II. Power Spectrum Analysis and Wave Search

be identified with periodic features such as vortex chains and the equatorial plumes. The origin of others is less certain. We We study power spectra and search for planetary waves in present a table of our best wave candidates.  1996 Academic Press, Inc. images of Jupiter’s cloud opacity. The observation wavelength of 4.9 mm senses thermal emission from the p5-bar level; overlying clouds attenuate the emission. Our companion paper (J. INTRODUCTION Harrington, T. E. Dowling, and R. L. Baron, 1996, Icarus 124, 22–31) describes 19 nights of observations (6 with 3608 longiJupiter’s middle and deep atmospheric regions strongly tude coverage) and new reduction techniques. Atmospheric influence the types of dynamics that occur above them seeing limits resolution to p2500 km. Zonal power spectral density at planetary wavenumbers higher than p25 follows a (Dowling and Ingersoll 1989). The direct study of these power law in the wavenumber. Eastward jet–power laws averregions is inhibited by the presence of the ammonia clouds, age 22.71 6 0.07 and westward jet–power laws, excluding which reflect most visible wavelengths and whose tops are cloud-obscured regions, average 23.14 6 0.12. Wavenumbers located near 250 mbar (West et al. 1986). The deepest 1–24 roughly follow power laws near 20.7 for both jet direcprobing light we can receive from Jupiter is thermally emittions, but with many superposed discrete features. The meridioted near the 5-bar level at wavelengths close to 5 em nal spectrum similarly breaks around wavenumber 25, with (Kunde et al. 1982). This light is attenuated as it passes power law trends of 20.36 and 23.27. However, a pattern of through the various cloud layers, giving us our best source undulations is superposed over its linear trends. of information on the optical thicknesses of the clouds. L. D. Travis (1978, J. Atmos. Sci. 35, 1584–1595) established an empirical correspondence between power spectra of atmoAlthough spectral and photometric work at this wavespheric kinetic energy and those of cloud opacities for the Earth length has been progressing for some time (Beer and Tayand analyzed Venus cloud data under this assumption. We do lor 1973, Terrile 1978), only in the past decade have electhe same for Jupiter. If the Rossby deformation radius, Ld , tronic infrared imagers achieved the sensitivity and spatial were an energy input scale, as baroclinic instability theory preresolution necessary for studies of the horizontal variation dicts, one would expect energy and enstrophy cascades (power of Jupiter’s cloud opacities at this wavelength. laws of 25/3 and 23, respectively) on opposite sides of the The first paper in this series (Harrington et al. 1996, wavenumber corresponding to Ld. If the top of our highhereafter Paper I) describes the acquisition of maps of wavenumber power law is Ld , its value is p2100 km at 458 latitude. Jupiter’s cloud opacities on 19 nights between January and Our spectra show persistent features with phases moving April of 1992. The maps were taken at a wavelength of 4.9 linearly over the 99-day observation period. Some of these can em with the ProtoCAM instrument at the NASA Infrared Telescope Facility. On 6 of these nights we obtained complete longitude coverage. Although there is much work at 1 Current address: Code 693, Goddard Space Flight Center, Greenbelt, MD 20771-0001. optical wavelengths involving the tracking of features and