With ACS/HRC coronagraphy, we have achieved the direct detection of a planet candidate in F606W and F814W around a bright nearby star with a debris belt. The planet candidate lies 18 astronomical units interior to the dust belt and we detect counterclockwise orbital motion in observations separated by 1.75 years. The candidate has mass no greater than three Jupiter masses based on an analysis of its luminosity and the dynamical argument that a significantly more massive object would disrupt the dust belt. Using recent model predictions for 100-300 Myr old planetary atmospheres, the planet candidate has a temperature of ~400 K and a mass 1.6 - 3.4 M_J. Variability at optical wavelengths suggests additional sources of luminosity such as H-alpha emission or the episodic accretion of cometary material. A key surprise is that the planet candidate is NOT detected in Keck adaptive optics observations at 1.6 microns. Two model atmospheres predict a flux a few times greater than our detection limit, though the model predictions disagree with each other by a factor of five due to theoretical uncertainty in the strength of CH4 vibrational bands. These models predict the strongest emission centered on the F110W passband of NICMOS such that the F814W - F110W color will be red. Here we propose follow-up NICMOS observations to verify that the emission observed in F814W is due to the emergent flux from passive cooling of the planet, as opposed to other explanations, such as reflected light from a Saturn analog with a circumplanetary debris disk that would produce a bluer F814W - F110W color. Additional deep images in and and out of the 1.14 micron water trough using NIC1 narrowband filters will test whether or not the emission is produced from the passive cooling of a young massive planet.
No bigger than 3J. wow.