Progress in OMI Measurements of BrO, OClO and HCHO
Kelly Chance and Thomas P. Kurosu Harvard-Smithsonian Center for Astrophysics Cambridge, MA
EOS Aura Science Team Meeting
Pasadena, March 3, 2005
Acknowledgements
• Our colleagues in the U.S. and Dutch OMI Science Teams: Pieternel F. Levelt (PI), Pawan K. Bhartia, and The Team.
• The PGE/data processing team at SSAI: Peter Leonard, Mike Linda, Marghi Hopkins, Patrick Higgins. et al.
• NASA for providing the funds to carry out this research
The OMI Instrument: Swath Dimensions
• 60 across-track pixels • 2,600km total swath width
• 13x24km2 at nadir • 13x120km2 at edges
• 13x13km2 spatial zoom mode
The Retrieval Algorithm
Based on non-linear least squares, direct fitting
Direct Fitting: I = (I0 e-Absorbers + Ring + Pol1) * Pol2 Proceeds in three stages:
1) Direct Fitting: Modified Solar wavelength calibration 2) Radiance wavelength calibration Use a Radiance Reference R0 instead of the Solar Spectrum I0 3) Fitting of all scan lines/pixels -Absorbers
I = (R0 e
+ Ring + Pol1) * Pol2
Includes Ring effect Improves dark current correction (less cross-track “striping”) in the L2 product Includes undersampling correction
The Results Shown
Only BrO (344-359nm) and HCHO (337-357nm) , since OClO (363-393nm) is near or below the detection limit for the time period that OMI data are available (too late in the South, too early in the North) Exhibit “striping” due to calibration of the L1b products; expected to be solved with latest OPFs Use radiance reference R0 instead of solar reference I0 for improved cross-track radiometric stability
BrO:
Released from shelf-ice during polar Spring
(National Snow and Ice Data Center)
Sea ice concentration Oct 2004
OMI BrO – South Polar Region
OMI BrO – North Polar Region
HCHO: Biomass Burning, Anthropogenic Emission
26 Day Average: 24 September – 19 October 2004
HCHO: Principal intermediate in the oxidation of hydrocarbons
HCHO: OMI Retrieval, MODIS Fire Counts
Chongqing
(Red Basin)
Jakarta
Summary/Outlook
Early into its lifetime, OMI already provides very
encouraging results for BrO and HCHO
Results for OClO are currently inconclusive because of its low concentration during the time of available OMI observations The biggest problem for weak absorbers like BrO, HCHO, and OClO is the calibration of the L1b product; this is expected to be solved with latest OPFs. After improved radiometric calibration is in place, our “real work” starts, i.e., optimization of retrieval windows, quantification of fitting uncertainties, accounting for instrument effects, etc.
Backup
BrO:
Retrieval Errors
Relative error: • 25-30% on average, • ~15% for hot-spots