We have simulated near-infrared spectra of the emission from Venus' nightside with a radiative transfer program that allows for emission, absorption, and scattering by atmospheric gases and particles. Except for an O₂ airglow emission band near 1.27 µm, the emission is produced in the hot, lower atmosphere of Venus. Its intensity at the top of the atmosphere is substantially decreased by scattering and absorption within the main clouds and is concentrated within spectral "window" regions where the lower atmosphere is most transparent. The twin goals of this paper are to assess the adequacy of current gas databases for simulating Venus' nightside near-IR spectra and, within these limitations, to derive information on gas abundances below the main clouds and the optical thickness of the main clouds. We used the moderate resolution spectra of Crisp et al. (1991) for both these objectives and the very high resolution spectra of Bezard et al. (1990) for the first.

The HITRAN database for the permitted transitions of CO₂ is found to be totally inadequate for simulating the near-infrared emission from Venus' nightside. However, much improved simulations can be made when Wattson's high-temperature ("high-T") database of CO₂ is used instead. A major weakness of the current gas databases is the lack of accurate information on CO₂ and H₂O continuum opacity.

Even bright spots on the nightside have substantial cloud-scattering optical depths. We derive cloud optical depths of about 25 at a reference wavelength of 0.63 µm by fitting spectra of three different spots. We find that the H₂O mixing ratio has a constant value of 30±10 ppm in the altitude range from 10 to 40 km. We also infer mixing ratios for SO₂ of 180±70 ppm at 42 km, HCl of 0.48±0.12 ppm at 23.5 km, HF of 1-5 ppb at 33.5 km, and upper limits on the mixing ratio of CH₄ of 0.1 ppm at 30 km and 2 ppm at 24 km.

We show that that it is possible to infer both the mixing ratios of some gas species and their vertical gradients at the centroid of emission in a given spectral window. In particular, we find that OCS has a mixing ratio, α, of 4.4±1.0 ppm and a gradient, dα/dz, of -1.58±0.30 ppm/km at an altitude of 33 km; i.e., the OCS mixing ratio strongly increases with decreaseing altitude. In addition, we find that the corresponding values for CO are 23±5 ppm and +1.20±0.45 ppm/km at an altitude of 36 km. Therefore, within a factor of 2 CO is decreasing toward the surface at the same rate as that as which OCS is increasing. These results are in good agreement with theories of lithospheric buffering and gas thermodynamic equilibrium at the surface that predict a substantial abundance of OCS (several 10s of ppm) at the surface due to reactions in which CO is one of the reactants.