2010PASP..122.1341B

We discovered evidence for a possible additional transiting planet in the NASA EPOXI observations of the known M dwarf exoplanetary system GJ 436. Based on an ephemeris determined from the EPOXI data, we predicted a transit event in an extant Spitzer Space Telescope 8 µm data set of this star. Our subsequent analysis of those Spitzer data confirmed the signal of the predicted depth and at the predicted time, but we found that the transit depth was dependent on the aperture used to perform the photometry. Based on these suggestive findings, we gathered new warm Spitzer observations of GJ 436 at 4.5 µm spanning a time of transit predicted from the EPOXI and Spitzer 8 µm candidate events. The 4.5 µm data permit us to rule out a transit at high confidence, and we conclude that the earlier candidate transit signals resulted from correlated noise in the EPOXI and Spitzer 8 µm observations. In the course of this investigation, we developed a novel method for correcting the intrapixel sensitivity variations of the 3.6 and 4.5 µm channels of the Infrared Array Camera (IRAC) instrument. We demonstrate the sensitivity of warm Spitzer observations of M dwarfs to confirm sub-Earth-sized planets. Our analysis will inform similar work that will be undertaken to use warm Spitzer observations to confirm rocky planets discovered by the Kepler mission.