Line mixing effects in solar occultation spectra of thelower stratosphere: measurements and comparisonswith calculations for the 1932-cm-1CO2branchCurtis P. Rinsland and L. Larrabee StrowLine mixing effects have been observed in a CO2Q branch recorded in 0.01-cm-1 resolution IR solaroccultation spectra of the lower stratosphere. The spectral data were obtained by the Atmospheric TraceMolecule Spectroscopy Fourier transform spectrometer during the Spacelab 3 shuttle mission in the spring of1985. Analysis of the 1932.47-cm- (11102) (00001) band Q branch ofl2Cl602shows absorption coefficientsbelow the band origin -0.62 times those calculated using a standard Voigt line shape function. Calculationsof line mixing using the Lorentz halfwidths of the lines and a simple energy gap scaling law to parametrizerotational energy transfer reproduce the observed absorption coefficients to -10%. The present resultsprovide the first quantitative information on air-broadened line mixing effects in a Q branch at lowtemperatures (210 K) and show that these effects are significant even at the low pressures of the lowerstratosphere (100 mbar).I. IntroductionA wide variety of remote sounding experiments uti-lize measurements of infrared vibration-rotationbands to deduce atmospheric parameters, such aspressure, temperature, and atmospheric composition.In the inversion of the data, the assumption is usuallymade that the absorption or emission can be calculatedby summing the contributions of isolated Voigt-shaped lines. Although this assumption is satisfac-tory for many studies, laboratory investigations haveshown that significant deviations from the Voigt shapeoccur in the infrared (see, for example, Refs. 1-3).One of the mechanisms that can produce deviations isinterference effects caused by the overlapping of lines.These interference effects, often termed line mixing orcoupling, are predicted to be important in the far wingsof lines4 5and in Q branches where the line spacing issmall.6 7Q branches are often used for remote sounding be-cause of the high density of absorption within thesefeatures. For example, the Cryogenic Limb ArrayEtalon Spectrometer (CLAES) instrument will flyaboard the Upper Atmosphere Research SatelliteCurtis Rinsland is with NASA Langley Research Center, Atmo-spheric Sciences Division, Hampton, Virginia 23665-5225, and L. L.Strow is with University of Maryland, Physics Department, Balti-more, Maryland 21228.Received 21 June 1988.(UARS)8and use the C02 Q branch at 792 cm-1forsounding of both temperature and pressure in thestratosphere. Q branches have frequently been usedfor detection and quantitative measurement of tracespecies in the atmosphere,9-14but line mixing effectshave not been included in these studies. Neglectingline mixing effects in atmospheric radiative transfercalculations can lead in some cases to significant re-trieval errors, as has been demonstrated in a theoreti-cal study of atmospheric emission in the 15-,um 2 Qbranch of CO2.15Line mixing effects in Q branches have been studiedin the laboratory using both Raman16l8and directabsorption techniques.19-22One analysis approachhas been to compare the measured spectra with syn-thetic spectra calculated with mixing effects modeledusing a simple energy gap scaling law and the knownpressure broadening coefficients, line positions, andintensities. Absorption coefficients computed in thisway were found to agree within 10% with values mea-sured from high-resolution room-temperature absorp-tion spectra of a self-broadened CO2Q branch, a N2-broadened C02 Q branch,19and a self-broadened N20Q branch,20indicating that line mixing effects in Qbranches can be accounted for fairly accurately nearroom temperature with this model, which has no ad-justable parameters. However, for analysis of atmo-spheric observations, it is necessary to include thevariation of line mixing with atmospheric temperaturein the calculations. Unfortunately, to date no labora-tory measurements have been reported at reducedtemperatures to test such calculations.1 February 1989 / Vol. 28, No. 3/ APPLIED OPTICS 457The purpose of the present paper is to report the useof high-resolution solar occultation spectra of the low-er stratosphere to analyze line mixing effects in the Qbranch of the 1932.47-cm-" (11102) -(00001) band of12C1602. These measurements provide the first quan-titative data on air-broadened line mixing effects atlow temperatures (210 K). They also allow an op-portunity to test methods proposed for including tem-perature effects in line mixing calculations.15II. ObservationsThe spectral data used in our analysis were acquiredfrom orbit by the Atmospheric Trace Molecule Spec-troscopy (ATMOS) instrument, a high-speed Fouriertransform spectrometer designed to obtain measure-ments of the composition of the earth's upper atmo-sphere through high-resolution solar occultation mea-surements covering 2-16 Atm. The spectra wererecorded on 30 Apr.-1 May 1985 during the Spacelab 3mission. Details concerning the ATMOS instrumentand its operation during Spacelab 3 are described byFarmer and Raper23and Farmer et al.2 4Processing ofthe ATMOS data and the analysis software will bedescribed in a future publication.2 5The essential elements of the ATMOS instrumentare a double-passed Michelson interferometer, tele-scope, and hemispherical Sun tracker. During eachoccultation the interferometer records successive dou-ble-sided interferograms every 2.2 s, which from theSpacelab 3 orbital altitude of -360 km corresponds tosuccessive tangent height separations of -4.2 km inthe upper atmosphere. In the lower stratosphere andtroposphere, the spacing between tangent heights ofsuccessive scans is reduced by refraction and drifts ofthe Sun tracker field of view on the flattened solar disk.The maximum path difference of each scan is 47.5 cm,yielding an unapodized resolution of 0.01 cm-'. Theradiation is detected by a HgCdTe detector cooled to77 K. Spectra were obtained during nineteen occulta-tions (twelve sunsets between 25.6 and 32.70N andseven sunrises between 46.7 and 49.0S) during theSpacelab 3 mission.ATMOS spectra are obtained from the inverse Fou-rier transform of two-sided interferograms with thesame zero path difference point. For the present anal-ysis, we used ratios of the atmospheric spectra to high-Sun spectra