Hot nights on extrasolar planets: mid-infrared phase variations of hot Jupiters

We present results from Spitzer Space Telescope observations of the mid-infrared phase variations of three short-period extrasolar planetary systems: HD 209458, HD 179949 and 51 Peg. We gathered Infrared Array Camera (IRAC) images in multiple wavebands (3.6 or 4.5 and 8 mu m) at eight phases of each planet's orbit. We find the uncertainty in relative photometry from one epoch to the next to be significantly larger than the photon counting error at 3.6 and 4.5 mu m. We are able to place 2 sigma upper limits of only similar to 2 per cent on the phase variations at these wavelengths. At 8 mu m, the epoch-to-epoch systematic uncertainty is comparable to the photon counting noise and we detect a phase function for HD 179949 which is in phase with the planet's orbit and with a relative peak-to-trough amplitude of 0.001 41 (33). Assuming that HD 179949b has a radius R-J < R-p < 1.2R(J), it must recirculate less than 21 per cent of incident stellar energy to its night side at the 1 sigma level (less than 26 per cent at the 2 sigma level, where 50 per cent signifies full recirculation). If the planet has a small Bond albedo, it must have a mass less than 2.4M(J)(1 sigma). We do not detect phase variations for the other two systems but we do place the following 2 sigma upper limits: 0.0007 for 51 Peg and 0.0015 for HD 209458. Due to its edge-on configuration, the upper limit for HD 209458 translates, with appropriate assumptions about Bond albedo, into a lower limit on the recirculation occuring in the planet's atmosphere. HD 209458b must recirculate at least 32 per cent of incident stellar energy to its night side, at the 1 sigma level (at least 16 per cent at the 2 sigma level), which is consistent with other constraints on recirculation from the depth of secondary eclipse depth at 8 mu m and the low optical albedo. These data indicate that different hot Jupiter planets may experience different recirculation efficiencies.