Transcript Gordon_Monday.ppt

A NEW “DIET” FOR AIRBROADENED HALF-WIDTHS OF
WATER VAPOR IN THE
HITRAN2004 COMPILATION
Iouli Gordon
Laurence Rothman
Robert Gamache
David Jacquemart
Why do we need accurate data for
air-broadened half-widths of water?
• Water is the principal absorber of longwave radiation in
terrestrial atmosphere and is important for remote
sensing applications
• Air-broadened half-width is one of the most important
parameters in determination of molecular profiles and
the largest source of uncertainty
• For accurate retrievals (in troposphere) the half-width
and its temperature dependence should be known with
3-10% uncertainty depending on line intensity
Original preference order for adding
air-broadened half-width data into
HITRAN2004
• Intercomparison database (averaging of 2-8
experimental data)
• Database of the single measurements
}
R. Gamache and J.-M.
Hartmann, Can. J. Chem.,
82 (2004) 1013
•
Smoothed values from R. Toth (http://mark4sun.jpl.nasa.gov/data/spec/H2O)
•
Database of complex Robert-Bonamy (CRB) calculations
(http://faculty.uml.edu/Robert_Gamache/)
•
Semi-empirical calculations that employ all available experimental
and theoretical data, D. Jacquemart, R.R. Gamache, and L.S. Rothman,
JQSRT 96, 205-239 (2005).
•
Default estimate
Discrepancies
Corroborated by Masiello et al., 2005
Sources of errors
• No available data is flawless
• Experiments are plagued by impurities and line overlaps,
which are especially difficult for weak lines
• Smoothed values are dependant on the quality of
experimental information
• CRB calculations are not perfectly applied to all types of
transitions
• Semi-empirical calculations inherit uncertainties from the
above sources (except smoothed values)
• Approach: Each transition have its own “metabolism”
and it has own “diet”=source preference
Sources of errors
Sources of errors
Step 1: Eliminate eight references that provide consistently bad data
Intercomparison
Filtering scheme
Step 1
Eliminate measurements that are >20% off semi-empirical calculations
For intercomparison database (Step 2)
Uncertainty (U)- separation (%) between average and the farthest point
• If for J (average J between levels)<8, U>7% then replace with CRB or
semi-empirical calculations
• If for 8<J<13, U>10% then replace with CRB or semi-empirical calculations
• If for J>13, U>15% then replace with CRB or semi-empirical calculations
For single measurements database (Step 3)
Uncertainty (U)- is reported experimental uncertainty (%)
Same procedure
Validation (AER)
N2O
CO2
Validation (ACE)
sr10063 H2O retrieval, 12.5 km
transmittance
0.9
0.8
obs
HITRAN 2004
0.7
0.6
0.8
0.7
0.6
0.5
0.4
0.4
1987.25
1987.3
1987.35
1987.4
-1
wavenumber (cm )
1987.25
1987.3
1987.35
1987.4
obs
HITRAN 2006
1987.25
1987.3
1987.35
1987.4
1987.45
1987.4
1987.45
wavenumber (cm-1)
0.02
0.01
0
-0.01
-0.02
1987.2
1987.2
1987.45
sr10063 H2O retrieval, 12.5 km
0.9
0.5
1987.2
residuals
1
residuals
transmittance
1
1987.45
0.02
0.01
0
-0.01
-0.02
1987.2
1987.25
1987.3
1987.35
Comparison (DLR)
1E-19
1E-20
Intensity
1E-21
1E-22
HITRAN2004-DLR
Update2006-DLR
1E-23
1E-24
1E-25
0
10
20
Absolute Percent Difference
 HITRAN  DLR
 DLR
100%
Conclusions and potential
improvements
• Air-broadened half-widths were improved significantly as can be
seen from application to the spectra of lower troposphere
• The measurements at the higher altitudes (11-12 km) have not
shown significant improvement which affirms the small importance
of air-broadened half-widths in these retrievals
• We have found the new data from DLR to be superior to any
previous experiments in that region and we are planning to include
this data into HITRAN
• More (THOROUGH) experiments are desirable
• Refinement of the CRB calculations, or development of more
accurate calculations is a promising alternative
• Update semi-empirical calculations once significant bulk of new data
becomes available
Acknowledgements
• M. Shephard, J. Delamere, S. A. Clough
(Atmospheric and Environmental Research (AER), Inc.)
• C. Boone and P. Bernath
(University of Waterloo, ACE)
• K. Jucks
(Harvard-Smithsonian Center for Astrophysics)
• NASA