Neptune's far-infrared spectrum from the ISO long-wavelength and short-wavelength spectrometers

Neptune was observed by the Infrared Space Observatory (ISO) Long-Wavelength Spectrometer (LWS) between 46 and 185 mum. At wavelengths between 50 and 110 mum the accuracy of these measurements is less than or equal to 0.3 K. Observations of this planet made by the ISO Short-Wavelength Spectrometer between 28 and 44 mum were combined with the LWS data to determine a disk-averaged temperature profile and derive several physical quantities. The combined spectra are matched best by a He/(H-2 + He) mass ratio of 26.4(-3.5)(+2.6)%, reflecting a He molar fraction of 14.9(-2.2)(+1.7)%, assuming the molar fraction of CH4 to be 2% in the troposphere. This He abundance is consistent with one derived from analysis of joint Voyager-2 IRIS and radio occultation experiment data, a technique whose accuracy has recently been called into question. For a disk average, the para-H-2 fraction is found to be no more than similar to 1.5% different from its equilibrium value, and the N-2 mixing ratio is probably less than 0.7%. The composite spectrum is best fit by invoking a CH4 ice condensate cloud. Using a Mie approximation to particle scattering and absorption, best-fit particle sizes lie between 15 and 40 mum. The composite spectra are relatively insensitive to the vertical distribution of the cloud, but the particle scale height must be greater than 5% of the gas scale height. The best models are consistent with an effective temperature for Neptune that is 59.5 +/- 10.6 K, a value slightly lower than derived by the Voyager IRIS experiment-possibly Neptune's mid- and far-infrared emission has changed during the seven years that lie between its encounter with Voyager 2 and the first spectra taken of this planet with ISO. The model spectra are also ostensibly lower than ground-based observations in the spectral range of 17-24 mum, but this discrepancy can be relieved by perturbing the temperature of the lower stratosphere where the LWS spectrum is not particularly sensitive, combined with the uncertainty in the absolute calibration of the ground-based measurements. (C) 2003 Elsevier Inc. All rights reserved.