Simon Calcutt

Abstract:

The Galileo Near Infrared Mapping Spectrometer (NIMS) has returned many spectra of the Jovian atmosphere in the range 0.7-5.2 mum. Although communications restrictions have limited the data return, several wide-area maps have been recorded at near full NIMS resolution. Using these data it is possible to determine both the average shape of the near-infrared (NIR) spectra with very thick clouds (and zero 5-mum brightness) and how these spectra vary as the 5-mum brightness increases.In most of the cases studied, we find that the variable part of the reflectivity has a very different shape to the mean part and may best be explained by variable reflectivity in the cloud layers at pressures greater than 1 bar. In these cases it would thus appear that a variable opacity in a cloud deck based between 1 and 2 bars is mainly responsible for the NIR albedo variations, and not a higher ammonia cloud based above 1 bar as has often been previously suggested. While the composition of this main variable cloud deck could well be ammonium hydrosulphide, other candidates include ammonia (should the much higher estimate of its deep gaseous fractional abundance resulting from the Galileo probe mission be correct), and perhaps even the upper reaches of a deeper water cloud. (C) 2002 COSPAR. Published by Elsevier Science Ltd. All rights reserved.

Abstract:

NIMS is the Near-Infrared Mapping Spectrometer on board the Galileo spacecraft in jovian orbit. We have selected four maps of warm-to-hot regions of the North Equatorial Belt (NEB) for study, analyzing the spectra emerging in the low-opacity 5-μm window. Two methods for calculating the spectrum have been used. The first is a full-scattering radiative transfer forward model that is slow but accurate. The second method calculates spectra by interpolating on a grid of spectra precalculated using the first method for a range of model atmospheres. This method of forward calculation is more suited to analysis of large data sets where application of the full radiative transfer in every instance would be computationally prohibitive. The faster method is verified against the first before being used alone. A retrieval (inversion) algorithm is then used to match model spectra to data and obtain values for cloud opacities and gas mixing ratios. We first sum spectra with similar peak radiances to produce mean spectra representative of brighter and darker (at 5 μm) regions of the maps. These coadded spectra are then analyzed with the fast retrieval code to obtain the average variations in atmospheric parameters from the center to the edges of the hotspots. These analyses confirm that 5-μm hotspots are relatively cloud free, and that a medium level (1.5-bar) cloud layer of large NH4SH particles is the main absorber at these wavelengths. Variations in water vapor relative humidity and high (0.5-bar) ammonia cloud opacity are also derived. We then analyze single spectra over wide areas to produce spatial maps of parameter variations. We find that models that do not include a deep water cloud (~4 bar) do not match all the spectra to within the noise level. A deep water cloud therefore seems to be present in localized areas, toward the edges of the hotspot regions. We examine these findings in the light of results from other Galileo instruments, concluding that the deep cloud observed by the SSI instrument at several locations is likely to be the deep water cloud required by the NIMS data. © 2001 Academic Press.