The 3.6-8.0 μm Broadband Emission Spectrum of HD 209458b: Evidence for an Atmospheric Temperature Inversion

TLDR: In this article, the authors estimate the strength of the bandpass-integrated thermal emission from the extrasolar planet HD 209458b at 3.6, 4.5, 5.8, and 8.0 μm using the Infrared Array Camera (IRAC) on the Spitzer Space Telescope.

Abstract: We estimate the strength of the bandpass-integrated thermal emission from the extrasolar planet HD 209458b at 3.6, 4.5, 5.8, and 8.0 μm using the Infrared Array Camera (IRAC) on the Spitzer Space Telescope. We observe a single secondary eclipse simultaneously in all four bandpasses and find relative eclipse depths of 0.00094 ± 0.00009, 0.00213 ± 0.00015, 0.00301 ± 0.00043, and 0.00240 ± 0.00026, respectively. These eclipse depths reveal that the shape of the inferred emission spectrum for the planet differs significantly from the predictions of standard atmosphere models; instead, the most plausible explanation would require the presence of an inversion layer high in the atmosphere leading to significant water emission in the 4.5 and 5.8 μm bandpasses. This is the first clear indication of such a temperature inversion in the atmosphere of a hot Jupiter, as previous observations of other planets appeared to be in reasonably good agreement with the predictions of models without such an inversion layer.

Figures

Fig. 3.—Predicted emission spectrum for HD 209458b from Burrows et al. (2006). Squares show this spectrum integrated over the Spitzer bandpasses (response functions for these bandpasses are shown at the bottomof plot [dotted lines], scaled by a factor of 0.001), and circles show the estimated eclipse depths in these bandpasses. The 24 mpoint is taken fromDeming et al. (2005). The two dashed lines give the planet-star flux ratio for the case in which the planet is a perfect blackbody with a temperature of either 1500 or 1900 K and the spectrum of the star is calculated from a model (available at http://kurucz.harvard.edu/stars/hd209458). Note that we have chosen not to plot the 7.5Y12.2 m spectrum measured by Richardson et al. (2007); this is because the spectrum was scaled to match a preliminary value for our 8.0 m eclipse depth and thus does not contain independent information about the absolute strength of the emission from the planet at these wavelengths.

TABLE 1 Best-Fit Eclipse Depths and Times

Fig. 2.—Top to bottom: Secondary eclipse of HD 209458b on UT 2005 November 27, observed at 3.6, 4.5, 5.8, and 8.0 m,with best-fit eclipse curves overplotted. Data have been normalized to remove detector effects (see discussion in xx 2.1 and 2.2) and binned in 7 minute intervals.

Fig. 1.—Top to bottom: Secondary eclipse of HD 209458b on UT 2005 November 27, observed at 3.6, 4.5, 5.8, and 8.0 m, binned in 7 minute intervals and normalized to 1. The overplotted curves show the best-fit corrections for detector effects (see xx 2.1 and 2.2).