Prof. Patrick Irwin

Abstract:

The upper jovian atmosphere is particularly transparent at wavelengths near 5 μ m. Levels well below the cloud layers, which are situated between 0.5 and 2 bar, can be sounded. Large spatial variations of the brightness are observed, which are directly related to the opacity of the overlying cloud layer. Yet, the nature of the 5- μ m absorber in the jovian atmosphere has been subject of much debate. The cloud layer has been modelled many times as a thin, non-scattering layer, the opacity adjusted to fit the overall radiance level. This has proven to work well for individual spectra. Data from the Galileo near infrared mapping spectrometer (NIMS), covering the 0.7- 5.2 μ m range, include a number of observations of the same areas, separated by several hours, at different emission angles. Should the 5 μ m absorber be a thin absorbing layer then, apart from a change in radiance level, the overall shape of the 5- μ m spectrum is also expected to change significantly with emission angle. However, comparison of the 5- μ m spectra measured by NIMS of the same location but at different viewing angles reveals that while the overall radiance level decreases with increasing emission angle, the shape of the spectra remain unchanged. In this paper we present atmospheric models that include scattering to explain this effect. We show that the 5- μ m absorbing cloud particles must be significantly scattering ( ω = 0.9 ± 0.05 ) in order to explain these observations, and find that the base of the cloud layer must reside at pressures less than 2 bar. Furthermore, we show that the scattering within this cloud has important consequences on the retrieval of gas abundances from spectra in the 5- μ m region. © 2006 Elsevier Ltd. All rights reserved.

Abstract:

Near-IR absorption of methane in the 2000-9500 cm-1 spectral region plays a major role in outer planet atmospheres. However, the theoretical basis for modeling the observations of reflectivity and emission in these regions has had serious uncertainties at temperatures needed for interpreting observations of the colder outer planets. A lack of line parameter information, including ground-state energies and the absence of weak lines, limit the applicability of line-by-line calculations at low temperatures and for long path lengths, requiring the use of band models. However, prior band models have parameterized the temperature dependence in a way that cannot be accurately extrapolated to low temperatures. Here we use simulations to show how a new parameterization of temperature dependence can greatly improve band model accuracy and allow extension of band models to the much lower temperatures that are needed to interpret observations of Uranus, Neptune, Titan, and Saturn. Use of this new parameterization by Irwin et al. [Irwin, P.G.J., Sromovsky, L.A., Strong, E.K., Sihra, K., Bowles, N., Calcutt, S.B., 2005b. Icarus. In press] has verified improved fits to laboratory observations of Strong et al. [Strong, K., Taylor, F.W., Calcutt, S.B., Remedios, J.J., Ballard, J., 1993. J. Quant. Spectrosc. Radiat. Trans. 50, 363-429] and Sihra [1998. Ph.D. Thesis, Univ. of Oxford], which cover the temperature range from 100 to 340 K. Here we compare model predictions to 77 K laboratory observations and to Uranus spectra, which show much improved agreement between observed and modeled spectral features, allowing tighter constraints on pressure levels of Uranus cloud particles, implying that most scattering contributions arise from pressures near 2 bars and 6 bars rather than expected pressures near 1.25 and 3.1 bars. Between visible and near-IR wavelengths, both cloud layers exhibit strong decreases in reflectivity that are indicative of low opacity and submicron particle sizes. © 2006 Elsevier Inc. All rights reserved.