Methane from the Tropospheric Emission Spectrometer

1

• Methane from the Tropospheric Emission
  Spectrometer
• Vivienne Payne, Mark Shephard
• Atmospheric and Environmental Research, Inc.
• Tony Clough
• Clough Associates
• Brendan Fisher, Greg Osterman, Bill Irion
• JPL
• Ray Nassar
• University of Toronto
• Jennifer Logan
• Harvard University
• AURA Science Meeting
• 27th-30th October 2008

2
Methane CH4
Long-term trend

• Important greenhouse gas
• Important for atmospheric chemistry
• Influence on OH concentration
• Factors influencing growth rate
• Wetland emission
• Biomass burning
• Significant increase in growth rate in 2007
• Observed from NOAA flask network
• Long lived
• fairly well mixed in troposphere
• variations are small (lt10 globally)
• latitudinal gradient sources in NH

3
TES CH4 Averaging kernels and DOFS
Tropical, over ocean
Low latitude, over desert

• Antarctica (polar winter)

Siberia (polar summer)
TES is most sensitive to CH4 in mid to upper
troposphere.
4
Representation of retrievals with limited
vertical information

• TES Level 2 products are supplied on a fine grid
  (67 pressure levels)
• Number of fine grid levels is much greater than
  number of DOFs
• Data on single level of the reported fine grid
  may contain very little measurement information
• Comparisons with other data or global models
  without the use of sophisticated data
  assimilation systems
• e.g. visual inspection of maps, analysis of
  latitudinal gradients
• Move to a representation that better reflects the
  number of DOFS
• Report retrieval state in terms of one element
  per DOFS (Von Clarmann and Grabowski, ACP, 2007)
• For TES CH4, report one quantity a
  representative tropospheric VMR
• Representative of the mid-upper troposphere
• There are a number of different approaches
• SVD (Ceccherini et al., 2003) partial
  eigen-decomposition (Joiner et al., 1998),
  analysis of leading eigenvectors (Pan et al.,
  1998)
• Utilize a straightforward post-retrieval mapping
  process that
• Reduces the influence of the a priori
• Is performed in physical space
• Retains the sensitivity (level and vertical
  resolution) of each profile
• Varies spatially and temporally

5
Representative Tropospheric VMR resultsPayne et
al. (2008), submitted to JGR
RTVMR for July 2006
July zonal means
Latitudinal Gradients
Difference from GEOS-Chem, after bias removal at
50S-60S
Effective pressure
Latitude
GEOS-Chem CH4 Wang et al. (2004), Xiao et al.
(2004)
6
Global spatial features
TES (2006) minus GEOS-Chem (2001), after 3.5
bias removal
7
TES CH4 representative tropospheric VMR
TES RTVMR minus GEOS-Chem 2001 RTVMR field (after
removal of 3.5 high bias from TES)
High CH4 over Indonesia in October 2006
associated with increased biomass burning during
the El Nino?
8
Northern Hemisphere CH4
High CH4 values Possible sources? Transport
across Pacific?
TES CH4 RTVMR (July 2006)
Effective pressure
Difference from GEOS-Chem
9
DACOM flights for INTEX-B Comparisons between
DACOM, TES v003, AIRS v5.0

• DACOM Differential Absorption CO Measurement
• PI Glen Sachse, NASA Langley
• INTEX-B DACOM flights March/April/May 2006
• TES/AIRS
• Both mid-IR instruments, sensitive to CH4 in mid-
  to upper troposphere
• AIRS CH4 has already been compared with aircraft
  measurements from a range of latitudes
• Xiong et al, JGR-B (2008)
• Different spectral resolution
• TES 0.06 cm-1 unapodized
• AIRS 1.0 cm-1 in CH4 region
• Different forward models
• AIRS v5.0 CH4 absorption coefficients tuned using
  tropical aircraft profiles
• Different retrieval techniques
• Retrieval approach
• Retrieval grid

10
TES/AIRS/DACOM comparisons example case

• Comparable sensitivity
• TES sensitive lower down
• TES has slightly more DOFS
• Helpful at high latitude
• TES biased high wrt DACOM
• N2O interference
• Plans to address in future versions
• AIRS biased low wrt DACOM
• Plans to re-examine tuning in future versions

11
TES/AIRS/DACOM scatter plots

• TES 3.5 bias wrt DACOM
• Investigate physical reasons for bias
• Update N2O profiles in supplemental files
• Try to use microwindows that avoid all N2O lines?

AIRS -2 bias wrt DACOM Tuning to be
re-evaluated for v6.0
12
NDACC FTIR comparisons(Network for the Detection
of Atmospheric Composition Change)
FTIR column data publicly available from
http//www.ndsc.ncep.noaa.gov/

• -77.825 166.65 Arrival Heights, Antarctica
• -45.038 169.684 Lauder, New Zealand
• 28.3 343.52 Izana, Tenerife
• 43.66 280.60 Toronto, Canada
• 53.107 8.854 Bremen, Germany
• 60.200 10.800 Harestua, Norway
• 67.84 20.41 Kiruna Sweden
• 76.52 291.24 Thule, Greenland
• 78.92 11.92 Ny-Aalesund, Spitsbergen
• 80.050 273.58 Eureka NDSC

Future work Comparisons with FTIR tropospheric
VMRs through collaboration with instrument teams
(A. Goldman, U. Denver)
13
Summary/Future Work

• TES CH4 product contains useful info when viewed
  appropriately
• TES CH4 is most sensitive in the mid to upper
  troposphere with 1 DOF
• Representative tropospheric VMR Single
  troposphere VMR value specified at an effective
  pressure
• Sensitivity varies from profile to profile
• TES CH4 RTVMRs show evidence of interesting
  features on a global scale
• TES latitudinal gradients are realistic
• TES closer to GEOS-Chem than to MOZART a priori
• GEOS-Chem surface values have been validated
  against NOAA CMDL measurements (Wang et al, 2004)
• Validation of the TES CH4 product is ongoing..

14
Summary/Future work

• TES shows high bias of wrt GEOS-Chem
• 3.5 bias at 50S-60S in January
• TES shows high bias of 3.5 wrt DACOM in-situ
  aircraft data
• Investigation of the high bias in the TES CH4
• Radiance closure studies using DACOM data
• N2O interference in TES microwindows
• Update N2O to present-day values
• Validation against aircraft and ground-based
  measurements
• FTIR tropospheric volume mixing ratios
• Larger number of aircraft comparisons
• AIRS/SCIAMACHY comparisons on a global scale
• Check whether features of interest are observed
  in both datasets