Separating recent emissions from background CO and V6 validation

Juying Warner, Zigang Wei, Fabien Carminati - Dept. of Atmospheric and Oceanic Science, Univ. of Maryland College Park

Recent Publications: •  Warner, J., Carminati, F., Wei, Z., Lahoz, W., and Attié, J.-L.: Tropospheric carbon monoxide variability

from AIRS under clear and cloudy conditions, Atmos. Chem. Phys., 13, 12469-12479, doi:10.5194/acp-13-12469-2013, 2013.

•  Warner, J. X., Yang, R., Wei, Z., Carminati, F., Tangborn, A., Sun, Z., Lahoz, W., Attié, J.-L., El Amraoui, L., and Duncan, B.: Global carbon monoxide products from combined AIRS, TES and MLS measurements on A-train satellites, Atmos. Chem. Phys., 14, 103-114, 2014.

•  Warner, J. X., Wei, Z., Carminati, F.: AIRS Version 6 CO Updates and Validation, Atmos. Measurement Technology, 2014 – in preparation.

Funded by: •  NASA NNH05ZDA001N ACMAP: An Inter-comparison of Tropospheric Carbon Monoxide Measurements from the Past and Current Satellite

Instruments: Using a Uniform Retrieval Algorithm. (2007-2011)

•  NASA NNH10ZDA001N, Earth System Data Records Uncertainty Analysis, Uncertainty Analysis of Tropospheric Carbon Monoxide Data Records Using AIRS and IASI from a Uniform Algorithm. (2011-2014)

•  NASA Jet Propulsion Laboratory – COEUS #19770, $100K, 10/01/2013-9/30/2014, Continued Efforts for AIRS CO Using OE Method.

Collocated AIRS Pixels with MODIS Cloud Masks

163.1 163.2 163.3 163.4 163.5 163.6 163.7

23.4

23.5

23.6

23.7

23.8

23.9

24.0

Longitude (° )

Latit

ude

(°)

•  For retrieval quality assurance. •  To develop a clear sky dataset for AIRS single FOV ret.

AIRS V5 CO trends under clear and cloud cleared skies - daytime

Southern Hemisphere Ocean

2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013Time (Year)

50

100

150

-75

-50

-25

0

25

clou

d-cl

eare

d - c

lear

airs cloud SHOcean : -0.23(±0.27)ppbv/yr

airs clear SHOcean : -0.30(±0.27)ppbv/yr

Southern Hemisphere Land

2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013Time (Year)

50

100

150

CO (p

pbv)

-75

-50

-25

0

25

airs cloud SHLand : -0.29(±0.29)ppbv/yr

airs clear SHLand : -0.07(±0.27)ppbv/yr

Northern Hemisphere Ocean

2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013Time (Year)

50

100

150AIRS cloud-clearedAIRS clearDifference

-25

0

25

50

75

clou

d-cl

eare

d - c

lear

airs cloud NHOcean : -1.07(±0.32)ppbv/yr

airs clear NHOcean : -1.01(±0.32)ppbv/yr

Northern Hemisphere Land

2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013Time (Year)

50

100

150CO

(ppb

v)

-25

0

25

50

75

airs cloud NHLand : -1.32(±0.33)ppbv/yr

airs clear NHLand : -1.28(±0.33)ppbv/yr

•  Globally cloud clearing cause very little differences in trends; •  It does causes positive bias over land and negative bias over ocean; • CO shows decreasing trends, more in the NH than in the SH indication of economic slowdown.

AIRS V5 CO trends under clear and cloud cleared skies - nighttime

Southern Hemisphere Ocean

2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013Time (Year)

50

100

150

-75

-50

-25

0

25

clou

d-cl

eare

d - c

lear

airs cloud SHOcean : -0.22(±0.27)ppbv/yr

airs clear SHOcean : -0.27(±0.28)ppbv/yr

Southern Hemisphere Land

2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013Time (Year)

50

100

150

CO (p

pbv)

-75

-50

-25

0

25

airs cloud SHLand : -0.17(±0.30)ppbv/yr

airs clear SHLand : -0.10(±0.29)ppbv/yr

Northern Hemisphere Ocean

2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013Time (Year)

50

100

150AIRS cloud-clearedAIRS clearDifference

-25

0

25

50

75

clou

d-cl

eare

d - c

lear

airs cloud NHOcean : -0.98(±0.32)ppbv/yr

airs clear NHOcean : -0.77(±0.32)ppbv/yr

Northern Hemisphere Land

2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013Time (Year)

50

100

150CO

(ppb

v)

-25

0

25

50

75

airs cloud NHLand : -0.78(±0.32)ppbv/yr

airs clear NHLand : -0.62(±0.32)ppbv/yr

• Similar conclusions to the daytime cases, but smaller CO difference between clear and cloud cleared skies.

DOFS comparison: clear vs cloud-cleared Clear Cloud-Cleared

•   CO  VMRs  and  DOFS  show  higher  values  under  clear  sky  cases.  

CO VMRs comparison: clear vs cloud-cleared Clear Cloud-Cleared

Separating CO fresh emissions from background - PDF distribution and two Gaussian Fits  

NH Day Land

0 50 100 150 200 2500.0

0.2

0.4

0.6

0.8

1.0

1.2Clearclear BGclear FECloudycloudy BGcloudy FE

NH Day Ocean

0 50 100 150 200 2500.0

0.2

0.4

0.6

0.8

1.0

1.2Clearclear BGclear FECloudycloudy BGcloudy FE

SH Day Land

0 50 100 150 200 2500.0

0.2

0.4

0.6

0.8

1.0

1.2Clearclear BGclear FECloudycloudy BGcloudy FE

SH Day Ocean

0 50 100 150 200 2500.0

0.2

0.4

0.6

0.8

1.0

1.2Clearclear BGclear FECloudycloudy BGcloudy FE

CO VMRs (ppbv)

Nor

mal

ized

PD

Fs

AIRS V5 and NOAA IASI at 500 hPa CO trends for recent emission (right Gaussian: red-AIRS and yellow-IASI) and background (left Gaussian: blue-AIRS and cyan-IASI)

Southern Hemisphere Ocean

2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013Time (Year)

30

80

130

airsBG SHOcean : -0.54(±0.26)ppbv/yr

airsFE SHOcean : -0.78(±0.29)ppbv/yr

iasiBG SHOcean : -0.57(±1.06)ppbv/yr

iasiFE SHOcean : -0.68(±0.91)ppbv/yr

Southern Hemisphere Land

2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013Time (Year)

30

105

180

CO (p

pbv)

airsBG SHLand : -0.17(±0.26)ppbv/yr

airsFE SHLand : 0.01(±0.29)ppbv/yr

iasiBG SHLand : 5.81(±1.17)ppbv/yr

iasiFE SHLand : 2.72(±1.00)ppbv/yr

Northern Hemisphere Ocean

2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013Time (Year)

50

105

160

AIRS emissionsAIRS backgroundIASI emissionsIASI background

airsBG NHOcean : -0.94(±0.32)ppbv/yr

airsFE NHOcean : -1.55(±0.34)ppbv/yr

iasiBG NHOcean : 4.11(±1.27)ppbv/yr

iasiFE NHOcean : 3.20(±1.11)ppbv/yr

Northern Hemisphere Land

2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013Time (Year)

70

120

170CO

(ppb

v)

airsBG NHLand : -0.97(±0.32)ppbv/yr

airsFE NHLand : -1.20(±0.34)ppbv/yr

iasiBG NHLand : 4.32(±1.42)ppbv/yr

iasiFE NHLand : 2.30(±1.30)ppbv/yr

Length of record matter to trend studies.

CO Emissions: AIRS 500 hPa vs GFED3\MACCity

80

113

147

180

CO (p

pbv)

NH

4

6

8

10

x10-1

1 kg.

m-2.s

-1

2002 2003 2004 2005 2006 2007 2008 2009 2010 2011Time (Year)

60

100

140

180

CO (p

pbv)

SH

0.0

3.3

6.7

10.0

x10-1

1 kg.

m-2.s

-1

CO Emission in NH

4 5 6 7 8 9GFED3\MACCity CO (x10-11 kg.m-2.s-1)

80

100

120

140

160

AIRS

CO

(ppb

v)

r = 0.726

CO Emission in SH

0 2 4 6 8GFED3\MACCity CO (x10-11 kg.m-2.s-1)

50

75

100

125

150

r = 0.915

Recent emission from AIRS correlated With standard emission inventories

Implications of separating fresh emissions from the background

50 100 150 200CO VMR (ppbv)

0.00.20.40.60.81.0

Num

ber o

f Cas

es BKGEMSNEW

-120 -100 -80 -60

-120 -100 -80 -60 -40

3035

4045

50

3035

4045

50

North_America: Background for 20060304

-120 -100 -80 -60

-120 -100 -80 -60 -40

3035

4045

50

3035

4045

50

Raw Data ( UT: 18:59:34~19:05:27 )

40.0 60.0 80.0 100.0 120.0 140.0 160.0 180.0

AIRS V6 CO Validation increase high values in NH & decrease low values in SH

AIRS V6 CO – primary upgrades: Hemispheric uniform first guesses

& upstream improvements

•  AIRS V5 CO uses 1 global FG leads to low bias in the NH and high bias in the SH. •  V6 made significant improvements.

AIRS CO V5 and V6 validated against HIPPO-HIAPER Pole to Pole Observations

Summer Flights 4 & 5) Spring Flight 3) Winter Flights 1 & 2)

1.  Jan. 8 - 30, 09 2.  Oct. 31 – Nov. 22, 09 3.  Mar. 24 – Apr. 16, 10 4.  Jun. 14 – Jul. 11, 11 5.  Aug. 9 – Sept. 9, 11

Ave Differences INTEX-A/B & ARCTAS

OE for V7 Trace Gases!!