determined by a combination of factors: rapid increase in opacity through the troposphere, aerosols and cloud tops obscuring the sun, lack of accuracy in assigning tangent heights and pressures to individual spectra, and the difficulty in retrieving stratospheric constituent volume mixing ratios (vmrs) below their peak concentrations. Issues limiting the lower altitude range will be addressed in future analyses. The data processing and inversion methods employed with the ATMOS data follow those described by Norton and Rinsland [1991]. As some changes have been incorporated to enhance routine data processing and to distinguish between this current dataset and previous published analyses, the ATMOS data utilized in the papers in this issue have been ascribed to 'Version 2'. In brief, after transformation of the interferograms to power spectra, an average exoatmospheric spectrum is formed for each occultation against which ratios are made with successive atmospheric spectra. The resulting atmospheric transmission spectra are essentially free of solar and instrumental spectral features. The inversion of these transmission spectra follows a general sequence: assignment of a first-guess tangent height based on spacecraft ephemeris data, refinement of the tangent height (or more correctly, tangent pressure) from spectral fitting to the observed CO2 absorption bands [Abrams et al., 1996], and retrieval of atmospheric temperature-pressure profiles [Stiller et al., 1995], followed by a final tangent pressure assignment. Retrievals of vmr profiles for individual species use spectral intervals that are chosen based on the shape and strength of target gas absorption, the insensitivity of these features to errors in temperature, and their relative freedom from interfering absorption from other constituents [e.g., Gunson 2333

The infrared spectrum of the hydroxyl radical OH, between 1850 and 9000 cm-1 , has been measured with a Fourier transform spectrometer. The source, a hydrogen-ozone diffusion flame, was designed to study the excitation of rotation-vibration levels of the OH Meinel bands under conditions similar to those in the upper atmosphere which produce the nighttime OH airglow emission. Twenty-three bands were observed: nine bands in the Ap = 1 sequence, nine bands in the Ap = 2 sequence, and five bands in the Ap = 3 sequence. A global nonlinear least-squares fit of 1696 lines yielded molecular parameters with a standard deviation of0.003 cm-1. Term values are computed, and transition frequencies in the Ap = 3, 4, 5, 6 sequences in the near-infrared are predicted. Subject headings: laboratory spectra-molecular processes 351

No abstract

Numerous absorption lines of stratospheric sulfur dioxide (SO2) have been identified in solar occultation spectra recorded by the Atmospheric Trace Molecule Spectroscopy (ATMOS) Fourier transform spectrometer during the Atmospheric Laboratory for Applications and Science (ATLAS)‐1 shuttle mission (March 24‐April 2, 1992). Based on their analysis, a volume mixing ratio profile of SO2 increasing from (13 ± 4) p.p.t.v. (parts per 10−12 by volume) at 16 mbar (∼ 28 km) to 455 ± 90 p.p.t.v. at 0.63 mbar (∼ 52 km) has been measured with no significant profile differences between 20°N and 60°S latitude. The increase in the SO2 mixing ratios with altitude indicates the presence of a source of SO2 in the upper stratosphere. Profiles retrieved from ATMOS spectra recorded during shuttle flights in April‐May 1985 and April 1993 show similar vertical distributions but lower concentrations. Two‐dimensional model calculations with SO2 assumed as the end product of H2SO4 photolysis produce SO2 profiles consistent with the ATMOS measurements to within about a factor of 2.

CH4 and N2O are useful as dynamical tracers of stratospheric air transport because of their long photochemical lifetimes over a wide range of altitudes. The cryogenic limb array etalon spectrometer (CLAES) instrument on the NASA UARS provided simultaneous global measurements of the altitude profiles of CH4 and N2O mixing ratios in the stratosphere between October 1, 1991, and May 5, 1993. Data between January 9, 1992, and May 5, 1993 (388 days), have been processed using version 7 data processing software, and this paper is concerned with the assessment of the quality of this data set. CLAES is a limb‐viewing emission instrument, and approximately 1200 profiles were obtained each 24‐hour period for each constituent over a nominal altitude range of 100 to 0.1 mbar (16 to 64 km). Each latitude was sampled 30 times per day between latitudes 34°S and 80°N, or 34°N and 80°S depending on the yaw direction of the UARS, and nearly all local times were sampled in about 36 days. This data set extends the altitude, latitude, and seasonal coverage of previous experiments, particularly in relation to measurements at high winter latitudes. To arrive at estimates of experiment error, we compared CLAES profiles for both gases with a wide variety of correlative data from ground‐based, rocket, aircraft, balloon, and space‐borne sensors, looked at the repeatability of multiple profiles in the same location, and carried out empirical estimates of experiment error based on knowledge of instrument characteristics. These analyses indicate an average single‐profile CH4 systematic error of about 15% between 46 and 0.46 mbar, with CLAES biased high. The CH4 random error over this range is 0.08 to 0.05 parts per million, which translates to about 7% in the midstratosphere. For N2O the indicated systematic error is less than 15% at all altitudes between 68 and 2 mbar, with CLAES tending to be high below 6.8 mbar and low above. The N2O random error is 20 to 5 ppb between 46 and 2 mbar, which also translates to 7% in the low to midstratosphere. Both tracers have useful profile information to as low as 68 mbar, excluding the tropics, and as high as 0.2 mbar (CH4) and 1 mbar (N2O). The global fields show generally good spatial correlation and exhibit the major morphological and seasonal features seen in previous global field data. Several morphological features are pointed out for regions and conditions for which there have been essentially no previous data. These include the differential behavior of the tracer isopleths near and inside the Antarctic winter vortex, and local maxima in the tropics in 1992, probably associated with the Mount Pinatubo sulfate aerosol layer. Overall, the results of this validation exercise indicate that the version 7 CH4 and N2O data sets can be used with good confidence for quantitative and qualitative studies of stratospheric and lower‐mesospheric atmospheric structure and dynamics.