra

AERONET

Aqua

GISS GCM Aerosol at SGP


0.03

0.06

0.09

0.12

0.15

Sulfate

Sea Salt

NitrateBlack Carbon

Organic Carbon

Dust

Absolute value


0.5

1

1.5

2

2.5Terra

GCM

Aqua

AERONET


10

20

30

40

50

60

70

Sulfate

Sea Salt

NitrateBlack Carbon

Organic Carbon

Dust

Percentage

Figure 10. GISS GCM along with MODIS Terra, MODIS Aqua, and AERONET at selected sites.

Schmidt. G. A., et al. (2005), Present day atmospheric simulations using GISS ModelE: Comparison to in-situ, satellite and reanalysis data, J. Climate, in press.

0

0.1

0.2

0.3

0.4NW Africa

τ(0.

55 µ

m)

0

0.2

0.4

0.6

0.8Persian

0

0.2

0.4

0.6S Africa

0

0.3

0.6

0.9

1.2Sahel

0

0.2

0.4

0.6East Asia

τ(0.

55 µ

m)

0

0.1

0.2

0.3

0.4E USA

0

0.1

0.2

0.3

0.4W Europe

0

0.1

0.2

0.3

0.4N Atlantic

0

0.1

0.2

0.3

0.4N Pacific

τ(0.

55 µ

m)

0

0.05

0.1

0.15

0.2C Pacific

0

0.1

0.2

0.3South Ocean

0

0.1

0.2

0.3South Ocean

0

0.25

0.5

0.75

1Sahara

τ(0.

55µm

)

0

0.2

0.4

0.6

0.8Persian

0

0.2

0.4

0.6S Africa

0

0.3

0.6

0.9

1.2Sahel

0

0.2

0.4

0.6

0.8S America

τ(0.

55µm

)


0.2

0.4

0.6E USA


0.1

0.2

0.3

0.4W Europe


0.2

0.4

0.6

0.8East Asia


0.1

0.2

0.3South Ocean

τ(0.

55µm

)


0.1

0.2

0.3South Ocean

GCMTERRAAQUAMISRPOLDERAVHRR

NOTE: Top three rows present the aerosol optical depthsaveraged over the water surface while the bottom threeones are the τ averages over the land area for eachregion shown in Fig.6.

A physically based aerosol climatology is essential to addre ss the questions of global climate changes. In this study, we use a vailable satellite and ground-based measurements, i.e., MODIS, M ISR, POLDER, AVHRR, and AERONET data, to characterize and validate the geographic distribution and seasonal variability of the GISS ModelE [Schmidt et al., 2005] aerosol optical depth (AOD) and particle size via Ångström exponent (A). Our analysis of satellite and ground-based observations shows that there is considerable “diversity” in observed global distributions of AOD, and in particular, the Ångström exponent. Given the uncertainties associated with satellite retrieval results, both the global optical depth and the Ångström exponent distributions of GCM aerosols are qualitatively reasonable. The Ångström exponent of the GISS GCM aerosol is clearly biased low compared to satellite data, implying that the GCM aerosol climatology sizes might be overestimated. We have also compared the GISS ModelE aerosol single scattering albedo climatology versus TOMS Aerosol Index (AI) and AERONET data. This inter-comparison study points to the need to readjust the size specification of different aerosol species in the GCM to produce agreement between the model derived aerosol climatology and those retrieved from satellite and ground-based measurements, and requires improvement of the chemical transport model simulations upon which the GCM aerosol climatology is based. On the other hand, the existing diversity among different satellite products indicates an urgent need for improved retrieval of tropospheric aerosol radiative properties from satellite measurements.

MISR Angstrom Exponent data are not currently available

Documents

NASA Goddard Institute for Space Studies Improving GCM ...€¦ · Figure 1. Qualitative agreement for the overall seasonal mean in Summer (JJA) aerosol optical depth at 0.55 mm complied