Balancing the energy budget of short-period giant planets: evidence for reflective clouds and optical absorbers

Abstract

We consider fifty transiting short-period giant planets for which eclipse depths have been measured at multiple infrared wavelengths. The aggregate dayside emission spectrum of these planets exhibits no molecular features, nor is brightness temperature greater in the near-infrared. We combine brightness temperatures at various infrared wavelengths to estimate the dayside effective temperature of each planet. We find that dayside temperatures are proportional to irradiation temperatures, indicating modest Bond albedo and no internal energy sources, plus weak evidence that dayside temperatures of the hottest planets are disproportionately high. We place joint constraints on Bond albedo, $A_{B}$, and day-to-night transport efficiency, $\varepsilon$, for six planets by combining thermal eclipse and phase variation measurements (HD 149026b, HD 189733b, HD 209458b, WASP-12b, WASP-18b, and WASP-43b). We confirm that planets with high irradiation temperatures have low heat transport efficiency, and that WASP-43b has inexplicably poor transport; these results are statistically significant even if the precision of single-eclipse measurements has been overstated by a factor of three. Lastly, we attempt to break the $A_{B}$-$\varepsilon$ degeneracy for nine planets with both thermal and optical eclipse observations, but no thermal phase measurements. We find a systematic offset between Bond albedos inferred from thermal phase variations ($A_{B} \approx 0.35$) and geometric albedos extracted from visible light measurements ($A_{g} \approx 0.1$). These observations can be reconciled if most hot Jupiters have clouds that reflect 30-50 per cent of incident near-infrared radiation, as well as optical absorbers in the cloud particles or above the cloud deck.