[Previous] | [Session 30] | [Next]
M. Roos-Serote, J. Pimentao (OAL), S. Vicente (Univ. of Lisbon)
Models of the vertical structure of the jovian troposphere predict the existence of several cloud layers, with cloud base pressures a function of the abundance of condensibles in the atmosphere: a water cloud with a base at pressures higher then 5 bar, na ammonium hydrosulfide cloud somewhere between 1 and 2 bar and na ammonia-ice cloud near 0.7 bar. Analysis by many different authors of many different observations has not yet led to a coherent view on the vertical cloud structure inferred from these data: at visible wavelengths the cloud contrast seems to be due to a cloud located in the sub-bar region (the ammonia cloud), whereas at infrared wavelengths a cloud in the 1-2 bar region in found.
There is an unexplored region in the jovian spectrum that seems ideally suited to infer information on the vertical cloud structure: the region between 4 and 4.5 micron. This region is impossible to observe from Earth due to strong CO2 blinding and spacecraft data is also scarce: the Voyager IRIS spectra were very noisy here and the Galileo/NIMS data set does not always contain this spectral region. The region next to it, the so-called 5 micron window (4.6-5.3 micron) has been extensively studied. Radiation originates from well below the 2 bar level, being attenuated by the overlying clouds. In the 4-4.5 micron region, radiation comes from levels where the clouds are expected (1-2 bars).
We have performed a Principal Component Analysis on the entire 4-5 micron region using several NIMS data cubes. We find that about 75 % of the spatial variation is described by a first entirely positive Principal Component vector. This means that most of the variation seen in the 4-4.5 micron region has a similar origin as the variations seen in the 4.5-5.2 micron region: the opacity variations of the overying clouddeck. Yet, about 25 % is described by a second PC-vector, which shows an anti-correlation between the 4-4.5 and 4.5-5.2 micron region. We find that this can be modeled by introducing two cloud layers, one around 0.7 bar and another around 1-2 bar. We also find that the spatial variations of the opacities of these clouds must vary in an anti-correlated manner in order to reproduce the observed PCs. The physical interpretation must be linked to the way clouds are formed and the dynamics of the atmosphere.
[Previous] | [Session 30] | [Next]
Bulletin of the American Astronomical Society, 37 #3
© 2004. The American Astronomical Soceity.