HR: 0830h
AN: A21C-0986    [PDF]
TI: Evaluation and Correction of Spectral Undersampling by the OMI Instrument on Aura
AU: * Chance, K
EM: kchance@cfa.harvard.edu
AF: Harvard-Smithsonian Center for Astrophysics, 60 Garden Street, Cambridge, MA 02138 United States
AU: Sioris, C E
EM: csioris@cfa.harvard.edu
AF: Harvard-Smithsonian Center for Astrophysics, 60 Garden Street, Cambridge, MA 02138 United States
AU: Kurosu, T P
EM: tkurosu@cfa.harvard.edu
AF: Harvard-Smithsonian Center for Astrophysics, 60 Garden Street, Cambridge, MA 02138 United States
AB: Detector array-based spectrometers used in atmospheric remote sensing from space suffer from spectral undersampling: they do not sample high enough in spatial frequency to fully represent the information content of the instrument transfer function, as is required by the sampling theorem (Nyquist sampling). Thus they permit higher-frequency, aliased, information into the measured spectra. This is further complicated by the binning over the detector pixel response profiles that leads to the measured radiances and irradiances. Undersampling strongly impacts the comparison of radiances and irradiances, which is an integral step in the analysis of the spectra. This comparison requires resampling in wavelength due to Doppler shifts between radiance and irradiance measurements, plus smaller contributions from instrument changes in-orbit. Neglecting effects from undersampling can lead to substantial, often dominant, systematic residuals in the spectral fitting. Correction of undersampling for previous instruments (GOME and SCIAMACHY) has been handled in an {\it ad hoc} fashion, by resampling undersampled and correctly-sampled representations of a high-resolution Fraunhofer reference spectrum and differencing the results. We present a more systematic approach to undersampling characterization by frequency analysis of the instrument transfer function, sampling of a Fraunhofer reference spectrum, and convolution with the detector pixel response profile. This analysis is presented for the EOS-Aura Ozone Monitoring Instrument (OMI): We determine the extent of undersampling in the windows for fitting the atmospheric gases O$_{3}$, NO$_{2}$, SO$_{2}$, BrO, OClO, and HCHO, and provide corrections for inclusion in the data analysis algorithms.
DE: 0394 Instruments and techniques
SC: Atmospheric Sciences [A]
MN: 2003 Fall Meeting